Adam Matthew Hart

A Constructivist Approach to Developing

Interactive Digital Technology for
Musical Learning

A thesis presented in partial fulfilment of the

requirements of the University of Salford for the degree
Doctor of Philosophy

School of Arts & Media

University of Salford, UK

September 2018
 Table of Contents

List of Figures vi
List of Tables xii
Acknowledgments xiii
Declaration xiv
Abstract xv

1. Introduction 1
1.1 About the Project 2
1.1.1 Overview 2
1.1.2 The Relevance of Constructivism 3
1.1.3 Validation and Dissemination 5
1.1.4 Limitations 6
1.1.5 Structure and Approach 8
1.1.6 Ethics 10
1.2 About the Researcher 10
1.2.1 Qualifications and Experience 10
1.2.2 Theoretical Orientations 11

2. Research Context 1: Constructivism & Creativity 13

2.1 What is Constructivism? 14
2.1.1 Overview 14
2.1.2 Emergence 14
2.1.3 Cognitive and Social Constructivism 16
2.1.4 Constructivism and Digital Technology 19
2.2 Constructivism and Music Education 22
2.2.1 Curriculum and Instruction 22
2.2.2 Musical Creativity in the Classroom 28
2.2.3 Supporting Musical Creativity with Digital Technology 31

i
 2.3 Summary 39

3. Research Context 2: Environment & Interaction 42

3.1 Digital Technology in Creative Learning Environments 43
3.1.1 Visible Learning 43
3.1.2 Tangible (‘Hands-on’) Learning 45
3.1.3 Augmented Learning 48
3.2 Digital Technologies in the Music Classroom 49
3.2.1 Overview 49
3.2.2 Listening and Appraising 51
3.2.3 Performance 52
3.2.4 Composition 54
3.3 Learning from Games 56
3.3.1 Sandbox Games 56
3.3.2 Educational Applications 58
3.4 Summary 59

4. Methodology 63
4.1 Choosing a Methodology 64
4.1.1 Considerations 64
4.1.2 Software Development as Educational Research 65
4.1.3 Grounded Theory 67
4.1.4 Mixed Methodology 67
4.2 Research Tools 68
4.2.1 Programming Platform 68
4.2.2 Additional Software 69
4.2.3 Hardware 69
4.3 Research Model 70
4.3.1 Timeline 70
4.3.2 Data Collection and Analysis 74

ii
5. Phase 1: Initial Design 78
5.1 Defining the Activity 79
5.1.1 Overview 79
5.1.2 Experiential Composition 79
5.1.3 Expected Outcomes 81
5.1.4 Domain Knowledge and Data Structures 82
5.2 Initial Development 84
5.2.1 Glossary 84
5.2.2 A Note on the Programming Process 85
5.2.3 Sequencing in GEM 86
5.2.4 ‘Painting’ Sound 87
5.2.5 Scale and Register 89
5.2.6 Dynamic Patching 89
5.2.7 Remote Communication 90
5.2.8 Designing Voices 91
5.2.9 Expanding the User Interface 92
5.2.10 Alternative Models 93
5.3 Summary 93

6. Phase 2: Development 96
6.1 Exhibition at University A, September 2015 97
6.1.1 Context 97
6.1.2 Method 98
6.1.3 Results 98
6.2 Exhibition at University B, November 2015 101
6.2.1 Context 101
6.2.2 Method 102
6.2.3 Results 104
6.3 Focus Group at School A, January 2016 105
6.3.1 Context 105
6.3.2 Method 106

iii
 6.3.3 Results 107
6.4 Scaffolding with Simple Structures 111
6.4.1 Tune-blocks 111
6.4.2 Editing Tune-blocks 113
6.4.3 Other Developments 114
6.5 Focus Group at School A, May 2016 115
6.5.1 Method 115
6.5.2 Results 116
6.6 Summary 119

7. Phase 3: Application, Refinement, Expansion 124

7.1 Refinement of the Interface 125
7.2 Storyboard Composition at School B, June 2016 127
7.2.1 Context 127
7.2.2 Method 127
7.2.3 Results 129
7.3 Melodic Composition at School C, July 2016 140
7.3.1 Context 140
7.3.2 Method 140
7.3.3 Results 141
7.4 Performance and Composition with iPads at School B, December 2016 146
7.4.1 Context 1 146
7.4.2 Method 1 148
7.4.3 Results 1 152
7.4.2 Context 2 155
7.4.3 Method 2 155
7.4.4 Results 2 158
7.5 Sound Objects and Keynote Sounds 164
7.6 Storyboard Composition at School D, March 2017 172
7.6.1 Context 172
7.6.2 Method 173

iv
 7.6.3 Results 174
7.7 Performance and Composition with iPads at School C, June-July 2017 184
7.7.1 Context 1 184
7.7.2 Method 1 187
7.7.3 Results 1 187
7.7.4 Context 2 190
7.7.5 Method 2 190
7.7.6 Results 2 191
7.8 Theoretical Summary 197

8. Conclusion 199
8.1 Summary of Outcomes 200
8.1.1 Practical Outcomes 200
8.1.2 Theoretical Outcomes 200
8.1.3 Other Outcomes 202
8.2 Research Questions 202
8.2.1 Question 1: Interactive Digital Technologies and ‘Musical Behaviour’
202
8.2.2 Question 2: Sandbox Environments 205
8.2.3 Question 3: Constructivism and Design 207
8.2.4 Question 4: Creative Expression and the Musically Untrained 208
8.2.5 Recommendations 210
8.3 Critical Evaluation 211
8.3.1 Scope 211
8.3.2 Methodology 212
8.3.3 Design and Development 213
8.3.4 Application 215
8.3.5 Originality 216

Bibliography 217

v
 List of Figures
The following is a list of the figures contained in this document. At the head of each chapter
is a figure which serves to represent, in some way, the portion of the narrative which it
introduces. All other figures are illustrative of points within the text, and are referenced
therein. Unless stated, figures (including screen recordings) are created by the author. Clip-
art images are all public domain. Permission has been sought for any images which may be
subject to copyright restrictions, and every effort has been made to credit the original
sources of images. If you see any material which may be subject to copyright or uncredited,
please contact the University of Salford Institutional Repository at www.usir.salford.ac.uk

1. Introduction
1.1 – Word cloud generated from the text of chapter 1 as an indication of the themes of the
project 1

2. Research Context 1: Constructivism & Creativity

2.1 – Photograph of a child’s toy building brick, a symbol of constructivism and creativity
13
2.2 – Kolb’s cycle of experiential learning. Adapted from Kolb 1984 18
2.3 – Describing a square in LOGO. Retrieved 14 November 2015 from
www.teacherlearnstocode.com/2014/10/13/logo-something-to-get-excited-about/ 19
2.4 – Bloom’s 21. Retrieved 10 October 2016 from
www.plpnetwork.com/2012/05/15/flipping-blooms-taxonomy/ 24
2.5 – Exploding Art jam2jam. Retrieved 14 October 2015 from
www.explodingart.com/jam2jam/jam2jam/Home/Home.html 32
2.6 – Drumsteps. Retrieved 14 October 2015 from
www.bbc.co.uk/music/childrens/games/drumsteps/ 32
2.7 – Yelsi AG eJay 35

3. Research Context 2: Environment & Interaction

3.1 – Touchscreen tablets in a classroom music environment 42
3.2 – Generating a spiral structure in Scratch. Retrieved 20 September 2017 from
www.makelearn.org/2015/04/08/kids-and-coding-what-might-seymour-papert-say/ 44

vi
3.3 – Bloom 50
3.4 – Ars Nova Practica Musica 51
3.5 – Rising Software Auralia 51
3.6 – The sequence of progression in Charanga Music World, alongside the sequence of
musical development. Adapted from The Sequence of Musical Development: A Study of
Children’s Composition by K. Swanwick and J. Tillman, 1986. British Journal of Music
Education, 3(3), p.331. 52
3.7 – Assistive tools for composing and performing in Charanga 52
3.8 – Home Concert XTREME for iPad. Retrieved 19 December 2014 from
www.itunes.apple.com/us/app/home-concert-xtreme/id443017184?mt=8 53
3.9 – Yousicion. Retrieved 03 September 2017 from
www.youtube.com/watch?v=nlAXGJ8LuwM 53
3.10 – Synthesia. Retrieved 03 September 2017 from www.voicesinc.org/yousician-vs-
synthesia/ 53
3.11 – Avid Sibelius Ultimate 54
3.12 – Arranging clips in GarageBand 54
3.13 – Arranging drums in GarageBand 55
3.14 – Morton Subotnick’s Pitch Painter. Retrieved 19 December 2014 from
www.itunes.apple.com/us/app/mort-subotnicks-pitch-painter/id519738403?mt=8 55
3.15 – Impromptu 55
3.16 – Sim City. Retrieved 18 August 2018 from www.gambitmag.com/2016/10/simcity-
buildit-adds-lakes-rivers-forests-oh/ 56
3.17 – Minecraft 57
3.18 – Victorian House by Planet Minecraft forum member Paul_zero. Retrieved 07 August
2018 from www.planetminecraft.com/project/victorian-hospital-4091132/ 57
3.19 - A class of children, potentially across multiple schools, explore a teacher-generated
world in MinecraftEdu. Retrieved 19 December 2014 from
www.mrsboothsclass.weebly.com/building-bridges-minecraftedu/week-1-minecraftedu 58

4. Methodology
4.1 – Cutting cogs from coloured card 63
4.2 – Gantt chart of phases 1-3 70
4.3 - The research model for data collection, analysis and application. Studies across phases 2
and 3 are indicated, along with the emerging data types 73

vii
5. Phase 1: Initial Design
Note that most of these figures are screenshots of Pd patches from the earliest stages
development.
5.1 – An early design sketch from my notebook, 22 April 2015 78
5.2 – An example of graphic notation, where symbols represent musical dimensions. I
frequently used this particular example as a music teacher 80
5.3 – A suggested initial framework for mapping musical dimensions to graphic notations 81
5.4 – Data structures in Pd 82
5.5 – Solitude by Hans Christoph-Steiner. Retrieved 20 November 2015 from
www.at.or.at/hans/solitude/ 83
5.6 – A basic 16 step sequencer 86
5.7 – The canvas window in my earliest patches 87
5.8 – A sine wave oscillator and amplitude envelope, used as a sound source in early patches
87
5.9 – 'Painting' in the canvas window 87
5.10 – Using the mouse to create geos 88
5.11 – Text file showing X axis position, MIDI pitch and amplitude 88
5.13 – Dynamic creation of abstractions 90
5.14 – UDP messaging between patches 91
5.15 – VST patches made for Graphick Score 91
5.16 – The user interface of Graphick Score in August/September 2015 92
5.17 – The mixer panel 93
5.18 – MetroNotes 93

6. Phase 2: Development
Note that most of these figures are screenshots of Graphick Score in development, or of pupil
compositions using the programme. Any figures with pupil names also have an
accompanying video file of the piece, which can be found in the digital appendix.
6.1 – Processed photograph of the author presenting Graphick Score in November 2015 96
6.2 – Graphick Score in September 2015 97
6.3 – The canvas window in slendro tuning 101
6.4 – Exhibiting Graphick Score at University B, November 2015 102
6.5 – Keyboard commands printed onto a mouse-mat 103
6.6 – The proposed design for SoundWorlds 103

viii
6.7 – SoundWorlds uses the Leap Motion to allow the user to ‘grab’ and ‘move’ 3-
dimensional geos, altering their musical parameters 105
6.8 – The Leap Motion uses infra-red to map hand motions to a XYZ axis 105
6.9 – Different geos; simple geometric shapes, complex shapes, 3-dimensional .mod files 106
6.10 – Alice 108
6.11 – Jessica 108
6.12 – Bill 109
6.13 – Hattie 110
6.14 – A tune-block for the song Oh When the Saints 111
6.15 – A sequence generated using the tune-block shown in Figure 6.14 112
6.16 – Block chords and broken chords 112
6.17 – Fruit rhythms making the rhythmic sequence for ‘pat-a-cake pat-a-cake baker’s man’;
coconut coconut apple pear 113
6.18 – Fruit rhythms 113
6.19 – Rotations of the tune-block shown in Figure 6.14 114
6.20 – The instrument menu 114
6.21 – A message directs the user to select a channel 114
6.22 – Graphick Score in May 2016 115
6.23 – Lara 116
6.24 – Lee 116
6.25 – Hattie 117
6.26 – Jessica 117
6.27 – Jessica 118
6.28 – Arranging a block of ‘sound painting’ 122

7. Phase 3: Application, Refinement, Expansion

Note that most of these figures are screenshots of pupil compositions made with Graphick
Score. Any figures with pupil names also have an accompanying video file of the piece, which
can be found in the digital appendix.
7.1 – Graphick Score as published in The London Review of Education (Hart 2017a) 124
7.2 – Graphick Score in June 2016 125
7.3 – The tune-block for Somewhere over the Rainbow is changed from a major to a minor
key. Arlen, H. & Harburg, Y. (1939) Over the Rainbow. London: Francis 125
7.4 – Keyboard stickers for operating the programme 126

ix
7.5 – Animal tempos, moving from slow to fast; tortoise, man, hare, cheetah 126
7.6 – The instrument menu 126
7.7 – The example story, Incy Wincy Spider, told with ‘sound painting’ 128
7.8 – Carly 129
7.9 – Imogen 130
7.10 – Hayley 130
7.11 – Kody 131
7.12 – Rob 134
7.13 – Renee 135
7.14 – Barry 135
7.15 – Emma 136
7.16 – Aisha 137
7.17 – Millie 137
7.18 – Ronan 138
7.19 – Harvey 138
7.20 – The demonstration from the session 141
7.21 – Natasha 142
7.22 – Ethan 142
7.23 – Rickie 142
7.24 – Sammy 144
7.25 – Lewis 145
7.26 – The verse of Little Donkey. Boswell, E. (1959) Little Donkey. London: Chappell & Co.
Ltd. 149
7.27 – The verse and chorus of Little Donkey in iPad tune-blocks 150
7.28 – The chord interface 150
7.29 – Rhythmic variations 154
7.30 – Melodic variations with the same rhythmic structure 154
7.31 – Melodic variations with the same harmonic structure 155
7.32 – The Rainbow Keyboard 156
7.33 – Arranging tune-blocks into a complete structure 158
7.34 – Composing their own carols 160
7.35 – Original composition by group 161
7.36 – Original composition by group 161
7.37 – Original composition by pupil 161

x
7.38 – Graphick Score in 2017. A version with a white canvas window was requested by the
London Review of Education due to printing considerations 164
7.39 – Kurtág’s Alapelemek from Játékok vol. 6. Retrieved 17 February 2018 from
www.junttu.net/_/_/jatekok_1_kirja.html. Kurtág, G. (1997) Játékok vol. 6. Budapest: Editio
Musica Budapest 165
7.40 – Group 1 177
7.41 – Group 2 178
7.42 – Group 5 179
7.43 – Group 7 180
7.44 – Group 6 181
7.45 – Group 8 183
7.46 – Smart keyboard in GarageBand 185
7.47 – Sample-It 186

8. Conclusion
8.1 – Sequence representing creative kinaesthetic actions; sketch, sculpt, splash, swipe? 199
8.2 - The theoretical codes emerging from the studies and analytical process, and their
relevance to the four research questions 201

xi
 List of Tables
The following is a list of tables used in this document. Clip-art images are all public domain.

5. Phase 1: Initial Design

5.1 – Scale degrees 89
5.2 – Basic musical dimensions represented in Graphick Score 94

7. Phase 3: Application, Refinement, Expansion

7.1 – Table of sound objects 166
7.2 – Table of storyboard images and sound files 167

xii
 Acknowledgements

Research depends on community, and as such, I owe a tremendous debt of gratitude to the
individuals and institutions that have supported me in so many ways throughout this project.
This research is funded by the Arts and Humanities Research Council and the North West
Consortium Doctoral Training Partnership. I will always be proud to have embarked upon
my research career with the support of these networks.

I could not have commenced or completed this project without the guidance and friendship of
Professor Alan Williams, and I greatly appreciate the time and thought he has given over the
past several years. Also, I thank Professor Steve Davismoon for continuing to be a source of
support and advice, even after moving on to another university. I am very much obliged to all
of the staff and students at the University of Salford who have had such a lasting influence on
my life, and who are too numerous to name here. I hope to continue to work with all of these
people for many years to come.

I will always be grateful to my network of schools, teachers and other educational

professionals for training me, employing me, and letting me conduct my research in their
classrooms. Teaching is the most important job in the world, and I count myself exceptionally
lucky to have such a varied and valuable time in education among such inspirational people.

There are a number of online communities who have proved to be vital resources throughout
this research. In particular, I would like to thank the endlessly insightful and helpful members
of the Pure Data online community for always having an answer to technical queries,
however obscure or incomprehensible, and for the profound influence that their work has had
upon my own.

In this and everything else I do, I thank all of my family and friends for their support.

xiii
 Declaration

This thesis is submitted under the University of Salford regulation for the award doctor of
philosophy. Unless otherwise stated in the text, I hereby declare that the contents of this
thesis are the result of my own work, and that no part of it has been submitted in support of
any application for other degree or qualification at this or other institutions of higher learning.

…………………………………………

Adam Matthew Hart

September 2018

xiv
 Abstract

This study explores the potential added value of interactive digital technology to early-years
music education through the development of software artefacts and their application in the
primary classroom. The aims of the study are to investigate and identify new designs and
approaches for classroom music-making with technology, to integrate educational theory and
software development in this field, and to promote musical creativities at the individual and
social level. Following a constructivist-interpretivist approach, a range of qualitative
methods are applied in the pursuit of these goals. The emerging educational methodology of
Software Design as Research (SoDaR) is combined with an emergent model of Grounded
Theory, adapting a new methodological approach for the unique purposes of this project, as
well as examining how this model could be applied to similar projects.

Findings can be summarised as follows:

i. Rudimentary and familiar actions can be harnessed, through digital interfaces, as a
means of affording creative expression in novice musicians
ii. Meaning-making structures are as vital as music-making structures in such activities,
allowing creative ownership and communication to emerge through basic narrative
elements
iii. Through this process of interactive and representational engagement, digitally-
supported environments can allow learners to construct musical concepts for
themselves, at the individual and social level.

This research has implications for music education: Appropriate digital technologies and
approaches can support the process of musical learning in social settings. Furthermore, such
technologies, and the framework for their use in the classroom, can be developed through a
process grounded in empirical educational research. Closer working links between
educational practice and research, and the development and testing of digital resources, are
recommended to ensure that a meaningful and active music education, and the valuable
opportunities for personal, social and creative development and expression that this affords,
is available and accessible for all learners.

xv
 1. Introduction

Figure 1.1

This chapter will frame the narrative of this project, which concerns two concurrent
trajectories of development and discovery: Firstly, the development of a software artefact for
musical learning and the findings which emerged from its application in the classroom and
secondly, the development of my practice as a programmer, researcher and music teacher,
and the findings which emerged as I triangulated these approaches into a single mode of
practice. I will provide a preliminary overview of the progression of this project from
proposal to completion, introduce my experiences, aims and theoretical orientations as the
conductor of this research, and explain the form and structure of this thesis.

1
1.1 About the Project

1.1.1 Overview
The aim of this research project is to investigate an approach for achieving musical learning
outcomes with digital technology. ‘Musical learning’ is defined, within this study, as learning
which enhances musical knowledge and skill arising from diverse modes of interaction and
engagement in creative environments. This is a practice-led project, one in which creative
practice is ‘in itself is a form of research and generates detectable research outputs’ and
where this practice ‘can lead to specialised research insights which can then be generalised
and written up as research’ (Smith & Dean 2009, p.5). In this instance, the practice concerned
is the development and application of bespoke digital technology for musical learning, the
insights generated from which are used to answer a set of research questions:
 To what extent have interactive digital technologies hitherto succeeded in
encouraging musical behaviour in the classroom?
 What conditions prompt musical learning in ‘sandbox’ environments?
 How can constructivist educational theory be applied to software design as a means
of facilitating creative musical learning?
 How might interactive digital technology be harnessed to facilitate creative
expression in the musically untrained?
As this is an exploratory project, where outcomes are directed, at least in part, by the changes
in this practice as it develops and responds to emergences, many of the more specific details
are arrived at during the narrative of this thesis. For example, the questions posed here, and
answered in the conclusion, are not the questions I was asking at the start of the research
project. At that time, I had a single question:
 How can digital technology be used to facilitate musical learning?
Compared to those which emerged from it, this question may be smaller, but it is certainly
not simpler1. It demands answers to a series of qualifying questions: With what kind of digital
technology are we concerned? What constitutes musical learning? Why are these two ideas
necessarily mutually relevant? How can I assume that we should we using digital technology
for this purpose? These details, along with the research questions, are addressed in

1
This distinction is present in the work of Bamberger (1996, p.42; see 6.6) and became a prominent influence in
this research project.

2
subsequent chapters. I will begin, in this chapter, by clarifying my decision to adhere to a
specific theory of learning.

1.1.2 The Relevance of Constructivism

In simple terms, constructivism is the idea, originally theorised by Piaget, that individuals
develop their understanding (of anything) from interaction between their experiences and
their ideas. As such, it has significant implications for the nature of education and
instructional design, and is often used as an umbrella term encompassing a wide range of
related pedagogies. The various aims, assumptions, adherents and applications of
constructivism will be explored in detail in chapter 2. Here, I will just explain how this
particular theory came to feature so prominently in my research.
The original proposal for this project, entitled Educational Applications of Interactive
Music Technology, drew from my experiences of music education, as both a student and a
teacher, and my continual interest in using digital technology to support learning. At this
time, the Office for Standards in Education (OFSTED) had recently published two reports
critiquing the lack of musical opportunities in the classroom, as well as insufficient ICT-
related advancements in musical learning (OFSTED 2012, p.54), and efforts to restructure
music education provision which were largely failing to address a fundamental lack of music
teachers and resources (OFSTED 2013). In my research proposal, I highlighted potential
solutions to these issues, and set out a plan for developing new resources for musical learning
and testing them in learning environments. Having spent the preceding five years working in
education in various capacities, including training and working as a secondary school music
teacher, I wanted to ensure a strong educational basis for my research, and so my first
semester as a PhD student was mostly devoted to reading education journals. Consequently, I
read much on constructivism. The dynamics of active experience and reflection to which it
referred resonated with my classroom experience, the outstanding teaching and learning
which I had observed, and what I understood to be indicative of best practice in education.
Although it has informed numerous pedagogies, constructivism is not by itself a pedagogy
but an ontology. If we subscribe to this ontology as a way of supporting teaching, by defining
learning structures, planning activities and promoting reflection, then could we not also apply
it as a measure of the resources we are using for teaching, to support their design,

3
development and application? In my research, this specifically concerns digital resources for
musical learning.
A disparity between the modes of practice I proposed to combine in this project
became apparent. When planning a lesson, my training and experience led me to approach it
as a constructivist, allowing the students to apply their own skills and ideas and arrive at
outcomes in their own way, while also ensuring that this was evaluated against the learning
objectives. When programming a musical interface, I did not approach the task with the same
attention to reflexivity. My patches were typically based on a single and fixed initial idea, and
consequently had a single mode of use and fixed outcome, with little room for variability of
input and output. As an example, here is a description of a computer programme, or patch,
which I completed in 2013 as part of a collaborative project at the University of Salford with
Professor Alan Williams, who would become my supervisor for this research. The project
was an installation entitled Sing to Me and I’ll Tell You My Story:

A vocal melody is played through speakers, inviting listeners to ‘sing to me and I’ll
tell you my story’. A listener then sings the melody back into a microphone, and the
patch analyses the frequency of the audio signal to check if it falls within a given
frequency range that can be perceived as an ‘accurate note’. A message on a screen
informs the listener how many notes they got right, and if they accurately sing all of
the notes, a pre-recorded audio file plays, ‘telling’ the listener a story.

Because the patch has an educative element – learning to accurately sing a melody in order to
receive the reward of a story – it came to mind when writing my proposal of how interactive
digital technology might support music education. However, though the process works well
as a game or installation, it is not a process that I would apply in a music lesson, for a number
of reasons: Firstly, the outcome, from the perspective of the listener, is not the ability to
accurately sing the melody, but the story it triggers, so they do not necessarily see the value
in any learning that may have taken place. Secondly, feedback is minimal, and is only given
when the melody is sung incorrectly, as when sung correctly, triggering the story, the melody
serves no further purpose for the listener. Thirdly, this involves no repetition or practice of
the correctly sung melody; once they have sung it accurately, they are silently listening to the
story. Fourthly, there is only one way in which the listener can approach the interface; they
sing the correct melody and are told a story, a single mode of input and a fixed output.
Looking at the patch as a teacher and not a programmer, I see these deficiencies.

4
 During the first semester of this project, I realised that the approach taken and
outcomes generated with this kind of patch contrasted with that of my lesson planning and
that I wanted to avoid this contradiction when exploring opportunities for musical learning in
digital technology. Therefore, the measure I set for myself in this project – to achieve an
ontological reconciliation between my teaching and programming practice, to develop this
combined practice, and to use it to produce research insights – was to approach it from a
constructivist perspective. Consequently, instead of producing several applications with
single modes of input and output, I focused my attention on a single versatile application with
multiple modes of use. This artefact, Graphick Score, is provided in the digital appendix
which accompanies this thesis.

1.1.3 Validation and Dissemination

My research proposal was awarded a three-year scholarship by the Northwest Consortium
Doctoral Training Partnership (NWCDTP) with funding from the Arts & Humanities
Research Council (AHRC) and I was accepted as a candidate for the award of PhD Music at
the University of Salford. I was supervised by Professor Alan Williams and Professor
Stephen Davismoon of the School of Arts and Media. I started work on this project in the
autumn of 2014, completing this thesis in 2018. This research has been widely disseminated
during this time, through its application in classrooms throughout the North West of England,
in a variety of national and international conferences and exhibitions, in two published peer-
reviewed papers and on social media. I have presented this research at each of the annual
NWCDTP postgraduate research conferences and Salford Postgraduate Annual Research
Conferences (SPARC) since commencing this project, winning the award for ‘Best
Interactive Presentation’ for two consecutive years at the latter. Other major conferences at
which I presented include the International Association for the Study of Popular Music
(IASPM) and the Nordic Educational Research Association (NERA).
Just prior to the third year of this research project, Professor Davismoon left his
position at the University of Salford to become Head of Performing Arts at Edge Hill
University, but remained in my supervisory team. As the university has a noted education
department, and given my previous links to the institution (my PGCE was awarded by Edge
Hill University) this presented an opportunity to expand my network with regard to education
research. I arranged consultations with professors of education Tim Cain at Edge Hill
University and Jonathan Savage at Manchester Metropolitan University, both of whom are

5
cited as prominent contributors to the field of music education research in my literature
review, and was able to greatly expand my knowledge and practice with respect to this field
during this time. This is reflected in my outputs within international conferences and journals
within the field of education research in the latter half of the third year.
This research was presented at three major international conferences on education; the
45th and 46th international congresses of NERA in Copenhagen and Oslo respectively, as well
as the 19th annual conference on education at the Athens Institute of Education and Research
(ATINER). Some of the research conducted for this project has already been published in two
open-access peer-reviewed academic journals. Towards an Effective Freeware Resource for
Music Composition in the Primary Classroom (Hart 2017a), describing many of the central
research outputs of this project, was published in the London Review of Education by the
Institute of Education at University College London. In what was a new approach for the
journal, the article was published with embedded hyperlinks to video examples of music
made by pupils using Graphick Score, which were uploaded to the journal’s YouTube
account. This novel approach, raising further possibilities for linking to interactive digital
media in open-access publishing, was taken by the editor to more effectively disseminate the
evidence I provided; the videos and software artefact which are submitted in the digital
appendix to this document. I also had The Tablet as a Classroom Musical Instrument (Hart
2017b), detailing additional research undertaken for this project, accepted for publication in
the Athens Journal of Education, as an outcome of my attendance at ATINER.

1.1.4 Limitations
In order to arrange a consistent series of case studies and produce sufficient results of a
reliable quality, this research is limited within certain boundaries. For example, research
focuses on music composition at key stage 2, as these were identified as areas on which to
focus during the initial stages of the project. It will be useful to extend this study to a wider
age group and range of activities in a follow-up project (see 4.3.2). Limitations were also
encountered as to where, when and for how long research could take place. Most of the
schools who were contacted were very accommodating, and glad to offer a novel musical
experience to the pupils, but an appropriate day had to be arranged sufficiently in advance,
and several schools had to cancel or postpone due to various circumstances. Due to busy
schedules and the pressures of examinations and other projects, most schools had specific

6
periods in mind, such as the weeks prior to the winter or summer breaks. Also, only schools
in the North West of England were contacted for this study, due to the logistics of travelling
wider distances for short studies, and because most of my school contacts are within this
region.
Sessions varied in length, from an entire day of studies, to afternoon sessions of
around 2 hours and, in the earlier stages, short ‘taster’ sessions of around 20 minutes.
Ultimately, the length of each study, and the number of pupils involved, was the decision of
the class teacher, who was in turn subject to various limitations, though many were happy to
accommodate requests. This meant that, in many cases, pupils only had a short window
within which to become acquainted with the digital resources used, and the outputs of such
sessions are limited in relation to those of longer sessions. It should also be noted that, due to
the investigative nature of the research project, and variable availability of equipment that I
was able to obtain from the school, university or other sources, the case studies often differ in
terms of lesson plan and resources. These differences are highlighted in the account of each
case study.
Lessons were primarily conducted by me acting as researcher and teacher, though the
class teacher was involved for some of the sessions. Most teachers were naturally more
receptive to the idea of a visiting music specialist taking the class than to the idea of
delivering a lesson using an unfamiliar resource themselves, though later sessions did see an
increased involvement (see 7.6). As such, I decided to focus mainly on the development of
these resources and student interaction with them rather than their use by the teacher, though
later research should certainly focus on this too. By acting as the teacher, I came to recognise
certain other limitations. My own training is largely in secondary school music, though I do
have previous experience working with younger children. Consequently, it must be
acknowledged that the pupils are working with someone who has a different set of skills from
their usual class teacher – a lack of personal knowledge of the pupils themselves, as well as
less pedagogical experience, but the additional knowledge of specialist musical training – and
the dynamic will not be that of a typical lesson. I must also acknowledge my own biases and
orientations with regard to teaching practice, educational theory, and musical ability. All of
these considerations are taken into account within this thesis.
Finally, it will be useful to extend this research to examine a greater number of factors
which might influence musical creativity. Some details, such as gender, ability, and musical
training, were known or became apparent, while other information was not available, either
due to ethical considerations regarding sensitive data such as medical or pastoral

7
requirements, or other logistical factors. As an initial study into a relatively new area of
research, I hope that this work generates useful findings which lead to a more substantial
body of research, in my future work and that of other academics working within similar
fields.

1.1.5 Structure and Approach

The conclusions derived from this project have been generated by two modes of research,
each supporting the other. A theoretical examination of educational research, focusing on
music and technology, but drawing from other possible areas of relevance, provides the initial
ideas, substantiates explorations, and gives validation to findings. This inquiry is primarily
covered by the second and third chapters of this document, which review the existing
literature and research in the field of educational music technology, as well as investigating
some technologies with hitherto untapped potential for music education. The second mode of
research is practical, involving the design, development and evaluation of software artefacts.
It is through this inquiry that the initial ideas are tested, honed and expanded upon. The
methodology is outlined in chapter 4, while my programming practice and field research in
the classroom is presented over chapters 5, 6 and 7. The conclusions emerging from this
research are summarised in the final chapter.
The digital appendix, which accompanies this thesis, contains two main folders. The
first, entitled ‘Graphick Score’2 is the principal software artefact, along with a text file of
instructions for installation and operation, as well as any auxiliary files required for it to
function. The second folder, entitled ‘Classroom Research’, contains the outputs of classroom
studies (see chapters 6 and 7) in a series of sub-folders. Each of these folders corresponds to
the study name given in the chapter (e.g. ‘Storyboard Composition at School B, June 2016’,
see 7.2) while the file names correspond to the (anonymised) pupil names used when
presenting these findings (e.g. ‘Carly’, Figure 7.7). These are audio-visual files of the pupil
compositions made during these studies; as such, the reader can use the appendix to view any
pupil work which is referenced within this document.

2
©Copyright 2017 Hart . All software artefacts presented within this thesis or the accompanying digital
appendix are distributed under the terms of the Creative Commons Attribution Non-Commercial License, which
permits distribution and modification of the work for non-commercial purposes, providing the author is credited.
The outputs of this research are intended for educational purposes, in the spirit of open-source development and
distribution.

8
 This project can be characterised as qualitative, case-study led educational research
undertaken within an interpretivist paradigm (see O’Donoghue 2006). The interpretivist
position, common within social science research, is that individuals and society influence one
another in a meaningful way, and that the researcher should seek to understand this process of
mutual influence rather than any quantifiable absolute with which the positivist researcher, by
contrast, is concerned. Therefore, interpretivist research focuses on the behaviour of
individuals within specific contexts and environments. To a greater extent, this research
works towards the definition of theory from active practice and observation, as opposed to
research models which test existing theories:

Qualitative researchers build toward theory from observations and intuitive

understandings gained in the field.
(Merriam 1998, p.7)

This involves a prominent component of individual practice stemming from my experience of

interface design for composition, meaning that much of this research can be characterised as
practice-led research within the creative arts (see Nelson 2013). A mixed methodological
model for conducting and analysing this practice and its outcomes is taken from Brown’s
(2007a) software design as educational research model (SoDaR) supported by grounded
theory (see chapter 4). Furthermore, observations are derived from the application of this
practice, and its outcomes, in different case studies. These studies are often sequential
sessions returning to the same school, but where the emerging outcomes of one session,
analysed through a grounded theory-based approach, inform the structure and content of the
next. In this respect, the approach has links to the method of action research in education,
where issues are addressed dynamically within the learning environment (Mills 2000). To
some extent, this situates the findings within specific case studies, and means that the process
must be applied to different contexts in order to verify broader hypotheses. Bassey tells us
that these ‘fuzzy’ generalisations are of empirical value, as they can provide insights into the
effectiveness of educational practices within specific settings, as well as the understanding
that comes from applying these practices across a broader range of settings (Bassey 2001).

9
1.1.6 Ethics

Ethical approval for this research project was obtained from the College Ethics Panel at the
University of Salford during the first year of study. This application detailed the methodology
for developing resources and using them to conduct research as a school visitor. At this time,
I also obtained new documentation from the Disclosure and Barring Service via the School of
Health and Social Care, certifying me to carry out research with children of less than 18 years
of age.
Schools were invited to take part in the study by email, and an electronic copy of a
leaflet detailing the aims of the project at a key stage 1 reading age was attached. When a
school agreed to take part and arranged a date, forms were sent to the school again giving this
information and requesting consent from the parents or guardians of the pupils to allow them
to take part in the study. These were given to the pupils by the teacher, who then collected
them prior to the date of the study. In early studies, surveys were used to gather information
on the musical ability of the pupils taking part. No other personal information was collected,
and the names of the pupils and their schools are anonymised in this thesis and the other
published works emerging from this research.

1.2 About the Researcher

1.2.1 Qualifications and Experience

Merriam (1998, p.205) tells us that one of the ways in which a qualitative researcher can
achieve validation is by stating their theoretical orientation and biases at the outset. To this
end, I feel that it would be helpful to give some insight into my ideas and experience prior to
commencing this research.
I have held some form of educational role for the entirety of my adult life. Several of
my immediate family are, or were, teachers in comprehensive schools similar to those in
which this research has been conducted, so the discussion of educational practice has been an
ever-present influencing factor. Naturally, I have always pursued occupations related in some
way to the educational environment, from work experience to later employment and study.
While attending university I worked as a peripatetic guitar and keyboard tutor. Upon
graduating, I was employed by an educational supply agency and worked in a variety of roles
for several years, including teaching assistant, cover supervisor, and individual support for

10
pupils with special educational needs. I also held a long-term position as learning support and
intervention for the English department of Prestwich Arts College during this time, which
involved the assessment of individual learning needs and tailoring of approaches to meet
these requirements. After much experience and development in such roles, I eventually
completed a Postgraduate Certificate in Education through Edge Hill University alongside a
Schools Direct training programme with the Kingsbridge Educational Improvement
Partnership. During this period, I taught key stage 3 and 4 music at Hawkley Hall High
School in Wigan, with an additional placement at Manchester Communications Academy. I
now teach on the BA Music Production course at the British and Irish Modern Music Institute
and the MA Music course at the University of Salford, specialising in interactive and
emergent technologies.
Music and interactive technology have both been major components of my academic
and professional development. I hold a BA (Hons) in Popular Music and Recording and an
MA in Music Composition. Becoming interested in programming languages around the time
of my MA, I focused on the design and development of interactive music systems as modes
of composition. As previously mentioned, I have written and delivered modules relating to
this area of study for BA Music programmes. My main area of interest and expertise concerns
visual programming languages such as Pure Data and Max/MSP/Jitter, though I also have
experience working with Java, C, Visual Basic and Python. I occasionally made use of these
skills when teaching secondary school music, as a way of incorporating interactive games and
activities into my lessons. In some respects, this influenced my research proposal for this
project.

1.2.2 Theoretical Orientations

As an educational practitioner, I naturally hold certain preconceptions about how learning

takes place, rooted in experience, knowledge and instinct. In line with the ideas of
constructivism, I feel that people learn best when given opportunities for active experience
and reflection, and that, wherever possible, the learner should have opportunities to direct this
process for themselves. These ideas are not revolutionary, having been at the heart of
educational reform for the better part of a century. However, it is also my experience that
they have emerged in different areas of education, manifesting successfully in some ways but
not in others. One might expect that a creative subject like music would be one of the

11
prominent areas for constructivist teaching practices, though research tells us that this is not
the case (see 2.2). In my opinion, there are certain barriers imposed by resources and musical
forms commonly used in music curriculums, which emphasise knowledge and skills that
stand in the way of active experimentation. The learning of notation, for example, is
undoubtedly an important musical skill, but becomes a barrier if we take it as a prerequisite
of active music-making.
The above point raises another question of theoretical orientation. As a classically-
trained musician, I have an understanding and experience of traditional musical forms, such
as standard notation and orchestral instrumentation. However, as someone who has been
actively involved in many varied musical pursuits, I do not believe that being a musician
begins and ends with this kind of knowledge. The experiences which have most significantly
defined my musicality have all been active and social in nature. My route to these
experiences involved traditional musical training, but for others who shared these
experiences, this was not the case. Therefore, I tend toward a broader interpretation of what
kinds of activity can be considered music-making, and what constitutes a musician. Again,
this will inevitably have some influence on the kind of issues I will perceive as problems, and
the kind of solutions I will perceive as appropriate.
I also hold certain ethical assumptions surrounding educational research, that the
validity of educational research is shown when it is conducted in an educational setting and
learning takes place. In other words, educational research is valid when it assimilates into the
everyday learning process of a child at school, and results in some form of added value. This
is a responsibility not only to the relevance of the research, but also to the wellbeing of the
child and the practices of the school. Therefore, as educational researchers, we do not have
the freedom to try out a multitude of possibilities expecting different results, especially where
the educational benefit of such possibilities is questionable. The ethical considerations of
ensuring clear and accurate results through varied testing is inevitably compromised by the
more pressing ethical considerations of ensuring that each child has access to quality learning
experiences. This is supported in the ethical guidelines published by the British Educational
Research Association (see BERA 2014).

12
2. Research Context 1: Constructivism & Creativity

Figure 2.1

As an interdisciplinary project, the associated research context covers a range of

subjects, spanning musical creativity, educational theory and digital technology. Therefore,
this part of the narrative is divided up into two chapters; chapter 2 covers broader
educational and creative considerations underpinning this research to define questions, while
the chapter 3 focuses on a body of existing work, theory and practice necessary to develop
practical and methodological considerations.
This chapter will establish the research context for this project. I will provide an
overview of the constructivist paradigm, tracing the emergence and continual relevance of
the associated theories of learning. This discussion will focus on interactive digital
technology and music education, particularly what constructivist perspectives tell us about
musical creativity in the classroom, and how this may be supported by digital technology,
with reference to key examples from the field. Finally, I will summarise this context as a set
of initial research questions.

13
2.1 What is Constructivism?

2.1.1 Overview

The concept of constructivism, emerging from among the competing paradigms of education
over the last century, has come to signify a variety of evolving perspectives on the nature of
learning and instruction as an active, student-led, dynamic process, and continues to drive
educational reform, as well as informing new strategies for teaching, resource-design and
research. Duffy and Cunningham (1996, p.2) summarise these perspectives as ‘the general
view that (1) learning is an active process of constructing rather than acquiring knowledge,
and (2) instruction is a process of supporting that construction rather than communicating
knowledge.’ If we conceive of knowledge as communicated artefact, something which can be
transmitted from a teacher or a book to the recipient learner, like data from one computer disk
to another, then learning is the acquisition of this information; the teacher possesses this
knowledge and the student does not, so they must follow the instruction of the teacher in
order to acquire it, typically involving the recitation of facts or copying of processes,
otherwise known as rote-learning. Constructivist educators attempt to accommodate or
facilitate the means by which learners build this knowledge for themselves. Despite varying
perspectives across the constructivist paradigm, the general consensus within education is
that teachers should strive for understanding, and that this is something which is achieved
only by the learner, within a context and environment that enables them to do so.

2.1.2 Emergence

Constructivism does not claim to have made earth-shattering inventions in the area of
education, it merely claims to provide a solid conceptual basis for some of the things
that, until now, inspired teachers had to do without theoretical foundation.
(Von Glaserfeld 1995, p.15)

Constructivism is not, in itself, an approach, methodology or pedagogy. It is an ontology

which recognises the experiential dynamic of learning and understanding, which may be used
to direct or understand specific approaches. For example, the seminal pedagogical
methodology of Herbart (Ufer 1894, pp.81-7) might be described as ‘proto-constructivist’,

14
building upon the prior experience or understanding of the learner and abstracting from
concrete experiences to form new meaningful and functional processes3. The teacher, from
this perspective, is a skilled practitioner of education, not just the possessor of knowledge but
the facilitator of understanding. With the emerging acceptance of psychology as a science in
the late 19th century came the recognition of opportunities for a sound theoretical basis for
education (Thorndike 2013). However, such was the weight of behaviourist ideas in
psychology for the first half of the 20th century, the primary manifestation of influence in the
classroom during this time concerned discipline, or ‘learning as response acquisition’ (Mayer
1992, p.406-7).
With the latter half of the 20th century, the paradigm of educational psychology
shifted away from behaviourism and toward cognitivism, leading to a focus on ‘learning as
knowledge acquisition’ and ultimately ‘knowledge construction’ (Mayer 1992, p.407-8). The
continual influence of Dewey emphasised a view of education as an evolving system, where
progress is sustained through preoccupation with the experience of the learner, and therefore
reliant upon challenge and change:

Education is a social process; education is growth; education is not preparation for life
but is life itself.
(Dewey 1916, p.239)

Such ideas ultimately drew the trajectory of educational philosophy toward a notion of a
system which must grow and develop like the learner, and must therefore be derived from the
experience of the learner, as opposed to the maintenance of externally imposed stasis implied
by the behaviourist model. Dewey is critical of such prescriptive curriculums of learning
where they do not reflect the life-experiences of students:

A single course of studies for all progressive schools is out of the question; it would
mean abandoning the fundamental principle of connection with life-experiences
(1938, p.78)

This questions the coherence of any paradigm which seeks to impose knowledge artefacts as
having universal relevance. The content of the progressive curriculum, then, would be open

3
This taxonomy is comparable to Dewey’s ideas on formative life-experiences (English 2013, pp.105-13) and
Kolb’s experiential learning cycle (1984).

15
to challenge and change, and delivered with a context of authenticity to the life experiences
of learners. Within this emerging paradigm, the focus of attention is not the content itself, but
the perspective of the learner.

2.1.3 Cognitive and Social Constructivism

Widely acknowledged as the most influential figure within the constructivist paradigm is
Piaget, whose theory of cognitive development proposes that children intuitively employ a
different mode of learning from those consciously employed by adults, in that they construct
their understanding through a process of experience, assimilation and accommodation, and
develop structures (or schemata) through their play and explorations (Piaget & Cook 1952).
This highlights the importance of the learner perspective, the need to bridge the cognitive gap
between the adult teacher and the child who have fundamentally different thought processes,
and the recognition of play as an exploratory and developmental component of learning:

The prevalence of play among children is therefore to be explained not by specific

causes peculiar to the realm of play, but by the fact that the characteristics of all
behaviours and all thought are less in equilibrium in the early stages of mental
development than in the adult stage.
(Piaget 1962, p.147)

It is through this dynamic process that the individual maintains a state of cognitive
equilibrium between new experiences and environments and those understandings previously
acquired:

Equilibration is not a sequential process of assimilation, then conflict, then

accommodation; it is not linear. Nor is assimilation a process of ‘taking in
information’ as it has sometimes been described. Equilibration is instead, a non-linear,
dynamic ‘dance’ of progressive equilibria, adaptation and organization, growth and
change. It results from ‘coupling’ with our surround.
(Fosnot and Perry 1996, p.18)

16
This contradicts the traditional curriculum model of information as sequentially transmitted
artefact, and of the learner as passive ‘blank slate’. The nature of how the learner receives
information is influenced by experience, both prior and during the immediate educational
process, a factor which shapes the information as well as the learner. Within constructivist
theory, knowledge is generally defined as a cognitive and social process yielding an
idiosyncratic representation of a new experience or environment; it is the active process of
gaining information rather than the abstract end-product (Cunningham & Duffy 1996, p.2;
Perkins 1999, pp.7-8; Rieber 1992, p.94; Scott 2006, p.17; Webster 2011, p.6). In educational
practice, this has led to a revision of the student-teacher classroom dynamic; the learner must
be both active and reflective, and it is the role of the educator to mediate this process through
maintenance of the environment and pertinent inquiry (Von Glaserfield 1995; 1996, p.7;
Holt-Reynolds 2000; Honebein 1996; Richardson 1997). These ideas are further substantiated
by Bruner’s theory of discovery learning (1961), which states that effective learning takes
place when knowledge is arrived at independently, leading to a realisation and individual
cognitive interpretation on the part of the learner.
While the earliest constructivist theories focused exclusively on the cognitive
processes of the learner, a new perspective emerged with theorists such as Vygotsky, who
believed that it is through interaction with society and culture that an individual forms a
context of understanding for their experiences, thus emphasising the importance of social
structures in learning. Vygotsky’s ‘zone of proximal development’ (1978, p.86) distinguishes
between what the learner can and cannot do unaided by assistive structures, or ‘scaffolding’
(Wood et al., 1976), making completion of a task significantly more accessible. The
distinction is often made between the camps of cognitive and social constructivism within
academic discussion (Cobb 1996; Palincsar 1998; Powell & Kalina 2009) including criticism
of the apparent mutual conflict of these ideas (Phillips 1995). It has been suggested that both
interpretations constitute manifestations of the same didactic process, whereby the individual
interacts with a changing environment (Scott 2011, p.192) and that the separation of these
ideas is pedantic and dualist (Liu & Matthews 2005, p.389). For the purposes of this
discussion, it will be useful to establish common ground between these two camps.
In both cognitive and social constructivism, the learning environment is defined as the
situation or setting in which the learner constructs knowledge, through interaction with peers
or some other external stimuli. To further consolidate these perspectives, Jonassen (1994,
p.35) suggests that there are characteristics shared by both cognitive and social constructivist
learning environments:

17
 1. Providing multiple representations of reality, reflecting the complexity of real life-
experiences.
2. Emphasising the construction over the reproduction of knowledge.
3. Emphasising meaningful and authentic tasks over abstract knowledge.
4. Emphasising open-ended real-world environments over predetermined sequences of
instruction.
5. Encouraging reflection on concrete experience.
6. ‘[Enabling] context- and content- dependent knowledge construction.’
7. Supporting ‘collaborative construction of knowledge through social negotiation, not
competition among learners for recognition.’
This example of a constructivist
framework situated in practice
reflects Kolb’s theory of experiential
learning (1984), conceived as a four-
stage cycle of active experience and
reflection (Figure 2.2). Concrete
experiences with real-world contexts
(in this case, multiple realisations of
Figure 2.2 – Kolb’s experiential learning cycle such) and a process of meaningful
reflection are the means by which the
learner synthesises new knowledge from environment. The final point is of key importance,
highlighting the mechanism by which the learner also influences their environment as well as
adapting to it. It is for this reason that collaboration rather than competition is widely viewed
as preferable within constructivist models of learning environment; by influencing peers
rather than acting against them, the learner has the capacity to challenge and change their
zone of proximal development, so equilibrium is maintained not just by the learner but by the
environment itself. In modern perspectives on practice, the role of the teacher is to negotiate
and manage this environment to ensure that all students access opportunities for experience,
reflection and discovery (see Holt-Reynolds 2000; Powell & Kalina 2009, pp.247-8).

18
2.1.4 Constructivism and Digital Technology

A seminal convergence of constructivist learning

approaches and digital technology can be found in the
work of Papert, exploring the potential for meaningful
concrete experience adapted from abstract learning
materials (see 3.1). This involves virtual immersive
playground environments which the user navigates
through a process of experimentation, discovery and
assimilation (1980, pp.120-136; 1987). Papert’s
research incorporated the LOGO programming
language (Figure 2.3), the first programming
environment to follow a pedagogical design, with the
graphical avatar of a user-controlled turtle aimed at
Figure 2.3 – Describing a square in situating the learner in the virtual environment, and
LOGO
facilitating exploratory emergent behaviours:

My interest is in the process of invention of ‘objects-to-think-with’, objects in which

there is an intersection of cultural presence, embedded knowledge, and the possibility
for personal identification.
(1980, p.11)

Papert’s theory of learning, constructionism, emphasises the need for meaningful creative and
social learning outcomes:

Constructionism… shares constructivism's connotation of learning as "building

knowledge structures" irrespective of the circumstances of the learning. It then adds
the idea that this happens especially felicitously in a context where the learner is
consciously engaged in constructing a public entity, whether it's a sand castle on the
beach or a theory of the universe
(Papert & Harel 1991, p.1)

These opportunities arise from micro worlds (Papert 1980), software environments which
make objects-to-think-with – building blocks – out of challenging or abstract ideas.

19
Constructionism is arrived at by reconciling constructivist theory with emerging
technological possibilities, a consistent theme in 21st century interpretations of learning
theories, as demonstrated by the ‘digitisation’ of Bloom’s Taxonomy (Churches 2008; 2009)
and strategies for web-based learning (Churchill 2009; Duffy 2008; Greenhow et al. 2009).
Siemens (2006; 2014) and Downes (2008; 2012) present an interpretation of constructivism
‘for a digital age’ in the theory and principles of connectivism:

 Learning and knowledge rests in diversity of opinions.

 Learning is a process of connecting specialized nodes or information sources.
 Learning may reside in non-human appliances.
 Capacity to know more is more critical than what is currently known.
 Nurturing and maintaining connections is needed to facilitate continual learning.
 Ability to see connections between fields, ideas, and concepts is a core skill.
 Currency (accurate, up-to-date knowledge) is the intent of all connectivist learning
activities.
 Decision-making is itself a learning process. Choosing what to learn and the meaning
of incoming information is seen through the lens of a shifting reality.
 While there is a right answer now, it may be wrong tomorrow due to alterations in the
information climate affecting the decision.
(Siemens 2014, p.5)

This theory recognises the learning environment as being in constant flux due to the changing
nature of culture ideas and information exchanges in a digital world. The ability to see
connections and make decisions is defined as learning, because the knowledge itself is
subject to change. Siemens (2014, p.3) argues that this pursuit of ‘fuzzy’ learning situates
constructivist principles in the modern world:

Constructivism assumes that learners are not empty vessels to be filled with
knowledge. Instead, learners are actively attempting to create meaning. Learners often
select and pursue their own learning. Constructivist principles acknowledge that real-
life learning is messy and complex. Classrooms which emulate the “fuzziness” of this
learning will be more effective in preparing learners for life-long learning.

20
However, this interpretation is less likely to resonate with the critics who attribute evidence
for constructivist approaches as factual relativism (see Boghossian 2007). Phillips (1995, p.5)
argues that constructivism has ‘become something akin to a secular religion’ among its
followers. Some educationalists have expressed the view that constructivist researchers can
be so blinkered by an ontological stance as to distort empirical evidence (Phillips 1995; Fox
2001). Arguing from a constructivist position, Jonassen (2006) warns against a dichotomy of
competing paradigms, where constructivists seek to replace perceived behaviourist methods
on principle alone.
In taking on the concerns of critics, it is important that constructivist ideas are situated
not just in emerging technological possibilities, but also in measurable outcomes of
classroom practice. Drawing from an extensive meta-analysis of classroom practice, Hattie
(2009; 2012) emphasises the importance of making learning visible so as to demonstrate clear
impact. This relates not just to the measurable visibility of what has been learned, but in the
capacity of educators to see this from the perspective of their students and thus help them to
become motivated evaluators of their own learning. Hattie defines learning as ‘the process of
developing sufficient surface knowledge to then move to deep or conceptual understanding’
(Hattie & Yates 2013b, p.26), where students become self-directed and self-aware knowledge
constructors. In measuring impact, then, we must question the extent to which students can
demonstrably appropriate new knowledge to different contexts and outcomes. This is often
referred to as deep learning, defined by Fullan and Langworthy as instances of learning
which ‘[develop] the learning, creating and ‘doing’ dispositions that young people need to
thrive now and in their futures’ (2014, p.i). Constructivist ideas may help us to decide how
digital technology might be used to support specific learning outcomes, but it is then vital that
we provide sufficient supporting evidence of the impact of this approach (for example, the
ability of learners to apply their knowledge to new areas). For educational theorists
principally concerned with best practice, the researcher who attributes their findings as
evidence for their wider ontological stance may be guilty of factual relativism in the pursuit
of a theoretical agenda; certainly, this takes the discussion outside of the classroom, and
renders it unlikely to contribute to the forum between teaching practice and educational
research (see Welch et al. 2004, p.270).

21
2.2 Constructivism and Music Education

2.2.1 Curriculum and Instruction

Constructivism has had a significant impact on music education. In the United Kingdom, the
traditions of the conservatoire curriculum (instrumental training, music theory) were the
entirety of formal music education prior to the establishment of the national curriculum for
the study of music in 1992. This was heavily influenced by Paynter’s (1982) arguments for
composition as a mode of creative and personal expression and development. Paynter
advocates a distinctly constructivist pedagogy, where ‘making music is more important than
musical information’ (p.xiiii) and learners can ‘use the skills they have acquired as they
acquire them’ (p.123). This led to the establishment of a constructivist music curriculum (see
Garnett 2013) predicated on fostering creative decision-making and provoking emotive
discussion around musical ideas. Paynter’s constructivist approach is echoed in more recent
ideas with respect to instructional design in music (see, for example, Bower 2008; Brown
2012; Scott 2006; Rinaldo & Denig 2009; Webster 2011). It should come as no surprise that a
subject so heavily concerned with both individual creative expression (composition) and
meaningful social experience (‘jamming’) should inspire an affinity with constructivist
ideals. Indeed, both cognitive and social constructivism seem highly relevant to musical
practices, as separate ideologies and as a convergence, as evidenced in group composition,
for example (see Barrett 2006). In an educational paradigm where activity is the precursor to
understanding, we should expect to see an inherent experiential feedback cycle of music-
making and reflection. However, the 2012 nation-wide report on music education by
OFSTED contradicts this:

Too much use was made of verbal communication and non-musical activities. Put
simply, in too many cases there was not enough music in music lessons.
(p.4)

This issue prompts a possible research question: What does constructivism tell us about the
nature of learning that we can use to support specific musical learning?
Bloom’s Taxonomy tells educators to start with basic units of information, build an
understanding, and ultimately apply in creative contexts (Krathwohl 2002). If we are working
towards a creative learning goal with fixed logical steps, building a website for example, this

22
is an undeniably useful framework, but if we are teaching an inherently creative subject, we
most likely want to pursue a range of formative creative behaviours. As such, the delivery of
prescriptive and sequential musical curriculums can be something of a stumbling block to
meaningful musical learning. Webster (2002, p.4) highlights the importance of enabling
students to become self-directing in their music education, and to be able to make creative
decisions:

One obvious gage of how successful we are as teachers is the extent to which our
students can make aesthetic decisions about music as listeners, composers, and
performer/improvisers and to develop a sense of musical independence. Such
independent thinking does not happen if each decision is dictated. Teachers must
teach for independent thought.4

He proposes that the goal of music education is to teach students to ‘think in sound’, and that
‘this is possible only if students are encouraged to ‘create’ music through all the available
behaviours’ (pp.4-5). Studies into effective classroom composing show that educators
manage a creative environment by facilitating freedom of choice and making a range of
options as accessible as possible, but crucially directing students to consider the reasons for
their compositional choices (Berkley 2004; Bolden 2009; Burnard & Younker 2002;
DeLorenzo 1989, p.197; Reese 2003).
Improvisation with new ideas plays an important role in facilitating creativity.
Burnard (2000) demonstrates how children quickly ascribe meaning to improvised choices
and forge further creative decision-making. Borgo (2007) also advocates improvisation in
early-years music education, and is critical of the assumption that students must first master
lower-level features of music (chords, intervals and durations – the lower-order thinking
skills of Bloom’s Taxonomy) before attempting higher-level music gestures (melodic,
harmonic and rhythmic expression – creative higher-order thinking skills):

According to this logic, only after the pyramid of processing has been completed,
should students expect to achieve the less talked about (and rather more nebulous)
goal of improvising, to have conscious or unconscious experiences of a truly musical
kind.

4
For comparable views on independent (or student-directed) learning, see Barrett 2006; Bolden 2009; Brown, J.
2008.

23
 (p.76)

The implication here is that, if a student only finds the determination, inspiration and
confidence that are afforded by meaningful music-making experiences after learning the
curriculum content, then a passive, abstract and potentially discouraging experience can be
expected until the student is has the opportunity to apply this knowledge. Active
experimentation with new ideas and approaches may be the most effective precursor to
further creativity and understanding:

Pupils achieved the highest standards when they were given musical and
technological information as they needed it.
(Pitts & Kwami 2002, p.69)

This resonates with the findings of OFSTED (2012, p.46):

Survey evidence showed, very clearly, that pupils made the most musical progress
when they were taught in music, rather than about music.

If learners construct knowledge for themselves, as

constructivist theory claims, then musical
information would arise from reflections on active
music-making experiences. This is succinctly
illustrated by turning Bloom’s Taxonomy upside-
down (‘Bloom’s 21’, Figure 2.4) and starting with
the goal activity. Wright (2012) suggests that
learning activities should start with a creative task
and ultimately synthesise new knowledge from
these experiences, which can then be re-applied,
very much reflecting Kolb’s experiential learning
Figure 2.4 – Bloom's 21
cycle (1984).
While creative and practical activities can be more easily directed by teachers with
specialist musical training, this does not, by itself, lead to high-quality learning. OFSTED
found that the best music lessons were conducted by trained full-time teachers with a musical
specialism (2012, p.20), but that specialist music teachers also accounted for a large

24
proportion of inadequate lessons where a lack of classroom management or teaching skills
were observed (p. 19). Generalist teachers, accounting for two-thirds of music lessons,
demonstrated questioning and management skills but made less use of musical activities, and
therefore showed a greater degree of consistency in terms of quality of teaching (pp.18–9).
We can conclude from these findings that practical music activities only lead to outstanding
lessons when coupled with the teaching skills necessary to guide students beyond the ‘create’
stage of the ‘Bloom’s 21’ model. OFSTED make this a central argument of their follow-up to
the 2012 report, declaring that ‘performance and enjoyment are not enough’ if music is to be
upheld as a ‘rigorous, academic subject for all’ (2013, p.9). It is therefore vital to have a
qualified teacher overseeing music lessons.
Generalist teachers, or teachers with limited musical experience, have shown a greater
capability to deliver a quality music education than, say, music specialists with limited
teaching experience (2012, pp. 18–9), although studies continually address issues about
confidence in teaching the subject (Biasutti 2010; Biasutti et al. 2015; Garvis 2013; Hallam et
al. 2009; Hennessy 2000; Mills 1989; Russell-Bowie 2009; Seddon & Biasutti 2008; de Vries
2013). In recent years, government reforms have attempted to address this ‘musical skills
gap’ with initiatives such as the First Access scheme, in which schools employ visiting music
specialists from local music hubs to deliver music lessons5. A lack of pedagogical training
and knowledge of curriculum aims appears to have posed barriers to such lessons (OFSTED
2013, pp.16-8). Modern teacher training is intensively directed at helping students to
construct knowledge for themselves, so teachers possess the necessary pedagogical skills, but
do not necessarily have the resources at hand to deliver abstract musical ideas as active and
experiential learning environments. This would account for the overreliance on explanatory
classroom activity described by OFSTED (2012, p.4).
The introduction to the 2013 national curriculum for the study of music in the United
Kingdom cites the vital formative influence of creative musical opportunities, describing
music as ‘a universal language that embodies one of the highest forms of creativity’ with the
potential to impart ‘self-confidence, creativity and sense of achievement’ (Department for
Education 2013, p.1). The 2011 National Plan for Music Education follows similar rhetoric
with a quotation attributed to Aristotle: ‘Music has a power of forming the character and
should therefore be introduced into the education of the young’ (DfE and DCMS 2011, p.2).
5
OFSTED (2013, p.17) report that, among the schools who were aware of this policy, staff were often relieved
to defer wholesale the provision of what is arguably the most cumbersome subject to external specialists. It is
another indication of uncertainty around musical teaching that it was so commonly assumed that this would lead
to a higher standard of musical learning.

25
In the original 1992 curriculum, music is presented as a practical subject with composition as
a central activity (Swanwick 1992, pp.162–3). At 77 pages, this is a detailed document,
providing creative frameworks, sample schemes of work and lesson structures, and specific
guidance on SEN and ICT, among other important considerations. The 2013 version consists
of 4 pages, two of which are identical, to cover the aims, attainment targets and breadth of
study for key stages 1 to 4.
The substantial reduction of the curriculum comes at a cost to the specific guidance on
creative pedagogy present in previous versions. We are told that students should ‘improvise
and compose music for a range of purposes using the inter-related dimensions of music’ (DfE
2013, p.2). This refers to what were described in previous versions as the combined elements
of music: pitch, dynamics, duration, timbre, texture, structure and tempo. The 1999
curriculum encouraged reflection on how these elements are combined to produce musical
decisions or impressions, placing emphasis on subjective language of ‘mood’, ‘feelings’ and
‘intention’ (DfEE 1999, pp.16–9). It is clear that, in this curriculum, the recognised value of
music education is in its capacity to promote forms of interpersonal and intrapersonal
development (p.8). These values suggest a drive to encourage musical reflection as a means
of promoting social issues; inclusion, independent thinking, and readiness to challenge
immoral or irresponsible ideas. It is perhaps a reaction to OFTSED criticism of excessive
verbal communication (see 2.2.1), augmented by the reflexive ‘political football match’ as
responsibility for education passes from one party to another, that the 2013 curriculum
dispenses with this line of thinking altogether in favour of an approach that might be
described as ‘keeping it to-the-point’. The concise targets of the 2013 curriculum suggest a
narrowing of focus to the more traditionally academic aspects of music education: developing
an understanding of the history of music and great composers through listening and
appraising, and being able to ‘use and understand staff and other musical notations’ (DfE
2013, p.2). Staff notation, in particular, has hitherto not been seen earlier than key stage 3.
Previous curriculums have always placed greater emphasis on alternative notations, while the
staff system has in the past proved too daunting a prospect for many teachers (see Mills 1994,
p.194), and is arguably not an aim to introduce without substantial reasoning and support.
Among the greatest reductions, especially when compared with the 1992 curriculum,
is in specific guidance on the use of technology. Though OFSTED reported ‘insufficient
improvements in the quality of learning through the use of technology’ (2012, p.54), the only
technological guidance in the 2013 curriculum is that it should be used ‘appropriately’ (DfE
2013, p.1). The 1992 curriculum, by contrast, contains a section on ‘using IT in music and

26
developing IT capability’ (DES 1992, p.67), referencing what were (at the time) emerging
technological practices, and also containing various detailed references to the use of assistive
technology in the music education of pupils with special educational needs (see, for example,
pp.1, 57, 70). These points were arguably more pertinent in 2013, as educators pointed out
the ‘digital nativism’ (Prensky 2009) and ‘new pedagogies’ (Fullan & Langworthy 2014)
presented by digital technology in the classroom, and with research suggesting a prevalent
culture of digital music activity thriving outside of school life yet largely unexplored in music
classrooms (Daubney & Mackrill 2012).
Consider the following example of a music lesson activity, taken from the 1992
curriculum, which follows a framework of guidelines for improvisation, composition and
arrangement (p.47):

Year 9 pupils are set the task of creating a piece called Hiroshima in which the
musical depiction of a nuclear explosion is set between a threatening prelude and a
reflective epilogue: a given structure. Pupils worked in groups for half an hour and
then reassembled to perform their compositions, which were video-recorded. In one
group, seven pupils play timpanum, side-drum, tom-tom, cymbal, wind chimes,
trombone and electronic keyboard. The piece opened with very quiet chord clusters
on the keyboard alongside barely audible wind chimes. Percussion instruments
gradually imposed an ominously repetitive rhythm. The trombone added insistent
long notes on a single pitch. A fierce climax was reached. There was a long silence.
The epilogue echoed the prelude with the rhythm fading away into nothing. The class
analysed the outcome with the teacher prior to listening to Penderecki's Threnody.

The central components of musical activity, composing, performing, listening and appraising
(Swanwick 1979) are all present, though crucially, the example begins with the practical
exercises and concludes by inviting reflection. As such, it follows Kolb’s (1984) experiential
learning cycle. Notice that the activities are a prelude to listening to Penderecki’s Threnody –
a more traditional lesson structure might have the pupils listen to the piece and then create
something similar, resulting in a preoccupation with the lower order recall and apply skills of
Bloom’s taxonomy. Instead, we have an example of flipped Bloom’s taxonomy (Wright
2012), starting with a creative task and relating this to the source content at the end. This is
much more likely to result in original, idiosyncratic and unexpected outcomes, and therefore
a more varied and meaningful discussion – it is a lesson structure akin to the software

27
development model of bottom-up design and complex systems (see 4.1.1), demonstrating the
interconnected ideas and suggestions present in the 1992 curriculum. Also, the inter-related
dimensions of music are employed in a truly inter-related and expressive manner in this
example, and it notably demands very little in the way of formal knowledge regarding
harmony, key signature, staff notation, music history, and other more traditionally-academic
aspects of musical education while presenting, in the conclusion, opportunities to engage with
these aims. The 2013 curriculum is, by contrast, akin to the top-down design model (see
4.1.1). It presents its outcomes in the abstract, focusing on what students should learn but
neglecting the question how they could learn it. Garnett (2013, p.169) argues that this
constitutes a return to learning as response acquisition:

[Year 9 students] were learning how to perform a task correctly – essentially a piece
of behavioural learning – rather than experiencing an encounter with music that
prompted them to engage and expand their own concepts of how musical material can
be organised. The constructivist curriculum was thus taught in a behaviourist manner
fuelled by conceptually driven learning objectives.

2.2.2 Musical Creativity in the Classroom

Jonassen (2006, p.44) is critical of educationalists who ‘claim that constructivist solutions
should replace behaviourist methods’, though he also argues that effective solutions may be
founded on constructivist principles (p.43). Constructivism is not a way of teaching, but a
view of learning, and therefore offers no solutions by itself, but may bring to light instances
of learning which may benefit from new approaches. We may apply this lens to musical
creativity, and see an aspect of musical learning that has hitherto proved troublesome
(Shively 2015, p.129). Specialist, as well as generalist, teachers of music report uncertainty
on how to teach composition (Winters 2012). The curriculum gives us a framework of
musical devices, and we may direct students to apply these devices, but how do we argue that
they have done so creatively? Set against this question, in successive curriculums, is the
benchmark of the ‘great [Western] composers’, and the principle of developing a discerning
appreciation for their works, ‘to listen with discrimination to the best in the musical canon’
(DfE 2013, p.1). Burnard (2012) refers to ‘myths of creativity’ which disseminate an
unrealistic notion of composition:

28
 [Musical] creativity is rarely, if ever, a matter of the lone composer, or self,
composing in isolation. Rather, it is a matter of composers working and playing with
and with respect to others, intentionally or unintentionally... Creativity is not the
product of single individuals, but of social systems making judgements about
individuals’ products.
(p.24)

Myths such as the one identified here contribute to an idea of creativity that is wholly
unquantifiable; we might call it inspiration. This is not traditionally viewed as something
which can be taught, but which must emerge from the gifted individual. However, the same
can be said of understanding within the constructivist paradigm (see the start of this chapter,
for example) and yet this is not seen as a scarcely-attainable abstract but rather the very
objective of every lesson. When we view musical creativity through the lens of
constructivism, we may see through the veil of myth and observe something which is very
much attainable; even, part of an experiential process.
Burnard’s explanation of compositional process here resembles a social equilibration;
collaboration not just in the direct sense, but also in terms of indirect influence – reaction,
response, cultural ideas – all of the ways in which our social environment is engaged in a
dynamic interplay with our creative efforts. Borgo (2007, p.71) argues that, if we are shaped
by our social environment and ‘perceived place’ within it, then ‘all musical encounters… are
inherently social, since other listeners are always either present or imagined.’ While we know
that musical creativity can be fostered by a vast range of personal, social and cultural
circumstances, research is needed to contribute toward a framework for understanding how
this occurs. Previous research provides us with some insight into this field.
A study by Gall and Breeze (2008, p.35) found that students with less musical training
were discouraged from contributing to the task when grouped with more experienced
musicians, as they felt ‘less ownership of the final product’ due to the more experienced
students ‘[taking] control of the compositional process.’ In such scenarios, the more
experienced students are seen to be skipping the ‘exploratory phase’ of the activity, to the
detriment of its educational value for their less musically-confident peers6. The research
suggests that bridging this gap between students who are inclined and disinclined to

6
Research by Odam (2000, p.118) also suggests that more musically-confident pupils tend to find shortcuts
through the compositional process, and should be challenged with alternative approaches.

29
experimentation is an important consideration in the design of any resource aimed at
supporting creative-decision making. Also, this may be of particular relevance to the
transition between primary and secondary education (key stage 2 to 3) where differences in
provision of key stage 2 music education lead to potential disparity in skills and
understanding at the start of key stage 3.
A study on the student interaction with compositional computer programmes by
Scripp, Meyaard and Richardson (1988) found that students with no formal music tuition
were more exploratory in their use of the programme than those who had received musical
training. Mellor (2008, p.453) suggests that ‘those participants with musical training
produced responses which ‘fitted’ their self-perception of the ‘musical’ expectations inherent
in the task’ thereby diminishing their need to experiment. These findings are further
substantiated in several other studies (Folkstad et al. 1998; Hewitt 2002; 2009; Seddon &
O’Neill 2001; 2003; 2006). A study by Hewitt (2009, pp.19-20) concludes that students with
musical training tend to spend less time in the ‘exploratory phase’ of a music-making
activity, and proposes further determining factors: The older children in the study were also
less prone to exploration and experimentation (the age range of the group was between 7 and
12 years), as were those who were more familiar with the activity. This suggests that the
factors which encourage exploratory musical learning should be further examined.
Researchers have observed a need for case studies which examine the individual
experiences of students engaged in the music-making process, to understand how intuition
and personal perspective affects creative decision-making (Hallam in Welch et al. 2004,
p.250; Higgins 1992, p.491; O’Neill in Welch et al. 2004, pp.257-8). Hewitt (2009, p.21)
recommends that there is a need for research which focuses on both the macro- and micro-
levels of analysis; that is, the music-making experience of the individual and the wider social
process. What is clear from this discussion is that the social environment, be it present or
virtual, should be a key consideration in any research into pedagogical creativity and music
technology. The value of what teachers and other educational practitioners can bring to this
discussion should not be underestimated; Welch (et al. 2004, p.270) identifies the need for a
‘symbiotic research environment’ formed by closer links between researchers and
practitioners and is critical of the ‘ongoing ‘conservative’ influence of music conservatoire
curricula… largely untouched by pedagogic innovation’ (p.269; see also Cain 2004, p.220;
Savage 2007, p.72).

30
2.2.3 Supporting Musical Creativity with Digital Technology

One of the functions of this chapter is to generate research questions. I have stated that this
project does not aim to supplant behaviourist methods with constructivist ones, but rather to
view musical learning from a constructivist perspective to evaluate where a new approach
may produce successes. Similarly, digital technology should not be assumed to be inherently
preferable, and it is not to be taken for granted that any aspect of learning should be digitised,
but many aspects of the curriculum have been supported by digital technology with empirical
success (in the work of Papert, for example). This prompts further questions:
 To what extent have interactive digital technologies hitherto succeeded in
encouraging musical behaviour in the classroom?
And if we are concerned, like Paynter, with supporting creativity in the first stages for formal
music education, we may ask:
 How might interactive digital technology be harnessed to facilitate creative
expression in the musically untrained?
Burnard (2012, pp.252-3) suggests that implementations of digital technology for musical
creativity has new implications for collaboration, assessment and the role of the teacher in
digitally-augmented creative environments. Digital technology, then, has implications for
music education beyond merely ‘modernising’ existing structures. The importance of a
curriculum inclusive of relevant technology and informed by empirical research is further
highlighted by Cain (2004, p.291):

Curriculum change is necessary if the world of the classroom is to keep pace with the
world outside. And it is also necessary to have a clearly defined theory which allows
teachers to commit themselves intellectually to the change.

Similarly, Savage (2007, p.75) warns of the ongoing need to implement relevant technologies
into effective teaching practices:

If educators fail to grasp this major cultural shift, music as a curriculum subject will
become increasingly alienated from young people’s lives and they will find their
music education elsewhere.

31
There is a clear need for further research examining student interaction with new
technologies, the possibilities of integrating social networking in the interest of meaningful
engagement and wider social participation, and of course, close dialogue with students and
teachers.
There have been many software solutions
aimed at assisting both specialist and generalist
music teachers in delivering a more creative and
experiential curriculum. The most prominent
example of this in the United Kingdom is
Charanga (see 3.2.2), which offers a suite of
sequential virtual lessons based on the music

Figure 2.5 – jam2jam

curriculum. Many promote collaborative learning
in digital environments. In an age where
technology and virtual networks are increasingly significant, there is a demand for music
education to extend beyond ‘the constraints of classroom and curriculum’, as ‘schools and
colleges cannot be musical islands’ (Swanwick in Welch et al. 2004, p. 242). The same
metaphor is used by Dillon (2009, p.297):

Musicological research tools provide a bridge between these islands [of virtual and
present music-making environments] and have the potential for directing the
development of educational software that enables learning within both contexts.

Much has been written on the educational benefits of

collaborative music-making with respect to digital
technology (Barrett 2006; Blaine & Fels 2003; Dillon 2003;
Hewitt 2008; Savage & Challis 2001; Sawyer 2008;
Wiggins 1994). Emerging technology provides
opportunities to extend the benefits of collaboration beyond
present and virtual ‘islands’ through networked jamming
and composing, as seen with jam2jam (Brown 2007a;
Brown 2007b; Dillon 2004; Dillon 2009; Figure 2.5) and Figure 2.6 –Drumsteps
Networked Drumsteps (James & Stanton 2011; McCarthy et
al. 2005; Figure 2.6). These systems provide accessibility to both music-making experiences,
through the undemanding human-computer interface, and to wider social experiences,

32
through virtual networks of interaction and collaboration. Brown (2007b, p.296) suggests that
such opportunities are key to the development of personal schemata:

The ways in which [computer music systems] are used for musical activities teaches
students about how musical activities are conducted, for example whether or not
music is a private or social activity, individual or collaborative, interpretive or
inventive, unimportant or important, and so on. The integration of computer music
systems into the music program is, as with all other activities and resources, not
simply about the provision of a tool or about the transference of ideas; it is about
situating the student in a musical culture.

The sharing of compositions, over social networks for example, also appears to have highly
positive pedagogical implications regarding confidence and creativity (Kardos 2012, p.150;
Maag 2006; Partti & Karlsen 2010, p.375; Salavuo 2008).
Educationalists have also written about the potential of digital technology to support
new modes of musical interaction in the classroom. Savage and Challis (2002, p.2) express
concerns over the limitations imposed by the omnipresent MIDI-keyboard workstation:

MIDI is a highly flexible and powerful system for controlling all aspects of electronic
sound. However, within schools MIDI seems to have become synonymous with
keyboards linked to sequencing software. Recent research by the Fischer Family Trust
suggests that for 75% of high school pupils this is their primary means of performing
and composing with ICT. In this example pupils are relying on a 400-year-old
interface (the keyboard) for accessing the sounds contained within a computer!

Their research project Dunwich Revisited (Savage & Challis 2001; 2002) afforded a range of
compositional approaches, including modern electroacoustic methods, as well as diverse
stylistic genres, as part of a concept-based classroom composing activity:

The ways of working that pupils adopted when composing with the various
technologies represented a significant change from what might be called ‘notation-
based’ compositional tasks. This encouraged a greater sense of freedom both in
composition and performance… Even those performers playing the more
conventional parts in the final performance ‘improvised’ within the structures they

33
 had decided to adopt… The uses of different technologies empowered the majority of
pupils. They provided tools to engage in this sense of play through speculating,
affirming, selecting, rejecting and evaluating musical ideas both individually and
corporately.
(2001, p.147)

The feedback suggests that the range of familiar and unfamiliar approaches prompted a
highly democratic and exploratory social learning process7, with even the more experienced
musical performers being compelled to experiment. The inclusion of sound design and
manipulation technologies are identified as highly influential in the development of a creative
vocabulary among the students (Savage & Challis 2001, pp.142-3; Savage 2005, pp.171-2)8.
As an educational project, Dunwich Revisited is an effective response to Qualifications and
Curriculum Authority (QCA) recommendations made around the same time:

Teachers need to choose a context of relevance to young people’s lives, select an

interesting challenge and ensure that pupils have the necessary artistic skills.
Providing choice, ensuring autonomy, encouraging teamwork, allowing
experimentation and encouraging perseverance are key components of fostering
creativity within the arts.
(QCA, 2000, p.9)9

However, OFSTED’s 2012 report on the quality of UK music education found ‘insufficient
improvements in the quality of musical learning through the use of technology’ since their
previous assessment (p.54), and found no evidence of exploring electroacoustic music-
making techniques in the 194 institutions inspected, ‘even at key stage 4 and A-level’
(p.55)10. The findings of OFSTED suggest that the implementation of the most recent
technologies, which may in themselves present opportunities for emerging musicalities
relevant to the lives of young people, have gone largely unexplored (see also Daubney &

7
For another study on collaborative music-making projects maximising student potential by allocating a variety
of roles, see Tobias 2012.
8
For further ideas on the importance of sound design within the evolution of musical language see Théberge
1997, p.188.
9
A project similarly involving electroacoustic methods, cultural relevance and collaboration, also with
promising implications regarding engagement and teamwork, though less pedagogically-focused, can be found
in Freeman et al. 2011.
10
It has been suggested that there is a general deficiency in compositional software incorporating
electroacoustic methods (Eaglestone et al. 2001).

34
Mackrill 2013). More detailed empirical evidence is required on how children of different
age, gender and experience respond to a range of music-making technologies (Webster 2012).
The availability of a range of music technologies not only encourages
experimentation and versatility, as suggested by the research of Savage and Challis (2001;
2002) but creates an environment of accessibility for students who may otherwise struggle to
develop a musical identity:

Technology provides pupils with access to a musical environment that is not

dependant on their own performance skills
(Hodges 1996, p.77)

Several studies have focused upon the effectiveness of the sequencing software eJay on
promoting musical creativity in schoolchildren (Dillon 2003; Gall & Breeze 2008; Mellor
2008; Figure 2.7). The software employs pre-programmed ‘tape loop’ structures, which the
user arranges, thus providing an accessible environment for exploring form, structure, style
and texture:

For some, especially those without musical keyboard skills or without experience of
playing a traditional musical instrument, the experience of using Dance eJay not only
opened up a possibly for seeing themselves as a musician with music technology as a
main instrument, it also opened up the possibility to see themselves as a musician.
(Mellor 2008, pp.468-9)

Furthermore, such alternative approaches

may offer certain students the possibility of
a uniquely personal creative identity,
through the application of methods which
traditional resources do not permit (Wise et
al. 2011, p.125). There is limited room for
such emergences if classroom
compositional approaches continue to be
generally prescribed:
Figure 2.7 – eJay

35
 …our results show that the ways in which music is created varies between
individuals, and between different kinds of music. An important implication of this is
that there is no such thing as 'right' or 'wrong' with the respect to method or strategy
by which music should be created, and consequently that school should not teach the
method of composition, but rather create a context in which the pupils can explore
their own ways into music composition.
(Folkestad et al. 1998, p.95)

In order to support creativity, educators must ensure that students have access to approaches
that they consider meaningful and authentic (Crow 2006). This may be something that
emerges in the classroom, or it may already exist outside of the context of the classroom in
the ever-growing proficiency of young people in their use of personal communications
technology. It seems a logical progression to assimilate this proficiency into learning
environments (Gouzouasis & Bakan 2011; Norris et al. 2011; Zhou et al. 2011). Prensky
(2001; 2009) asserts that the ‘digital native’ (or ‘digital wisdom’, to use the more age-
inclusive updated terminology) demands an educational environment which reflects the
context within which they have come to understand and interact with the wider world, and
argues that educators must acknowledge and embrace this need in order for students to see
their learning experience as authentic. Recognition of this ideal may be seen in the recent
emergence of ‘bring your own device’ policy in some schools, despite a tradition of
conservative resistance to schoolchildren using personal communications technology onsite,
for educational purposes or otherwise (Tierney 2012). In any case, if educators are to create
an environment in which creativity is truly supported, it will be necessary to be open to
approaches which may prove to hold a context of authenticity to young people (Custodero
2002, p.6).
In promoting a context of authenticity, and also ensuring that students have an
understanding of the musical concepts they are applying, successful digital environments may
involve some form of student-directed instruction (Brown 2008). A more open-ended format
may allow students to build their own framework of musical concepts, addressing the issue of
suppressed experimentation in more experienced musicians, and allowing creating an
environment for improvisation non-dependent upon the prior mastery of low-level musical
skills (Manzo 2010). Rudi (2007, p.140) is critical of instructivist interface design which
relies on sequential information processing, and advocates systems based upon a nonlinear
learning model. Savage (2005, p.175) expresses a similar view, exploring the potential of the

36
computer programme as a ‘meta-instrument’ which can be customised to meet individual
needs. This may also have promising implications concerning the role of interactive
technology in meeting specific barriers to learning (Anderson & Smith 1996; Collins 1992;
McKnight & Davies 2013, pp.24-44) and in providing opportunities for cross-curricular
learning (Lehrman & Ryan 2005).
The apparent suitability of customisable interfaces to classroom applications is
reflected in the anecdote of a teacher consulted in a study by Savage (2007, p.70) whose
students ‘found more interesting things to do within the technology’, subverting the
prescribed learning outcome (see also Mills & Murray 2000). These unexpected occurrences,
pivot points, may provide valuable opportunities for student-directed learning:

A constructivist classroom is one in which the teacher welcomes learners to enter with
the teacher into the learning process, as full participants in exploring our, their, and
others’ musical worlds.
(Shively 2015, p.133)

As students become increasingly accustomed to experimenting with digital technologies,

perhaps this should be reflected in the use of ICT in the classroom (Prensky 2001; 2009;
Beckstead 2001, pp.48-9). Teacher attitudes to these occurrences are examined by Savage
(2007, p.70):

A number (39 percent) spoke of their pupils knowing more than they did about a
particular piece of technology. This did not threaten the majority of the teachers (57
percent of this 39 percent), who saw it as a positive opportunity to encourage pupils to
move towards a greater degree of independence in their learning. They described this
shift from teacher-dependence to learner-independence as accompanying a shift in
their teaching role from instructor to facilitator.

The views expressed have significant implications for interface design, but are far from
unanimous. Clearly, this student-teacher dynamic and its relationship to emergent technology
demands further study.
The usefulness of visual representation in pedagogic music-making activities cannot
be underestimated (Gall & Breeze 2005, p.429). Savage and Challis (2002, p.1) distinguish
the intrinsic act of composition from its extrinsic aspects (score representation, notation etc.)

37
Within a constructivist framework, the extrinsic aspects of composition should not pose
barriers to the intrinsic act:

In my opinion, we only err when we place the cart of music theory before the horse of
musical experience…
(Borgo 2007, p.78)

Supporters of traditional notation have suggested ways of using active composition to instil
knowledge of its formation, using MIDI for example (see Brophy 1996). The dynamic
creation of score material from alternative and highly original methods of input has also been
explored (Farbood et al. 2004; 2007; Jennings 2005; Kelly 2011). The pedagogic value of
alternative notational methods has been examined in some detail (Jennings & Tangney 2001;
Jennings 2006; McCarthy et al. 2005) as well as the use of abstract graphics in supporting
engagement and focus (Blaine & Perkis 2000). From a constructivist perspective of the
dynamic nature of knowledge, it may make sense to make use of different modes of
representation for different modes of musical learning. Brown (2012, p.31) advocates a
choice of notational representations, and offers examples of software which uses none
whatsoever, such as jam2jam. Bamberger’s (2010) explorations into ‘student-invented
notations’ prompt questions regarding the development of personal understanding, and calls
for further research into the scope for implementation of these strategies in educational
resources (p.35). In terms of facilitating the intrinsic act of composition, many researchers
favour intuitive and tangible input methods, which rely upon familiar actions such as drawing
(Farbood et al. 2004; 2007, Raffle et al. 2007; Rosenbaum & Silver 2010), moving physical
items (Bean et al. 2008; Catala et al. 2011; Costanza et al. 2003; Marshall 2007; Zuckerman
et al. 2005) and bodily gestures (Bevilacqua et al. 2007, Fischman 2013; Ip et al. 2005). As
commercially-available interfaces become more technologically sophisticated, there is a
growing interest in the re-humanisation of the interactive process through recognisable
metaphor, a concept referred to as transparency (Fels et al. 2002; Gadd & Fels 2002).
However, such aims must also be reconciled with reasonable expectations as to what can be
made available in the music classroom and harnessed by the generalist teacher.

38
2.3 Summary

This research project aims to investigate a constructivist approach to the development of

interactive digital technology (i.e. user-friendly software resources) for supporting specific
musical learning. The context established in this chapter should further outline what is meant
by ‘a constructivist approach’. Constructivism is not an approach or method, but a lens
through which to view teaching and learning, and to provide some insight into the
effectiveness of specific methods with respect to learning impact (Shivley 2015, p.129).
Therefore, we can apply this to existing practices and resources to identify an area in which
we may support musical learning through digital technology:
 To what extent have interactive digital technologies hitherto succeeded in
encouraging musical behaviour in the classroom?
In this chapter, I have partially answered this question. Research suggests that real musical
behaviour is lacking in the classroom, most directly stated in the reports of OFSTED (2012;
2013). The pursuit of more experiential musical behaviours is therefore desirable, so it is
appropriate that we should consult a constructivist perspective on this matter. I have explored
an interpretations musical learning in this context, and some of the barriers currently facing
music educators. To further address this question, I will now examine, through the
constructivist lens, the range and effectiveness of interactive digital technologies widely used
in music classrooms (see chapter 3).
Earlier in this chapter, I posed a question: What does constructivism tell us about the
nature of learning that we can use to support specific musical learning? The research that I
have cited indicates that musical learning is effective when students are engaged in active
music-making as part of a social dynamic. However, certain barriers become apparent, such
as disinclination to experiment among more experienced students, and limited accessibility
among those with less experience. Then there is the suggestion of a deficiency in the creative
use of technology in music classrooms. Most students’ interaction with composing
technologies is limited to the standard MIDI keyboard and sequencer workstation. However,
studies indicate that access to a range of technologies and methods may encourage creative
collaboration. It is also suggested that the need to understand low-level musical concepts and
traditional modes of representation is not an effective precursor to creative progress, when
compared to more active, dynamic and student-directed learning structures. Research is
required to explore how alternative modes of representation and dynamic, open-ended

39
learning environments facilitate creativity, through both accessibility and the possibility for
subversion of standard musical practices. We have discussed how creative activities
providing opportunities for meaningful reflection might be made accessible to the musical
novice by encouraging improvisatory, creative musical behaviours. If we follow the ‘Bloom’s
21’ model, we make a creative task the starting point for discovery of abstract musical ideas,
which will then have a context of relevance and understanding to the learner, rooted in
concrete experience, according to the theories of Bruner (1961) and Kolb (1984). Therefore,
we must ask what role interactive technology plays in formulating such opportunities:
 How might interactive technology be harnessed to facilitate creative expression in the
musically untrained?
This is a further aspect of design and application within the classroom, and may involve
considerations of transparency in interface design, as well as ensuring the visibility of
musical learning (Hattie 2009). In primary music education, we are faced with not only a
skills gap, but also a difficulty in translating abstract musical concepts to meaningful
activities, often leading to an overreliance on verbal communication as a means of addressing
curriculum aims (OFSTED 2012, p.4). Where discussion is present, we must ensure that it is
a reflective response to real musical behaviours, and that which effectively communicates
musical ideas as part of a reflexive and social creative process.
Having explored and consolidated perspectives on constructivist theory, we have
begun to address some of the implications for facilitating musical learning. Just as there are
numerous perspectives on educational theory, what constitutes musical learning may be
interpreted in many ways, and there are many approaches we could take. In this project, we
are concerned with interactive technological solutions which extend possibilities for learning
beyond that of existing musical workspaces. While constructivism is not a method for design,
we must consider how the relevant ideas contribute to an informed method of design:
 How can constructivist theory guide interface design as a means of facilitating
creative musical learning?
By applying this view of learning in this project, we are examining two aspects of
development and impact: Firstly, we are investigating the extent to which interactive digital
technologies support musical learning within the classrooms that we visit as part of this
research project. This will be achieved by evaluating the musical outcomes of these sessions
(see chapters 6-7). Secondly, we are investigating my own practice as a teacher, programmer
and researcher, and the extent to which this practice develops and is successfully applied in
this project, in recognition of my own position as a learner in this process. This will be

40
evaluated by the outcomes of this project, and a key measure of this is the musical outcomes
produced by the students.
The next chapter will further explore the role of digital technology as part of a
musical learning environment, aiming to provide answers to the questions posed here, and to
further establish how my design practice may support musical creativity in the classroom.

41
 3. Research Context 2: Environment & Interaction

Figure 3.1

We have established that the constructivist perspective of learning concerns maintaining

equilibrium through interaction with our environment. What this learning environment is,
and how we interact with it, are increasingly technological considerations. Late 20th century
perspectives on the constructivist paradigm, such as constructionism, focused on the new
educational dynamics presented by virtual learning environments, and present day ideas such
as connectivism are wholly situated in a context of e-learning. This chapter will continue to
answer some of the research questions previously established:
 How might interactive technology be harnessed to facilitate creative expression in the
musically untrained?
To do this, I will explore the extent to which digital technology forms our environment and
directs our interactions in the interest of experiential learning, and will examine the wider
dynamic considerations of instructional technology within educational environments. I will
also discuss how digital media has influenced music education, and look at the existing
software resources used in the classroom:
 To what extent have interactive digital technologies hitherto succeeded in
encouraging musical behaviour in the classroom?
Finally, I will discuss the gamification of digital learning resources, and the implications for
formulating a design strategy, in order to address the question:
 How can constructivist theory guide interface design as a means of facilitating
creative musical learning?

42
3.1 Digital Technology in Creative Learning Environments

3.1.1 Visible Learning

Students often look for a context of relevance in what they are expected to learn. ‘When will I
ever need to know this?’ they may ask themselves, their peers or their teacher when faced
with some unit of learning that possesses no apparent significance to their lives. Instructional
technology has, in many cases, allowed curriculums to evolve beyond the rote-learning of
abstract knowledge. For example, the teaching of mathematics was restricted to ‘pure maths’,
with limited application in real-world circumstances, before the advent of computer
modelling and subsequent focus on problem-solving from the 1970s onwards (Puhlmann
2014, pp.78-9). While this ultimately led to the development of more meaningful
curriculums, a great deal of attention to design was required before teachers could confidently
incorporate these technologies into their lessons, and scaffolding tools such as the calculator
faced opposition from conservative voices who felt that this constituted a ‘dumbing-down’ of
mathematical processes (pp.77-8). This is a good example of the educational importance of
scaffolding: Square roots are immensely difficult and tedious to work out on paper but easy
to do with a calculator. The only reason not to use a calculator is if the method serves some
purpose applicable to our lives, but if we have calculators in our lives and in our pockets,
then it makes sense to use them. When we do, more complex problems with a context of
relevance to our lives, such as calculating distances, become accessible. This allows us to
take mathematics out of the classroom and into the real world, or alternatively, we could
work with virtual representations of locations and their respective distances on a computer.
Either way, we have a much more experiential lesson with clear and valuable outcomes. Our
students are now much more likely to develop an interest in mathematics and see how square
roots can be useful in their lives, and if that is the case, they will find out independently how
to calculate square roots on paper should the need ever arise, being equipped with perhaps the
knowledge but, more importantly, the motivation to do so. In this circumstance,
technologically-augmented learning environments have made the learning aims visible to the
learner.
Hattie (2013) tells us that learning is made visible when ‘teachers see learning
through the eyes of students and help them become their own teachers.’ This means
facilitating a clear understanding on the part of the student of not only what they are learning,
but why they are learning it. Collins (1991, p.5) identifies the educational advantages of

43
 computer modelling as connecting how and why with visible representations of processes
which would otherwise remain unseen, allowing us to synthesise concrete experiences from
what would otherwise be abstract cognitive processes. From a perspective of experiential
learning, if we understand why we are doing something, the dynamics of how become clear,
because we see, or perhaps actively seek to understand, how these processes ultimately
connect to the goal we already have in sight. Papert (1980, p.74) clarifies how this occurs
within the LOGO virtual learning environment:

In the LOGO environment new ideas are often acquired as a means of satisfying a
personal need to do something one could not do before. In a traditional school setting,
the beginning student encounters the notion of variable in little problems such as:
5 + X = 8. What is X?
Few children see this as a personally relevant problem, and even fewer experience the
method of solution as a source of power. They are right. In the context of their lives,
they can't do much with it. In the LOGO encounter, the situation is very much
different. Here the child has a personal need: To make a spiral. In this context the idea
of a variable is a source of personal power, power to do something desired but
inaccessible without this idea.

Figure 3.2 – Generating a spiral structure in Scratch. Notice the use of an anthropomorphic cat avatar as
a cursor, a version of the pioneering 'turtle graphics' of LOGO.

This point is still relevant to modern pedagogical programming environments such as Scratch
(Figure 3.2), and highlights what Ackerman (2001, p.8) describes as the situation-dependent

44
definition of intelligence within constructionism, where ‘being intelligent means being
situated, connected, and sensitive to variations in the environment.’ Because the virtual
learning environment can be modified, customised, and interacted with in a multitude of
ways, the capability of the learner to engage with the learning materials is defined by the
suitability of the environment. To ‘dive into’ such an environment, the learner must have a
clear understanding of what their objective is, and how changes within the environment, or
responses to their interactions, constitute movement in the direction of this objective. Collins
et al. (1991, p.3) advocate an experiential mode of learning designed for intrinsic motivation
and impact, where teachers make processes visible and relevant to students. Scaffolding
tasks, which allow the learner to engage with simpler units to build a ‘conceptual map’ and
thus understand the wider global objectives before focusing on local skills, are used in
various forms of apprenticeship:

In tailoring, apprentices learn to put together a garment from precut pieces before
learning to cut out the pieces themselves. The chief effect of this sequencing principle
is to allow students to build a conceptual map, so to speak, before attending to the
details of the terrain.
(p.15)

This approach mirrors the use of pre-composed structures in music resources such as eJay
(see 2.2.2) which allow the user to engage with global musical skills such as structure and
texture. The idea of ‘global before local skills’ is also analogous to the notion of Bloom’s 21
(see 2.2.1) where the higher order thinking skills of Bloom’s Taxonomy are used as a creative
starting point. In all of these models, the environment, often through technological
augmentation, provides a ‘conceptual map’ which promotes more specific ‘local’ outcomes
through less instructive processes, such as active experimentation and discovery learning.
The students then have other opportunities to learn or transfer what is learned as the situation
or environment diversifies (pp.15-6).

3.1.2 Tangible (‘Hands-on’) Learning

From a perspective of experiential learning, the construction of a conceptual map by the

learner is the product of enhanced modes of interaction with the learning concepts within the

45
environment. Papert’s term ‘objects-to-think-with’ (1980, p.11) implies a hands-on approach
to cognitive processes, where otherwise abstract ideas are made not only visible but tangible
through interactive technology. Through the construction of ‘meaningful public entities’,
students, peers and teachers have a clear manifestation of learning as it takes place:

For a while, I dropped in periodically to watch students working on soap sculptures

and mused about ways in which this was not like a math class. In the math class
students are generally given little problems which they solve or don't solve pretty well
on the fly. In this particular art class they were all carving soap, but what each student
carved came from wherever fancy is bred and the project was not done and dropped
but continued for many weeks. It allowed time to think, to dream, to gaze, to get a
new idea and try it and drop it or persist, time to talk, to see other people's work and
their reaction to yours -- not unlike mathematics as it is for the mathematician, but
quite unlike math as it is in junior high school.
(Papert & Harel 1991, p.4)

This suggests an alternative to the modular and sequential approach to learning, instead
advocating a structure more commonly found in artistic subjects, which emphasise the
choices and decisions of students and allow them to apply skills in a variety of cross-
curricular contexts11. Though the structure employed within classes like the one which Papert
observes is, at least in part, a consequence of the less rigid control imposed on them as ‘non-
core’ subjects (auxiliary to the traditional Western educational triad of English, science and
mathematics) he values this structure as unrestrictive to the personal development of the
students, and seeks to incorporate it into his own subject, mathematics. By producing learning
materials which the student can see and ‘sculpt’, in this case, computer-generated, Papert
pursues a mode of learning rooted in ‘problem-creation’, where continual acquisition and
application of knowledge that is driven by the choices and orientations of the individual
learner (Harel & Papert 1990, p.24).
Consider the practice-based research model: The practitioner-researcher begins not
with a problem, but with their individual creative process, from which some knowledge is
extracted. They then apply this to a new creative process, and so the reflexive cycle
progresses:

11
Such an approach, termed project-based learning, is common in the Finnish educational system. See, for
example, Helle et al. (2006).

46
 The innovative and critical potential of practice-based research lies in its capacity to
generate personally situated knowledge and new ways of modelling and externalising
such knowledge while at the same time, revealing philosophical, social and cultural
contexts for the critical intervention and application of knowledge outcomes.
(Barrett & Bold 2014, p.2)

This model has afforded a rigorous mode of inquiry to the creative arts, one that allows the
practitioner-researcher to investigate how context shapes their practice and vice versa. In
social constructionist theory, this awareness is termed ‘reflexivity’ (Cousin 2016) but has a
clear parallel to the constructivist theory of equilibration, that of ‘coupling with our surround’
and ‘[acting] on new experiences and information’ (Fosnot and Perry 1996). Reflexivity can
be thought of as an analysis of this process, a constructed awareness of how it happens. It
seems, then, that this model has central relevance to educational practice within the creative
arts.
The crucial distinction in this form of learning is that it is not a matter of an individual
changing as they absorb an unchanging body of external information. It concerns interaction
with an environment (setting, context, task, resources, other people etc.) and the feedback of
this experience. This feedback is itself a dynamic process, something which the learner has an
influence over. It is shaped by their perspective, the mechanisms by which they internalise it,
assimilate it, or possibly reject it. This then manifests in some practical change, the means by
which they re-interact with their environment. But, in doing so, it is not just the individual
undergoing change, but also the environment with which they are interacting. This is the
archetypal mode of observation for the practitioner-researcher: How and why does my
practice influence my environment? How and why does my environment influence my
practice? We might say that the practice-based researcher is engaged in constructing (or
perhaps re-constructing) their own conceptual map through these practical interactions and
reflections, and this allows them to challenge concepts from their field and contribute new
knowledge.
The extent to which young learners can engage with the creative modes of enquiry
here described depends upon the provision of a suitable environment, in which they can apply
their existing skills to negotiate new concepts, allowing them to venture further still (i.e.
express creativity):

47
 ‘[Children’s] willingness to improvise and compose is a function of creating an
environment where children can express their creativity. By starting with activities
that are not too far removed from the child’s immediate experience, creativity
becomes integrated with the child’s existing musical experiences and skills.
Furthermore, by locating children in a range of musical settings they come to
recognise the multidimensional nature of what they already know, think and can do.’
(Burnard 2000, p.21)

This suggests maintenance of a ‘flow state’ (see Czikszentmihalyi 1996) in which present
challenge should increase in proportion to acquired skill; a change in environment in
response to the needs of the learner, or an environment which adapts to the learner, mirroring
the cognitive dynamics of equilibration. How might interactive digital technology contribute
to such an environment?

3.1.3 Augmented Learning

In e-learning, designs which change in response to the learner are referred to as ‘augmented
learning environments’. Augmentation in virtual learning environments may constitute, as we
have discussed, visualisation and concretisation of ideas, something as simple as an assistive
onscreen menu or as complex as a system which learns from the learner12. However, reliance
on this kind of ‘big data’ has led to criticism among educationalists wary of a ‘de-
humanisation’ of teaching and learning. Rouvroy and Berns (2013) refer to as ‘digital
behaviourism’ where our learning experience is defined by a codex of numerical ‘truths’.
Education is about emotional as well as cognitive connections, which occurs and is
interpreted on a person-to-person level. If we lose sight of this, we may see a revival of the
Pavlovian behaviourism that preceded the emergence of constructivism, albeit veiled by the
sheen of technological augmentation.
In order to help students to become actuators of their own learning, technology has to
form a scaffold which ultimately affords them more control, as opposed to directing their
interactions. Also, because we are primarily concerned with creative outcomes, we should
consider how our interventions augment the human element of learning along with the

12
This kind of technological responsiveness is present in emerging examples of highly software-oriented
education, such as the AltSchools educational start-up (see Robinson 2017) and AI teaching assistant (see
Backchannel 2017).

48
technological. Augmented learning environments may rely disproportionately on digital
technology to provide a dynamic component which could be achieved through social
interaction. For Collins et al. (1988, p.26) sociology is a ‘critical dimension that is often
ignored in decisions about curriculum and pedagogical practice.’ Modes of co-operative
learning feature exclusively within the instructional technology-oriented model of cognitive
apprenticeship (Collins et al. 1988; 1991). De Corte (1990) defines a ‘powerful’ problem-
based learning environment with reference to LOGO, but establishes methods centred around
human interactions such as expert modelling, guidance, feedback and reflective observation,
placing particular emphasis on any techniques which enable students to articulate what they
have learned (pp.12-3). This is echoed in more recent writings on new pedagogies driven by
technology:

The future of teaching may ultimately center in deeper relationships built between
teachers and students, developed through creative, collaborative, socially connected
and relevant learning experiences. Technology can enable and accelerate these deep
learning relationships — both between teachers and students and between students
and other “learning partners” such as peers, mentors, and others with similar learning
interests. Technology as a platform for more connected social learning experiences is
a far cry from the notion of technology supplanting teaching.
(Fullan & Langworthy 2013, p.14)

When augmented learning with technology, we should consider that which legitimately and
quantifiably augments the learning experience, or presents new modes of learning, and to
question that which is used for the sake of novelty and mere efficiency, especially where this
diminishes the effective social components of education.

3.2 Digital Technologies in the Music Classroom

3.2.1 Overview

Fullan and Langworthy (2013, p.4) argue that digital technologies present ‘new pedagogies…
enabled and accelerated by technology’ and promoting deep learning, which they define as
that which ‘develops the learning, creating and ‘doing’ dispositions that young people need to

49
thrive now and in their futures’ (p.i). This can be seen not just in the inherent engagement and
links to the extracurricular activities of students, but in the vast range of unique modes of use
across almost all subject areas. Relatively recent technologies such as touchscreen tablets
have been widely incorporated into the teaching of core subjects, such as literacy and
numeracy. We are only just starting to gather evidence on how tablets and other personal
devices may be beneficial within non-core subjects such as music: Riley (2013) documents
the various uses of the iPad in music lessons. Criswell (2011) examines teacher perspectives
on using the iPad for performance and composition. Most reports focus on the practices of
innovative teachers, on an individual case study basis. Riley (2016) suggests six apps that
may be used for classroom composition with young children, and provides sample lesson
plans. Some of these apps offer intuitive interactions with the touchscreen as ways of
manipulating timbre, such as
Singing Fingers and Brian Eno and
Peter Chilver’s Bloom app (Figure
3.3). Ruismäki et al. (2013)
document the use of the iPad for
practical learning in a Finnish music
class, noting the scaffolding
possibilities of GarageBand
(pp.1091-2).
Technology is always
Figure 3.3 – Bloom
subject to change, and the
commercial software resources used at the time of writing may not still be in use in years to
come. Nevertheless, it will always be useful to have some record against which to compare
new resources, so I will now examine the different types of software and applications which
may be employed in the music classroom currently. In order to categorise these resources as
coherently as possible, we can consider the type of musical activity they facilitate.
Swanwick’s (1979) influential definition of a music curriculum focused on three central
activities; audition (or listening and appraising), performance and composition. These are the
three basic strands of activity on which music education is founded; we are either perceiving
(usually, though not exclusively, by listening) and describing, playing or writing music.
There is, of course, a great deal of crossover of these approaches within the available digital
resources.

50
 3.2.2 Listening and Appraising

Digital resources for audition of musical

examples, such as listening exercises or
sight reading musical text such as staff
notation or tablature, often comes in the
form of software packages. These
packages are generally aimed at increasing
musical knowledge across a broad
Figure 3.4 – Practica Musica sequence of activities. However, many
focus on specific curriculum or assessment
content, of an awarding body such as the Associated Board of the Royal Schools of Music
(ABRSM) or the GCSE Music curriculum. As such, these packages are outcome-oriented,
designed to be used by an individual and making it easy for them to track their own progress,
and imparting a broad range of musical information through quizzes and other exercises.
However, this does mean that such packages tend toward a prescriptive and instructive
format. Software such as Ars Nova Practica Musica (Figure 3.4) and Rising Software
Musition is based on music theory, with
questions and exercises addressing the
identification of musical features. These
are often available as part of packages
containing other types of music training
software. Aural training resources such as
Rising Software Auralia (Figure 3.5) are
also common, which advertise as helping
users to develop perfect or relative pitch Figure 3.5 – Auralia

by identifying notes and intervals.

There are a number of software packages aimed at assisting teachers to deliver the
music curriculum. By far, the most prominent example is Charanga Music World, which
offers a sequential and learning outcome focused set of activities closely following the
curriculum. Charanga draws from research in music education, with the structure following
Swanwick and Tillman’s (1986) sequence of musical development, where learners alternate
between personal and social musical experience (Figure 3.6). It offers a range of activities as
pre-planned lessons drawing exhaustively from the curriculum, with music history and

51
listening exercises alongside performing and composing tools (Figure 3.7). Charanga
therefore has a much more collaborative and perhaps less instructive format than the other
software packages discussed.

Figure 3.6 – The sequence of progression in Charanga Music World, alongside the sequence of musical
development (Swanwick & Tillman 1986)

Figure 3.7 – Assistive tools for composing and performing in Charanga

3.2.3 Performance

Assistive digital resources for musical performance have tended to focus on tracking
accuracy and highlighting errors, such as timing issues and incorrect notes, so have been
designed around some means of digitising musical performance. As such, most have used

52
 MIDI, with input devices such as the
electronic keyboard or drumkit (Figure
3.8). Advances in frequency analysis and
pitch detection, as well as motion
tracking with web cameras, has led to the
capacity for computer analysis of more
complex and discrete features of musical
performance. Some researchers have
explored the possibilities of a de facto
Figure 3.8 – Home Concert XTREME for iPad digitisation of the classical music tutor in
the interest of enhancing performance
scrutiny and accuracy (Yin et al. 2005;
Fober et al. 2004; 2007). Other packages
utilise similar methods of input to
produce a more fun and casual format
aimed at the novice learner, for whom the
purchase of a piece of software might
prove more cost effective than ongoing
Figure 3.9 – Yousician
lessons (Figure 3.9).
The gamification of musical performance using interactive digital media was no doubt
propelled by the commercial success of Guitar Hero with the levels of dexterity and
‘virtuosity’ among habitual players arguably comparable to the results of legitimate musical
practice (Arsenault 2008)13. The
onscreen interface of falling notes
which must be synchronised by the
player using the connected
hardware has been adopted by other
performance tools such as Synthesia
(Figure 3.10) and popular
pedagogical apps such as Piano
Tiles. Other assistive performance Figure 3.10 – Synthesia

13
Further musical credibility is lent to the ability to skilfully operate a computer game controller or other
commercial input device by the possibility and widespread practice of mapping such devices to musical outputs
using audio-oriented programming environments such as Max/MSP.

53
resources take on a less linear format, such as Eno’s Bloom app and Brown and Dillon’s
jam2jam (see 2.2.3). The latter offers the freedom of ‘jamming’ in a collaborative virtual
environment by arranging musical ideas using a touchscreen, and as such offers much more
capacity for exploratory and creative performance than the more instructive resources
previously discussed.

3.2.4 Composition

The majority of commercial software

for music composition falls within two
categories; scorewriters and multitrack
Figure 3.11 – Staff notation in Sibelius Ultimate
sequencers. Both types tend to use
MIDI for note identification, instrument classification and imitation, as well as the control of
various musical parameters, such as volume and pitch-bending. Scorewriters (Figure 3.11)
are based upon the system of staff notation used in the Western classical tradition, allowing
users to create, edit and print musical scores. Such programmes tend to be aimed at trained
musicians, offering a sophisticated lexicon of musical representations but little in the way of
assistive support to the novice user. Multitrack sequencers (Figure 3.12) draw from studio
technology, offering the direct recording and processing of audio, and often a set of virtual
instruments (Virtual Studio Technology,
or VST) which are typically modelled
after popular hardware synthesizers, drum
machines and effects units. The usual
linear timeline and commands (‘cut’,
‘splice’, ‘loop’) draw from the traditions

Figure 3.12 – Arranging clips in GarageBand of audio tape recording and studio
practice, and therefore are also fairly
demanding for the novice user, though some examples of digital audio workstations have
taken on a more assistive format with scaffolding tools, such as eJay (see 2.2.3) or the drum
mixing matrix within GarageBand (Figure 3.13).

54
 Composing software aimed
specifically at children tends toward a
more straightforward mode of
arrangement, where musical ideas can be
easily differentiated from one another.
These are often built around the concept of
a more child-friendly activity, such as
painting (see Rosenbaum & Silver 2010;
Figure 3.14). Bamberger’s research
pioneered the use of ‘tune-blocks’ Figure 3.13 – Arranging drums in GarageBand

(Bamberger 1996; Bamberger &

Hernandez 2000), providing melodic
fragments to explore composition as a
form of puzzle solving (Figure 3.15). This
scaffolding approach has been shown to
facilitate an intuitive creativity in
musically-untrained students, who were
then able to provide well-reasoned
responses as to their creative decisions
(Bamberger 2003). Brown (2012) cites Figure 3.14 – Morton Subotnick’s Pitch Painter
Bamberger’s work as influential to the
development of jam2jam.

Figure 3.15 – Impromptu

55
3.3 Learning from Games

3.3.1 Sandbox Games

Games have long been recognised as an effective learning impetus. Teachers have been
gamifying learning materials with quizzes and other competitive challenges throughout the
history of the profession. When Piaget established play as a significant component of
educational discourse, the value of games beyond merely measuring knowledge or provoking
competition became apparent – The child playing in a sandbox, in developing skills and
ideas, however rudimentary, which this environment facilitates, is engaged in learning. Papert
and Harel (1991, p.1) compare the process of building a sandcastle to building a ‘theory of
the universe’, acknowledging how creative interactions with our environment underpin the
lifelong process of learning. We have discussed how, in modern constructivist approaches,
the virtual or augmented learning environment may be designed for and adapted by the
learner. This type of environment may be thought of as a sandbox, where the learning
materials are, in some sense, directed or ‘sculpted’ by the learner, but where imagination and
personal objective are the vital ingredients which the learner brings to the process.
In the commercial gaming
industry, ‘sandbox’ refers to a type of
gaming environment, where the player
sets their own objectives and defines
their own outcomes. The educational
potential of sandbox games is not
necessarily inherent. Rather, the link is
Figure 3.16 – Sim City derived from their usefulness as
resources in the kinds of open-ended
learning context that are valued by constructionist educators. The term fell into use with the
growing popularity of Sim City (Figure 3.16) and The Sims – open-ended games which
involve the maintenance or development of a state as the sole objective of gameplay. In Sim
City, for example, the player initiates development which they can then observe as it unfolds
further – They may build a hospital, upon which the autonomous sprites of the game add
roads and houses around the structure. Other generative aspects are entirely out of the
player’s control, such as natural disasters, and must be mitigated through protective or
restorative measures. However, ‘sandbox game’ has been generally applied as an umbrella

56
term for any open-world platform which does not follow the linear game format whereby the
sequence of events is always essentially the same. From this perspective, first-person shooter
or open-world games such as Grand Theft Auto might be categorised as sandbox games,
though the context of creative exploration implied by the term is largely absent.
As a demonstration of
how sandbox environments
might support or motivate the
novice user, let us consider
the gameplay in Minecraft.
The primary objective is to
build structures using a
catalogue of resources, most
of which are represented as 3-
Figure 3.17 - Minecraft
dimensional blocks, and other
items with specific properties or functions. The simple cubic graphics allow for the easy
construction of basic structures, such as small houses, while more complex designs can be
achieved by scaling up the number of blocks used, or the range of different items, so a vast
range of possibilities are present. When we begin a game as a new player, there is no
apparent objective or instruction in place. We find ourselves in one of several landscapes – a
field, a beach, a forest – with a number of onscreen status bars giving indication as to our
present condition (Figure 3.17). As we explore these surroundings, we will find that we can
collect some items, and combine them to craft other items, but that others are unavailable to
us until we obtain more resources. As these possibilities grow, we set further objectives for
ourselves. These objectives often have a sequential format, with some wider goal being
dependant on the fulfilment of
a set of smaller goals; e.g. I
need to find A so that I can
make B so that I can
accomplish C… As such, the
player acquires skills as they
move through the game, setting
their own objectives for
Figure 3.18 – Minecraft players often build and share complex
gameplay. structures

57
 Minecraft is a particularly pertinent example of a sandbox game due to the varying
levels of complexity and possibilities for social and personal creative activity. The player is
both an explorer and master of a virtual world14, which they can customise with built
structures of endless complexity (Figure 3.18). There is therefore a certain creative ownership
and personal identification in such gameplay, and the possibility of world-sharing and
networking, whereby players can enter and explore each other’s worlds, qualifies this as an
example of social creative practice. To use Papert’s term, the player is ‘constructing a
meaningful public entity’ (see 3.1.1).

3.3.2 Educational Applications

There is a growing body of research

concerning effective classroom
implementation of the commercial
open-ended games discussed here
(see, for example, Bos 2001; Denis &
Jouvelot 2004; Gaber 2007; Brand &
Kinash 2013; Ekaputra et al. 2013;
Schifter & Cipollone 2013; 2015).
The possibilities for engagement and
development afforded by Minecraft
Figure 3.19 – A class of children, potentially across multiple
schools, explore a teacher-generated world in MinecraftEdu have led to an unprecedented level of
commercial computer game
integration with educational practices (Schifter & Cipollone 2013). Joel Levin’s blog The
Minecraft Teacher, which started in 2010 as a way of demonstrating applications for the
game in various school subjects, eventually grew into MinecraftEdu (2011; Figure 3.19), an
officially-supported version designed for teachers. This format allows the teacher to take the
class into a virtual world, whereby they can use curricular content to produce something
meaningful. This is the same premise described earlier by Papert, where variables were used
to graph a shape in the LOGO programme.

14
Procedural generation in modern open-world games such as Minecraft means that the player explores a
theoretically infinite space, as the virtual landscape is determined not by direct manual design but by a
generative algorithmic procedure.

58
 Through MinecraftEdu, and other similar online communities such as Massively
Minecraft, educators share lesson plans and even Minecraft worlds (which can be shared as
with any other digital file format). The game has also been used to produce various
modifications of an educational nature. Circuit Madness, posted by user deco2000 on the
forum Planet Minecraft in 2012, harnesses the circuit-building functions of the game to pose
a series of challenges. With the block-breaking functions disabled, the player must rely on
logic and knowledge of circuitry to navigate the space and complete the game. This
demonstrates how sandbox games such as Minecraft have the potential not only for discovery
learning and development within an open-ended virtual environment, but also the capacity for
problem creation. Case studies exists of the game being used to teach programming skills
(Repenning et al. 2014; Roscoe et al. 2014), scientific concepts (Short 2012), mathematics
(Bos et al. 2014) and collaborative artwork (Overby & Jones 2015). Bos (2001, p.3), using
Sim City as an educational tool, suggests that feedback should be instrinsic, relevant to
accomplishment of the task at hand, rather than extrinsic, such as scoring points or winning
prizes. Where students are engaged by a task and have an inherent desire to do better at it,
effective teachers are able to reconcile learning outcomes with motivational feedback. Again,
this relates to Papert’s idea of ‘constructing a meaningful public entity’.

3.4 Summary

In this chapter, I have suggested that interactive technology best aligns with the constructivist
paradigm when it can be used to present an active and exploratory scaffolding environment,
and where it promotes the social functions of learning. Within this context, digital resources
which present musical concepts sequentially are less relevant than those provide a visible and
tangible means for interacting with such concepts at a global level, and opportunities to apply
these skills in a variety of open-ended settings. Resources like Charanga are very useful for
illuminating curriculum content, and for being accessible for generalist teachers and novice
learners. However, the open-ended format of digital audio workstations, not being
prescriptively focused on specific examples or exercises, present further opportunities for
discovery learning and creative ownership on the part of the student, where they can be
effectively utilised as part of a socially-driven educational environment.
While there are many music applications aimed at the very young, there are far fewer
music-making programmes specifically designed for school-age children (see 3.2). It is likely

59
that the reason why we have seen an insufficient impact of technology on music education is
because there is not enough music software designed specifically to support curriculum
learning, or to assist less musically experienced students and teachers. My own background is
principally in secondary school education, where I have found that each school tends to have
a subscription to a commercial composing software package, such as Logic or Cubase, which
is then used for all ICT-related music activities. Because these programs are not aimed at
children, new students face a difficult task in learning how they work, and such lessons often
focus disproportionately on the operation of the program rather than the wider relevance of
the music activity. This is perhaps an explanation for the ineffective applications of
technology that OFSTED (2012, p.54) criticized:

While they were kept engaged by the tasks and enjoyed working with the sounds,
most found it difficult to explain the reasoning behind their choices. Opportunities
were missed to develop understanding of the musical syntax, form or the sampling
processes that underlay the creation of the pre-composed loops and riffs that the
students were using. Even where pupils were creating their own musical ideas in step-
time or in real-time, limited evidence was seen of them going beyond the initial
inputting of notes to shape the dynamics, articulation or subtleties of tempo for their
ideas.

In this example, the students are unable to reflect upon their musical activity because the
structures being employed did not effectively communicate the musical concepts in question,
and the experiential learning cycle is un-propelled. Hence, further levels of sophistication in
shaping musical ideas remains unexplored.
The arrangement of ‘pre-composed loops’ can provide an accessible means to ‘select
and combine sounds’ (DfE 2013, p.2) that produces a coherent musical result. However, to
move beyond this rudimentary key stage 1 skill, we must explore how the approach functions
at different levels. We might ‘zoom out’ and ask, ‘How do these building blocks form a
verse, a chorus, an overall song structure?’ Conversely, we might ‘zoom in’ and ask, ‘What
are these building blocks made of, and can we alter them for some intended purpose? What
are these sounds? Can we cut them up into smaller blocks?’ Such an approach is employed in
the sequencing software eJay (see 2.2.3). To paraphrase the music curriculum, these
examples can be thought of as ‘other musical notations’ through which the ‘inter-related
dimensions of music’ (DfE 2013, p.2) may be made apparent, or may become interactively

60
attainable, providing that there are opportunities for students to further explore the shape and
subtleties of musical materials. There are several well-established examples of composition
using larger, meaningful structures that are later broken down into constituent components.
Notably, rhythm is traditionally introduced using word-rhythms that allow children to easily
conceive and memorize rhythmic groupings. Drink names are typically used: tea (one
crotchet), coffee (two quavers), lemonade (two semiquavers and a quaver) and so on. Simple
rhythmic compositions are then easily represented. Once a system of rhythmic patterns has
been established through musical activity, the concept of duration has some context of
relevance to the children, and the groupings can be further broken down into smaller
durational components. This approach has an empirical basis in Bamberger’s tune-blocks (see
3.2.4).
Reflecting on the questions already established, I have asked how interactive digital
technology may support creative expression in novice learners, and how constructivist theory
may inform an effective interface design for musical learning. In summarising the ideas
around interaction in digital learning environments that we have addressed in this chapter, I
propose the following principles for a creative music interface. It should:
 Be based on meaningful experiential activity
 Allow students to perceive, interact and engage with fundamental musical concepts
 Yield evident outcomes in terms of what is learned (in curricular terms), but also in
the production of a personally relevant creative achievement
 Be open-ended in format and encouraging exploration
 Facilitate social and collaborative learning
 Facilitate meaningful musical dialogue
 Support students in reflecting upon their musical work
 Have an accessible user interface for novice learners and non-specialist teachers
The kind of interface with which we are concerned would provide an environment in which
students can interact with musical concepts in an active, meaningful and social way.
We have also discussed how certain open-ended platforms, termed ‘sandbox’ games,
have been harnessed for educational purposes. This prompts a new principle for interface
design, as well as a new question:
 What conditions prompt creative musical learning in ‘sandbox’ environments?
When the term ‘sandbox’ is used in this project, it refers to learning environments which are
open-ended, exploratory and creative, but also accessible. In this type of environment, the
starting point is the improvisatory ideas of the user, and the learning materials are objects-to-

61
think-with, which allow the learner to produce a meaningful public entity and to form their
own conceptual map. Just as the child in the sandbox has tangible materials to shape
according to their imagination, so the learning materials provided in a sandbox environment
should allow the user to manipulate the creative concepts in play.

62
 4. Methodology

Figure 4.1

This chapter describes the research model adopted in this project. This is informed by two
methodologies: A newly-proposed model for software design as educational research,
supported by the established methods of grounded theory, which also contribute to the wider
questions and theoretical outcomes. An overview of the research tools, timeline and process
is also described.

63
4.1 Choosing a Methodology

4.1.1 Considerations

Through our examination of constructivism in music education, and within digitally-

augmented learning environments, we have arrived at a set of initial design principles (see
3.4). It is important to clarify that in this context of instructional design, we are concerned not
just with the production of a software artefact, but also a consideration of how this can be
applied in a music lesson, the cognitive and social dynamics of such application, and its place
within a wider framework of class and curriculum. Jonassen (1999, p.222) discusses this kind
of constructivist design as defining a ‘problem manipulation space [which] provides the
objects, signs, and tools required for the learner to manipulate the environment’ as learners
‘cannot assume any ownership of the problem unless they know that they can affect the
problem situation in some meaningful way.’ This further illustrates the definition of sandbox
as it is used in this thesis; a space for creative ownership over the problem itself as well as its
solution, such that the outcomes are intrinsically relevant or valuable to the learner. Such a
space must also employ what Jonassen (1999, p.222) terms authentic representation or that
which is ‘personally relevant or interesting to the learner’, akin to the capacity for personal
identification Papert describes in his objects-to-think-with (see 2.1.4). This is an important
form of scaffolding, as it helps the learner to engage from the point of first use by building
upon their existing comprehension.
This project requires a design methodology that will allow us to address these
considerations. Methodologies for design and development, for software but also as a general
rule, can be categorised as either top-down (starting with the goal design and then refining the
constituent details) or bottom-up (an emergent process of piecing together from developing
components). This model can also be applied to instructional design; Borgo (2007, pp.85-6)
refers to this dichotomy when comparing rote-learning of musical concepts with discovery by
improvisation and experiential learning:

Complicated systems tend to involve a top-down model of organization that requires a

strict hierarchy and the unerring execution of many sequential operations… Complex
systems, by contrast, tend to involve bottom-up or ‘self-organizing’ dynamics that
rely on extensive communication throughout a network of highly interconnected

64
 parts… In short, complex systems can produce ‘emergent’ behaviours; they offer the
possibility for surprise.

The constructivist orientation of this research project aligns it with the idea of complex
systems, or bottom-up design. We have a set of considerations, principles and initial target
outcomes, which we will bring together and measure in a learning context, to achieve
equilibrium between these component ideas and thus produce a meaningful solution.

4.1.2 Software Development as Educational Research

Reflecting on the developmental process of jam2jam (see 2.2.3), Brown (2007a, p.4)
proposes a methodological model which ‘is concerned with deliberately employing designed
systems that embody a hypothesis about educative experiences, in order to illicit new
research findings.’ Software Development as Educational Research (SoDaR) situates
software design in a context of constructivist educational research (see Brown 2012), an
iterative process which draws directly from student experience and learning outcomes. The
methodology aligns with complex system design, recognising the validity of emergent
behaviours and rejecting the notion of absolute faithfulness to the initial design specifications
present in top-down models:

An important SoDaR feature should be the maintenance of a healthy skepticism about

the degree of control the researcher has over outcomes. In this way control is
understood as improvisation rather than direction and there should be a view to
generating knowledge by capturing the opportunities that arise…
(p.10)

This sensitivity to learning opportunities comes from testing and developing the resource in
an experiential and educational setting15.
Brown defines the approach as having three stages, and suggests possible research
questions for each stage:

15
Brown (2007a, p.10) notes how the text messaging component of jam2jam became a prominent design feature
after its unexpected popularity among the students was observed.

65
 Stage 1: Define the activity
 How will the activity lead to the desired learning outcomes?
 What educational value would be provided by the software?
 Why is software the best medium for providing this experience?
 Is the activity described at an appropriate level of detail?
Stage 2: Software Design and Production
 What data structure best supports (technically and pedagogically) the domain
knowledge being represented?
 How does each software feature reinforce the design objective?
 What software platform will best enable production and deployment?
 Has the design and production process limited or expanded the educational
implications?
Stage 3: Usage and Refinement
 Are the activity and software mutually reinforcing?
 What are the differences between the expected and actual behaviour of the
students?
 How can the software and its use be improved?
 Are the students achieving the desired learning outcomes?
(Brown 2007a, pp.5-7)

This can be characterised as a qualitative research process which seeks to determine specific
added value from the application of developing software in educational settings. Being
situated in and emerging from particular environments and contexts, SoDaR is undertaken
with an interpretivist perspective that the experience of individuals and the influence of their
surround should be evaluated:

Research findings generated using the SoDaR approach are limited to the studied
contexts in the same way as other methods used within qualitative research studies.
While this can be alleviated somewhat by studying multiple sites or classes,
generalizations should be made with caution and context dependency taken in to
account.
(p.9)

66
The process has clear links to action research in education (see Mills 2000; Carr & Kemmis
2003). The working environment, in this context, may be taken to be the school environment
or the virtual environment. Receptiveness to unforeseen possibilities is emphasised within the
SoDaR method (Brown 2007a, p.10). Though generalisations are to be viewed with caution,
significant results demand further examination and should be presented in context. This is in
line with Bassey’s recommendations regarding ‘fuzzy’ generalisations (1998; 2001).

4.1.3 Grounded Theory

Grounded theory is a research methodology originally proposed by Glaser and Strauss (2017)
which has been described as compatible with the constructivist ontology (see Charmaz 2014;
Mills et al. 2006) and the work of educational practitioners, among other social fields. As the
grounded theory researcher collects data, often supported by the writing of memos, they look
for codes, or recurring units of qualitative data (words, ideas, etc.), which form concepts.
These are then grouped into categories, from which the theory ultimately emerges
(Breckenridge et al. 2012). Due to the exploratory and flexible nature of grounded theory, a
number of variations have been proposed. For Glaser (1992), an emergent approach,
characterised by an essentially consistent application of coding methods at micro and macro
levels of analysis, is favoured. This argument is made in his Basics of Grounded Theory
Analysis, a response to the methods elaborated by Strauss (1988). Charmaz (2000) proposes
that the grounded theory researcher acknowledge and examine his or her own perceptions to
mitigate bias. Glaser (2002) argues that pattern received by the researcher, and resultant
theory, are all part of the data and methodological process.

4.1.4 Mixed Methodology

SoDaR is a suitable model for this project, but also one which was only recently proposed,
and which has not undergone substantial testing. I therefore decided to base my practical
research model on SoDaR, but to combine this with grounded theory. This serves two
purposes:
1. This research is practice-led, and therefore has two main outputs; a practical output –
in this case, the software artefact – and also a theoretical output emerging from the
application and examination of this practice. While SoDaR enables me to produce the

67
 practical output, this requires support from a more theory-oriented methodology to
analyse and comment upon the process. The combination of these methods therefore
allows me to fulfil both aims.
2. Again, because I wish to comment upon my practice, grounded theory allows me to
step back and review the process at every stage, including critical analysis of the
methodology itself. I can therefore analyse in turn the effectiveness of this
methodological approach.

By combining SoDaR with the analytical coding methods of grounded theory, I aimed to
follow a complex system of design that is responsive to emergence and inclusive of the
interconnected considerations of software development, musical creativity, and educational
practice.

4.2 Research Tools

4.2.1 Programming Platform

The programming platform is a research tool which provides the software artefact, or
practical outcome. This in turn becomes a research tool for producing a theoretical outcome,
through application in educational contexts. The software platform used for developing this
resource will be Pure Data (Pd), an open-source, object-oriented programming language
suitable for audio interfaces. Pd presents a number of advantages over other possibilities:
 As an open-source language, it is freely available and developed by a supportive
online community. This allows for the production and sharing of a freeware resource
which can be further developed and customised by the wider user base (see Hart
2017a, p.414).
 Object-oriented coding languages allow processes to exist as manageable structures,
making the programme easier to edit in response to outcomes emerging from its use.
This makes it more suited to bottom-up programming than pure text-based languages
such as C or Python. It also allows me to design, develop and deploy software
artefacts using a single platform.

68
  Pd has many user-developed external libraries for a range of purposes. One of the
most widely used is the graphical processing library GEM (graphics environment for
multimedia) which will be useful for generating our objects-to-think-with.
 Pd runs on all major operating systems, and can be used as a processing platform for
mobile interfaces and apps.
However, there are also disadvantages to using Pd over other software platforms: As an
open-source platform with no commercial interest, developments can be unreliable, and
unstable releases are occasionally encountered. The platform Pd-extended, which combines
the core language with external user libraries to extend the functionality, was used for much
of this research project. The final stable release of this platform was in 2014, after which it
was abandoned by the community. Many users (including myself) continued to use Pd-
extended, as the same range of possibilities was not immediately available within Pd-vanilla
(the core language). Later releases of Pd-vanilla allow the user to install and connect these
external libraries. The developed version of the software artefact produced for this research
was eventually migrated to Pd-vanilla, to allow for further continual development. However,
this means that the external libraries have to be added by the user in order for the programme
to work, where in Pd-extended they were immediately available.

4.2.2 Additional Software

Besides Pd and the software artefacts produced in this platform, a number of other software
resources were used in this research. Bespoke applications were made using MobMuPlat, a
mobile music platform for building graphical user interfaces with audio engines made in Pd
(see 7.4; 7.7; 7.8). Ableton Live was used as a sound source in early exhibitions (see 6.1; 6.2).
Finally, the digital audio workstation GarageBand was used as a counterpoint to the earlier
research in a final series of studies (see 7.8).

4.2.3 Hardware

Many of the hardware resources used for this research could be provided by the schools;
percussion instruments and keyboards were typically available, as were headphones.
Electronic resources, such as laptops and touchscreen tablets, were often also offered by

69
schools. However, due to the need to develop bespoke applications, all laptops and tablets
were sourced prior to the studies taking place.
A stock of iPads was available for loan at the university, and were borrowed for some
of these studies (see 7.4; 7.8). A sufficient supply was available to provide one per pupil,
even when working with entire classes. Laptop computers presented a greater challenge to
source; several second-hand laptops were purchased and restored specifically for this project,
while others were borrowed temporarily. Consequently, the maximum number of laptops
used in this research was limited to eight (see 7.6). This issue was often resolved with smaller
group sizes, so headphone splitters were also frequently used (for iPads as well as laptops). A
set of eight 6-way headphone splitter units were obtained to enable group work on laptops.

4.3 Research Model

4.3.1 Timeline

In SoDaR, the software artefact is continuously weighed against learning outcomes to

identify the criteria for its refinement; the extent to which the desired outcomes are achieved,
the presence of any unexpected outcomes, and the question of whether software and activity
are ‘mutually reinforcing’ (Brown 2007a, p.7). As such, the developmental relationship
between the software and learning activity can be thought of as a form of equilibration.
Brown acknowledges the links to reflection-in-action (p.6), Schön’s (1987) model of
reflexive practice, which has parallels with the cycle of experience and reflection proposed by
Kolb (1984).

Figure 4.2 – Gantt chart of phases 1-3

70
This research project was characterised by distinct phases (Figure 4.2):

Phase 1: January – August 2015

During this phase, the activity was defined through analysis of relevant literature (including
the National Curriculum for Music). This also led to the formation of the research questions
(as documented in chapters 2 and 3). Initial developments and designs were explored to
produce a prototype software artefact, Graphick Score.

Phase 2: September 2015 – May 2016

During this phase, the software artefact was presented as a working prototype at exhibitions
and some initial classroom settings. These focused on the programme itself, with participants
being asked to complete some simple task with it, and then comment upon how it might be
improved. Questionnaires were also briefly used during this period, asking participants about
their own level of musical ability, as well as their opinions of the programme. A number of
studies were conducted:
 September 2015: Exhibition at University A (referred to as the pilot study)
 November 2015: Exhibition at University B
 January 2016: Focus group at School A (Year 7)
 May 2016: Focus group at School A (Year 7)

Phase 3: June 2016 – July 2017

During phase 3, the software had reached a state of sufficient user-friendliness that it could be
used as part of a music lesson without the need to acknowledge that it was a developing
prototype. Consequently, the focus within these sessions shifted from the software itself
toward the learning activity and outcomes. The extent to which these successes were present
in the lesson became the criteria for the refinement of Graphick Score, so development
occurred alongside being fully deployed for its intended purpose. It was during this phase that
the most valuable unexpected outcomes emerged, as the users were applying their creative
ideas and intentions to the environment, and looking for ways of doing what they wanted to
do, rather than their activities being bound by the limitations of what the programme could
do. It became apparent during this phase that Graphick Score was aligning with the
definitions of a sandbox environment described in chapter 3, and that the software and
activity were undergoing a process of mutual reinforcement.

71
 During the later stages of research, I began to use touchscreen tablets in the
classroom, utilising both commercial and bespoke apps to examine the implications for
musical performance as well as composition. These studies are interconnected with the
research on Graphick Score, contributing to the same research questions, or acting as a
counterpoint to earlier research. As such, these studies are included in phase 3, and presented
in chronological order.
All the studies conducted for phase 3 all took the form of music lessons rather than
exhibitions or focus groups. Some of these were single lessons, while others were split into
two sessions, or spread over an entire school day. The specific context of each study is
indicated in chapter 7:
 June 2016: School B (Year 6)
 July 2016: School C (Year 5 and 6)
 December 2016: School B (Year 5)
 December 2016: School B (Year 5) (re-visit)
 March 2017: School D (Year 5)
 June 2017: School C (Year 3)
 July 2017: School C (Year 6) (2 days)

The research model, indicating phases, individual studies, and data collection and
analysis, is described by Figure 4.3. A diagram of codes and concepts emerging from this
research, as well as the analytical process, can be found in section 8.1.2 (Figure 8.2).

72
Figure 4.3 – The research model for data collection, analysis and application. Studies across phases 2 and
3 are indicated, along with the emerging data types.

73
4.3.2 Data Collection and Analysis

A range of data collection methods were consistently applied; audio recordings were made
during studies, and later transcribed, and memos were also often written to record
observations. A single notebook was used throughout the entire research project, to
consolidate observations on literature review, programming and testing. This allowed
emerging codes to be identified through continual consultation of notes. All work produced
by pupils during phases 2 and 3 was saved and analysed, in addition to being sent back to the
school for distribution to the pupils. This mostly consists of files made with Graphick Score,
video exports of which can be found in the electronic appendix.
In accordance with the recommendations of Glaser (1992), a process of coding for
emergence was applied at all stages. Various data types are therefore analysed using the same
process:
 Audio transcripts
 Written memos (during reading, programming and testing)
 Software artefacts (produced through programming)
 Musical artefacts (produced by the pupils using the software)
 Unstructured interviews with pupils and teachers
This led to concepts being established, which are presented through the reflective summaries
at the end of each study (see chapters 5-7). Further coding of these concepts produced wider
conceptual categories, presented at the end of each phase of research. This ultimately led to
the formation of a theory grounded in this research data (see 7.9).
In total, the studies conducted during phases 2 and 3 involved approximately 200
participants, though a wider audience has been reached through dissemination at conferences,
and casual observers taking part in the earlier exhibitions. This is within the target sample
size given in the research proposal and ethical approval form. While it should be noted that
this constitutes a small sample size in relation to other qualitative research projects, other
considerations particular to this type of research model should be taken into account. In this
project, I have acted as a solo researcher and programmer to undertake an interdisciplinary
practice-led investigation. The value of this approach has been described in this chapter, but
is also recognised by Brown (2007a, pp.7-8):

Research is often a solo venture, where the researcher undertakes all tasks from
literature review, to data collection and analysis, to presentation of findings. In

74
 educational research this is reinforced by the culture of ‘solo’ teacher. As a result, the
software engineering skills required by SoDaR may seem a prohibitive boundary to
many researchers… It is worth emphasizing the benefits of developing software
engineering skills for those so inclined. Learning to program a computer changes the
relationship between the person and the computer, such that the person can create
what they need rather than having to accept only what features the machine provides.
Also, as a mode of expression, computer languages provide both an alternative mode
of thinking and opportunity to communicate ideas. These advantages apply to
software development as they do to developing skills in other modes of expression,
such as water-color painting, calculus, rhyming couplets, or music notation.

This acknowledges the creative and expressive nature of the practical programming
component of SoDaR, and its relevance to pedagogic considerations such as communication.
Also, Brown seems to suggest here that the educator undertaking all or part of the
programming task, even from a relatively amateur position, is gaining a relevant creative and
pedagogic insight, therefore presenting an opportunity for personal and professional
development. Adopting this model has necessarily restricted the number of studies – and
subsequently participants – compared to non-practical projects, as a proportion of research
time is dedicated to the programming task, which forms a prominent part of the narrative of
this project.
It should also be noted that issues of small sample size and subsequent concerns
around generalisation are common in educational research, and many theorists have proposed
allowances for this apparent deficiency (see Bassey 1998; 2001). Where the solo teacher
conducts research focused on their practice and outcomes, the sample is inherently restricted
to the local demographic of their school and students. If the experience of such a practitioner
is of value to research communities (see 2.2.2) it must be taken into account that these
outputs have a contextual dependency upon the particular practice of that individual; this is
seen in action research, for example. Bridges (2003, p.190) argues that educational principles
must be grounded in experiences of classroom practice or this presents a generalisation in
itself:

We probably do not even fully understand our educational values until we have seen
them implemented or seen the conflicts which arise in practice between different
principles to which we ascribe in general abstract terms. We can come to understand

75
 our philosophical principles differently by seeing them realised in practice, and hence
experience can come to change the principles we hold as well as being informed by
them. We can evaluate our experience by reference to our principles: and we can re-
evaluate our principles by reference to our experience of their realisation in practice.

This seems to describe an equilibration between practice (and emerging experience) and
educational principle, an idea to which the constructivist grounded theory-informed approach
adopted for this project conforms. Where this practice is part of the research model, as in this
case, a limited sample size may be expected. However, this type of focus has its own
particular value, according to some researchers:

Generalisations derived from much educational research based on large samples may
be positively misleading, since findings derived from large scale studies are not
necessarily reflected in the much smaller numbers that teachers are concerned with
(Foreman-Peck & Murray 2008, p.145)

With a close focus on specific situations and emerging outcomes directing practice, this
qualitative model, with its smaller sample size, is akin to the complex dynamics of my chosen
programming model (see 4.1.1) and therefore appropriate to an exploratory and creative
project. Findings may be used, albeit with caution, to make wider recommendations for
further research. As Pring (2000, p.131) notes, ‘no one situation is unique in every respect
and therefore the action research in one classroom or school can illuminate or be suggestive
of practice elsewhere.’
I have combined the areas of practice in this project to define an approach which is
exploratory yet has reasonable expectations and outcomes, with respect to the work which
can be achieved by the sole researcher and the reliability of the results. This requires certain
changes in situation and setting to define certain key variables, in line with the emergent
grounded theory model proposed by Glaser (1992). Therefore, I address concepts which are
‘fuzzy’, yet of value in informing further research; Bassey (1998) argues that such
generalisations are inherently valuable if situational dependency is taken into account in the
presentation of results. His approach is summarised by Mejía (2009, p.7):

76
 1. Suppose that in a situation s1, the carrying out of action x leads to result y.
2. The fuzzy proposition is drawn that in other situations like s1it is possible that x
may lead to y.
3. Suppose that a couple of replications are carried out in chosen situations s2 and s3
and it is found that in both, x leads to y.
4. The fuzzy generalisation is drawn that in other similar situations x is likely to lead
to y.
5. Suppose that in a further replication at s4 it is found that x does not lead to y.
6. The researchers examine in detail not only what happened in s4, but go back
through s1, s2 and s3 and try to modify the description of x to find an x’ such that in
s1, s2, s3 and s4, x’ leads to y.

Similarly, Brown (2007a, p.9) recognises that this is likely to be a concern, and advises
recognition of contextual dependency (see 4.1.2). Therefore, results can be presented by
providing an account of the situation and considering the circumstances under which
identified codes emerge. Concepts, too, are situated in these contexts, but are reinforced by
emergence from multiple situations.
In order to account for any situational dependency and mitigate generalisation,
presentation of each study is divided into three sections:
1. Context: What was the purpose/background of the study?
2. Method: How was the study carried out?
3. Results: What findings emerged from the study?
A discussion of the findings is combined with the results in each case, as the coding process
is drawn from a variety of outcomes, including pupil work, transcriptions of audio and
general observations. Wider discussion on emerging concepts observed in multiple situations
occurs at intervals. These results produce a refined software artefact, but also a theoretical
outcome grounded in this practical research. Crucially, this model can be replicated by other
researchers using the same software artefact, or with comparable resources.

77
 5. Phase 1: Initial Design
January – August 2015

Figure 5.1

Phase 1 of research concerned the defining of an activity where digital technology can
provide added value and producing a prototype artefact from these considerations. This
process was grounded in the review of literature, including the National Curriculum for
Music. As such, it should be noted that this process is partly concurrent with the development
of the research context outlined in chapters 2 and 3, though I focus in this chapter on the
influence of these wider theoretical ideas in shaping my practice. An account of this
exploratory process of developing an interface in Pure Data is given, culminating in a
prototype composing environment, Graphick Score.

78
5.1 Defining the Activity

5.1.1 Overview

The first step of the SoDaR method is to identify a learning situation which can be
supported by a new software application. This should be a situation where digital technology
presents a clear added value (see 3.1). I concluded chapter 3 with a list of design objectives
for the resultant artefact, noting that it should:
 Be based on meaningful experiential activity
 Allow students to perceive, interact and engage with fundamental musical concepts
 Yield evident outcomes in terms of what is learned (in curricular terms), but also in
the production of a personally relevant creative achievement
 Be open-ended in format and encouraging exploration
 Facilitate social and collaborative learning
 Facilitate meaningful musical dialogue
 Support students in reflecting upon their musical work
 Have an accessible user interface for novice learners and non-specialist teachers
We are concerned here with creative processes, resulting in a ‘meaningful public entity’ (see
3.1.1). This should lead to outcomes which are ‘intrinsically relevant’ (see 3.3) to the
interests and motivations of participant students. Furthermore, digitisation of these processes
should result in an added value. From an experiential perspective, this involves minimising
the need for non-musical activity (explanatory verbal communication), and maximising the
amount of real musical behaviour in the lesson. Digitisation should also support the reflective
process; Students should be able to refer to the musical activity with clarity, and use this to
progress on to further musical activity. The musical concepts in question should be visible
(clearly communicated through digital representation) and tangible (able to be manipulated in
an experiential manner, in order to further understanding of these concepts).

5.1.2 Experiential Composition

In chapter 2, I discussed perspectives and examples of how improvisation can lead to musical
creativity, and suggested that inspiration may be thought of as part of an experiential process.
The curriculum tells us that pupils at key stage 1 and 2 should compose with the ‘inter-related
dimensions of music’ (see 2.2.1). However, this implies a paint-by-numbers approach to

79
composition that is not reflective of creative process. These concepts are the result of
compositional process, not the tools which are used to engineer it. They are useful musical
descriptors, providing a framework for dialogue, but as isolated abstract concepts, they have
no inherent creative value; hence, inter-related. This accounts for what research tells us about
composition in education; even among specialist music teachers, it is a source of uncertainty
(see 2.2.1).
These musical dimensions result from creative process, from intention effectively
enacted in some way. This too is a nebulous idea, but the teaching of composition must
involve creative modes of interaction where some scaffold is provided, which enables
students to see the link between their creative intention and a musical result. If
‘compositional thinking’ is to be able to conceive of ideas as musical possibilities, then we
teach this by providing an environment which supports that connection, where interaction is
more likely to result in this construction of meaning on the part of the learner (see 3.1.1). For
the novice learner, this involves improvisation with new ideas or processes which have some
recognisable significance to what they can already creatively accomplish; what creative
processes does the learner already understand, and how can this be related to a new
compositional process. Transparency is the added value that interactive digital technology
can provide in this process (see 2.2.3).
Gall and Breeze (2005, p.416) suggest that compositional process (in general, as well
as in education) is inherently shaped by the affordances of the available software
environments. They quote Trouche (2003, p.2), who suggests that ‘tools shape the
environment’, and that the learner may in turn shape these tools to meet their creative
intentions. This provides an example of
creative interaction. A tool may be used, and
a result obtained. Differentiation or
progression occurs when the tool is shaped
to achieve a different conceived possibility,
exerting control over the ‘problem
manipulation space’ (see 4.1.1). Evidence
Figure 5.2 – An example of graphic notation, where
symbols represent musical dimensions for this is provided through reflection on this
process.
The key stage 2 curriculum tells us that students should compose with alternative
notations (DfE 2013, p.2). As a teacher, I make use of graphic notation to clarify musical
dimensions (Figure 5.2). These notations, and the means of producing them or manipulating

80
them, may provide a scaffold for compositional processes. These can be the tools which the
learner shapes, making abstract learning aims such as the inter-related dimensions of music
more visually and tangible accessible (see 3.1.1; 3.1.2).

5.1.3 Expected Outcomes

The expected outcome of phase 1 was a sandbox
interface for making a graphical score, allowing
students to immediately engage with a music-
making task and to reflect upon their decisions
with reference to a visual score which clarifies
musical concepts. The activity was conceived to
build an understanding of the inter-related
dimensions of music by allowing students to
engage with these concepts experientially, to see Figure 5.3 – A suggested initial framework
for mapping musical dimensions to graphic
them as a result of their creative decisions rather notations
than being exposed to these ideas in the abstract.
Graphic notations can provide accessible representations of musical concepts, so allowing
students to access these ideas through a software interface can provide a visual, audible and
kinaesthetic mode of interaction with otherwise abstract concepts. Staff notation, it should be
noted, is itself a graphical system, albeit highly codified and, consequently, more amenable to
behaviourist teaching methods than open-ended systems. With software, these representations
have the capacity to become interactive and responsive; a user can move a representation of a
sound vertically on a screen, or across the Y axis, to hear it change in pitch (Figure 5.3). This
reinforces the language we use to describe changes in pitch; up and down. Similar
representations can be explored for other musical concepts. As such, this provides the
possibility for objects-to-think-with, from which students can synthesise their understanding
of abstract concepts through experiential learning.
The wider aims of this research directed me toward an open-ended sandbox approach
and a bottom-up design (see 4.4.1), so I reasoned that the dynamics of learning, within this
activity, would become clearer as the software was developed, applied and refined, as long as
an area in which digital technology could provide added value had been identified. This is in

81
line with the emergent methodology described in chapter 4. At this stage, it was appropriate
to describe the activity as follows:

Students will be set a composing task using the software which directs them to make a
simple musical decision (a quiet or loud sound, a high pitch or low pitch etc.) By
being guided to reflect upon the connection between the musical result and the
creative intentions, further decisions will be reached, leading to the construction of a
graphic score which communicates a story through music. This will allow students to
reflect upon how specific musical features communicate this meaning, illustrating
how the dimensions of music function in combination.
(from my notes, January 2015)

5.1.4 Domain Knowledge and Data Structures

Brown (2007a, p.6) suggests that the SoDaR

researcher, upon commencing the
development stage, ask how the software
platform supports the domain knowledge. In
software engineering, domain knowledge
refers to the specific area of knowledge that
the software aims to address. In this
instance, our domain knowledge can be
summarised as follows:
 Pedagogical representations of the
inter-related dimensions of music;
pitch, dynamics, duration, timbre,
Figure 5.4 – Data structures in Pd. Three graphical
texture, structure and tempo values (x, y and w) are used to create envelopes for the
musical dimensions of pitch and amplitude
 Creative interaction in sandbox
environments
 Experiential musical learning
The most important element of domain knowledge here is the consideration of how to present
the dimensions of music in a way that makes them accessible to novice learners within this
sandbox environment. I have discussed how graphic notation can effectively represent these

82
concepts in pedagogical contexts, so the question that follows is how best to apply this
understanding to an interactive digital format, or, in simpler terms, how to move from the
stage of defining activities to the stage of pursuing a solution through programming. In
commercial software engineering, programmers do not typically possess the domain
knowledge in question, and consequently often work with highly detailed specifications in
order to accurately address this knowledge during the development process. One of the
advantages of the approach I have adopted is that it allows for reflexive interaction between
the intended outcomes and the domain knowledge, and is therefore more suited to bottom-up
design.
The domain knowledge can be represented by data structures; ways of combining and
representing data in order to make it more accessible. In Pd, lists of data corresponding to
different parameters can be represented as geometric scalars. This allows the user to alter the
scalar directly with the mouse, which changes the parameters of the sound. In the example
(Figure 5.4), the pink and black polygons represent pitch and amplitude envelopes of the
sound event represented by the blue rectangle. The values x, y and w (the co-ordinates and
width at any point) seen in the properties window denote variables, corresponding to both the
dimensions of the polygon and the sound envelope. Interacting with the scalar, rather than the
code in the properties window, offers a much more intuitive and accessible mode of control
over these values, which may be thought of as a form of scaffolding.

Figure 5.5 – Solitude by Hans-Christoph Steiner

This approach may be used to engineer a direct link between visual and musical
representation, where data structures can be copied, pasted, dragged and resized, resulting in
corresponding changes to the musical output. An example of this is Solitude by Hans-
Christoph Steiner, an interactive piece composed in Pd which consists of a complex sequence
of envelopes which can be edited and reassigned (Figure 5.5). In this project, my aim is to
make an interactive graphic score environment that is more suitable for pedagogical
applications, and where a clear relationship exists between the data structures and the domain
knowledge (the inter-related dimensions of music).

83
5.2 Initial Development

5.2.1 Glossary

The following basic Pure Data (Pd) terms are used in this account, so they are briefly
explained here for convenience:
 Patch – A patch is a computer programme made in Pd. This may contain patches
within patches, which are known as sub-patches.
 Object – Patches are populated by objects, each of which fulfils a specific purpose.
Objects are combined to build complex processes.
 Message – Messages are sent to objects to alter their behaviour.
 Bang – A bang is a basic message which serves as a trigger.
 Argument – An argument is additional data which defines how an object behaves. It
may be sent in a message.
 Variable – A variable is a non-constant value, which may be altered via messages. In
Pd, variables are represented by dollar symbols ($).
 Abstraction – An abstraction is a type of patch which is recalled, often to be used
many times with changing variables. These may be dynamically created (see A.2.4).
 Library – A library is a set of inter-related Pd objects. The open-source development
of Pd means that many external libraries other than the basic core library are
available. The now defunct Pd-extended collated many of these libraries, while in Pd-
vanilla these must be added manually.
 GEM – GEM (graphics environment for multimedia) is an external library for
graphics processing. GEM is used extensively in the development of Graphick Score.

Graphick Score, as well as most of the alternative designs described here, consists of the
following main components:
 Canvas window – The canvas window is the window in which GEM graphics are
displayed. It is the primary means of interaction by the user. The canvas window is
most often read from left to right. The X axis may or may not be divided into a metric
grid of bars of beats, while the Y axis may or may not be divided according to a
musical scale.

84
  Geo – A geo is a geometric GEM object displayed in the canvas window. It may be a
simple shape or a more complex 2- or 3-dimensional design. The visual
characteristics of geos, such as their size and position, are linked to musical variables.
 Painting/drawing… – Interaction with the canvas window, to create or manipulate
geos, is variously termed painting, drawing, sketching, sculpting, or any other suitable
term which suggests kinaesthetic, exploratory and creative behaviour.
 Scene – The user composes in different scenes, which can be looped or sequenced.
Each scene stores geos and settings to be recalled, allowing the user to switch
between them.
 Graphics menu – The graphics menu is usually located to the right of the canvas
window. It is the means by which the user makes selections relating to graphics,
which also alters musical variables.
 Instrument menu – The instrument menu is usually located below the canvas window.
It is the means by which the user chooses sound sources, such as VST or general
MIDI instruments.

5.2.2 A Note on the Programming Process

The following is an account of the programming undertaken during phase 1. This was a
largely exploratory process, based on the principle of manipulating graphics to achieve a
musical result, with transparency between mode of interaction and outcome. Ideas were
recorded in a notebook, alongside reflections on the literature review. Patches were saved in a
repository, and archived at present state whenever significant changes occurred. This was
typically at a rate of once per week, though in later phases, with less frequent programming
changes, archiving was undertaken at a rate of approximately once per month.
Brown (2007a, pp.10-11) suggests that ‘examples of computer code will not be useful
to most educational audiences’ but that they are still ‘likely to be a significant part of
examinable thesis presentations of work done using [SoDaR].’ The most prominent changes
to programming are given in this account, while later phases contain shorter summaries of
developments alongside field research. Incidentally, only patches and examples of
programming which in some way relate to phases 2 and 3 are presented here. Consequently,
many options were explored during this phase which have ultimately not been included as
part of this narrative; only those with significance to the educational outcomes of this project
are presented. Furthermore, due to the size of the resultant programme, the processes are
85
mostly presented for a non-specialist audience, with reference to the user-friendly interface
rather than the code underpinning it. However, the entire artefact is available in the electronic
appendix, along with an explanatory file describing how the code may be viewed.
It should be noted that patches were named on an impulsive basis, as is my usual
process when programming, as this helps me to recall their specific details. The names of
these patches became the name of the interface, where used for field research. This often
involves a simple description of the process (‘sound painting’). Where possible, I avoided
any names that might be difficult to locate in search engines; hence Graphick Score is
deliberately mis-spelt. The patch was originally one of several possible variations, but
ultimately emerged from this phase as the central research outcome. All titles are working
titles only, and are subject to change after this research project, where a new name may be
chosen for any artefacts which continue to be developed.
Finally, I have stated that Pd is an open-source platform, and that the software
outcomes of this project are similarly open-source (see 4.2.1). I have maintained this standard
at all stages of programming by using only royalty-free images and sound files. Therefore, all
external images and audio files used in the construction of Graphick Score are royalty-free.

5.2.3 Sequencing in GEM

My early experimentations for composing

environments in GEM were based upon 16 beat
sequencers, with a cycle of 16 sequential bangs
(Figure 5.6). These patches consisted of a canvas
window in which different geos could be set at
different present positions by clicking (Figure 5.7).
Turning on a geo opened a gate in a sub-patch for
Figure 5.6 – A basic 16 step sequencer
each of the 16 steps, allowing the bang message of a
particular step through the gate, and triggering a corresponding sound. These were often
audio files, or sine waves with basic amplitude envelopes (Figure 5.8). Each geo ‘lit up’, by
momentarily increasing RGB values, when its position in the cycle was reached.
These patches were conventional in design and offered little in the way of originality
or creative exploration. Later versions allowed the user to switch between different canvas

86
windows to build larger structures of sequential 16 beat loops, and offered more sound design
possibilities. However, the creative possibilities at this stage remained limited.

Figure 5.8 – A sine wave

oscillator and amplitude
envelope, used as a sound
source in early patches
Figure 5.7 – The canvas window in my earliest patches

5.2.4 ‘Painting’ Sound

Wanting to move away from the
metronomic format of sequencer
interfaces and toward a more open-
ended layout, I began to explore
how ‘painting’ might be achieved
with GEM. Alpha blending allowed
Figure 5.9 – 'Painting' in the canvas window
geo ‘trails’ to be drawn in the
canvas window, by following the path of the mouse position with the left button held down,
giving the appearance of brush strokes (Figure 5.9). The width and colour of these lines could
be altered with sliders mapped to the size and RGB values of the geo (Figure 5.10). The right
mouse button was used for clearing the buffer in GEM, clearing the canvas window. Sound
was achieved by mapping the Y axis of the canvas window to pitch and the geo size to
amplitude, with the left mouse button opening a gate so the oscillator could be heard; this
meant that drawing in the window produced the corresponding changes in pitch.

87
 My goal at this stage was to
find a way for the user to ‘paint’
musical gestures in the canvas
window and play back the contents.
This became the basis for much
experimentation. I eventually
achieved an acceptable result by
writing and reading .txt files. These
files contained a number of lines (e.g.
0-1000) corresponding to the X axis
of the canvas window. As the user
moved the mouse in the window to
draw a line, a pair of values (the Y
value and the geo size, which were
mapped to MIDI pitches and velocity
values) was written to the text file at
the line corresponding to the position
on the X axis (Figure 5.11).
Figure 5.10 – Using the mouse to create geos
Overlapping simply appended the
new pair to the line, generating the possibility for
polyphonic playback. The text file could then be
‘read’ through line by line at variable speed,
outputting the pitch and velocity pairs to an
oscillator. Different text files could be written as
the RGB values changed, generating the
possibility for playback of multiple voices. I
recognised that this was a very different mode of
interaction from earlier patches, and indeed from
conventional music software.

Figure 5.11 – Text file showing X axis position,

MIDI pitch and amplitude

88
5.2.5 Scale and Register

Initially, the Y axis of the canvas window controlled pitch

by changing the frequency of an oscillator. Later, MIDI
note numbers were used instead, limiting the range of notes
to the chromatic scale. At this stage, I experimented with
different divisions of the Y axis to produce different
musical scales; this meant that gestures painted into the
canvas window played back with a more musically-
coherent result. The scale could also be altered to compare
how gestures sounded in different scales. As the most Table 5.1 – Scale degrees
common scales in Western music contain seven notes, later
versions used the order of seven colours from the familiar children’s song I Can Sing a
Rainbow: red, yellow, pink, green, purple, orange and blue (Figure 5.12).
At first, the major and minor scales for all chromatic keys were used, with a separate
control for changing the register. I later experimented with less conventional tunings, such as
the quarter-tone scale, and non-Western tuning systems such as Indonesian slendro. Having
previously taught modules on the music of other cultures at key stage 3, including Indonesian
gamelan and Indian raga, I recognised that this interface might offer a means of composing
with scales that would otherwise be difficult to access.

5.2.6 Dynamic Patching

A significant progression point during this stage was the implementation of dynamic
patching. Until this point, all objects and patches were generated manually. Dynamic
patching is an advanced form of object-oriented programming which allows abstractions to
be generated from a set of variables. This means that the patch can be extended by user
operation, or by automatic processes, generating possibilities afforded by high-level
languages, such as C# or JavaScript, that are often used in commercial software
development.
Dynamic patching in Graphick Score occurs when the user interacts with the canvas
window; when a geo is created, a list of variables is used to generate an abstraction, which
defines the geo and also the musical properties. This was at first limited to its X and Y co-
ordinates and size, but later expanded to include various other inter-related properties. These

89
abstractions are then stored in a sub-patch, each with a unique identification. Playback of a
scene triggers a locator bar to move across the canvas window. As this bar reaches the X co-
ordinate of any abstraction stored in the subpatch, it triggers the abstraction to send its
variables to the sound source (triggering the corresponding timbre, pitch, dynamics etc.) as
well as momentarily ‘lighting up’. Each abstraction also stores the number of the scene for
which it was created, allowing it to be deactivated or reactivated as required. In later
developments, these geos can be selected (using the right mouse button) and dragged to a
different position, resized, and edited in other ways. Doing so changes the associated
variables in the abstraction. Consequently, text files were no longer required to store values,
as all variables were encoded directly into abstractions (Figure 5.13).

Lists of variables are generated when the user

interacts with the canvas window (Figure 5.9)

These variables (each represented here by a

dollar symbol) define the created abstraction,
with the exception of the first two, which
simply ensure that they do not overlap when
sent to the sub-patch ‘pool’

Figure 5.13 – Dynamic creation of abstractions

5.2.7 Remote Communication

With dynamic patching meaning that every geo could be edited individually, the possibilities
for the programme increased substantially, but so did the amount of graphical processing.
Lines and curves could be drawn, consisting of multiple geo abstractions, so alpha blending
was no longer required. However, this quantity of graphical processing began to interfere
with audio processing when complex scenes were created, leading to audio crackling and
dropouts. I resolved this issue by dividing the patch into two separate menus (the graphics

90
 menu and the instrument menu)
which were loaded by separate
instances of Pd. I then set up a
remote link between the two
patches using UDP (user datagram
protocol) messaging, which allows
communication channels to be
established without prior
communication protocols (Figure
Figure 5.14 – UDP messaging between patches 5.14). This channel is established
automatically whenever the two
patches are loaded on the same computer. The addition of remote communication allowed
complex graphics or audio processes to be undertaken without interfering with one another.
The graphics and instrument menus therefore developed independently, leading to further
variations.

5.2.8 Designing Voices

Synthesizer and sound design
patches (which may be
termed VST instruments) are
commonly produced in Pd. I
intended that these could be
downloaded and used as
voices, by loading them as
abstractions. I also designed
several voice patches myself,
Figure 5.15 – VST patches made for Graphick Score
such as a drum machine and
an FM (frequency modulation) synthesizer (Figure 5.15). These were intended to allow a
much greater degree of control over timbre.

91
5.2.9 Expanding the User Interface

Developments during this time were aimed at producing a prototype interface for a pilot
study (Figure 5.16). The graphics menu allowed the user to define the scale on the Y axis, as
well as dividing the X axis into any number of steps. A list of basic shapes and RGB settings
could also be selected. A cursor was designed for the canvas window, with the note name
corresponding to the current Y co-ordinate printed onscreen. The instrument menu consisted
of 16 slots for loading VST instruments, as well as 16 MIDI outputs, allowing the interface to
be used as a control surface for external software or hardware synthesizer.

Figure 5.16 – The user interface of Graphick Score in August/September 2015

A number of additional windows and menus were temporarily added. This included a mixer
panel, which could be opened from the instrument menu (Figure 5.17). The intention for this
panel was to allow channels to be routed through effects through send and return channels,
much like a conventional mixing desk.

92
Figure 5.17 – The mixer panel

5.2.10 Alternative Models

One of the alternative
graphics menus which
reached the testing stage,
titled MetroNotes, avoided a
linear timeline altogether
(Figure 5.18). Interaction
with the canvas window
was much the same as with
Figure 5.18 – MetroNotes
Graphick Score, with the Y
axis determining pitch. However, instead of triggering geos using a locator bar, each one
‘pulsed’ at a metronomic rate, determined by its position on the X axis, which sent the pitch,
velocity and other values to the instrument menu for playback. Moving a geo from left to
right caused this pulsing rate to increase. This design meant that complex generative patterns
could be created from a small number of geos. A rhythm composer was later added, allowing
the user to programme each geo with a rhythm rather than a steady pulse.

5.3 Summary

The design and development taking place in phase 1 resulted in a graphic score based
sequencing environment, presenting a novel mode of digital music composition. Simple

93
relationships between musical and graphical dimensions were explored from the earliest
versions of this interface, based on ideas that young children may intuitively understand, or
be able to refer to verbally. For example, I felt that different colours and shapes would clearly
indicate sounds (or timbres) of different character, while those which looked similar would
indicate that the sound is similar. Velocity was represented by the size of the shape, as I
reasoned that children would recognise the relationship between a big shape and a loud noise,
or a small shape and a quiet noise. Other relationships were based on more traditional musical
ideas. For example, pitch was determined by the position on the Y axis, relating to the
common metaphor of high and low pitches. Table 5.1 shows the various musical dimensions
and corresponding representations in the earliest versions of the graphic score programme, as
well as the intended verbal descriptions (i.e. how I expected a novice learner may refer to
them).

Musical Dimension Graphical Representation Verbal Description

Pitch Y axis, or relative height High and low

Structure X axis, or time line Left to right

Timbre Shape and colour Different sounds

Texture Quantity of visible shapes More and fewer

Dynamics Size Big and small

Duration Length of shape (lines could be Long and short

drawn for longer durations)

Tempo Initially operated by a slider, but Fast and slow

changed in later versions

Table 5.2 – Basic musical dimensions represented in Graphick Score

This system presented some intuitive modes of representation, which I hoped children
could easily understand. However, some obvious limitations were also present. I was
concerned that the idea of big and small sounds (dynamics) would be confused with longer
and shorter sounds (duration). Also, though the idea of fast and slow tempo seemed intuitive,
I wanted to find a way of representing this onscreen that could be immediately understood. I

94
decided that these issues would be addressed in a pilot study, and that decisions relating to
representation and interaction would be continuously readdressed as findings emerged.

95
 6. Phase 2: Development
September 2015 – May 2016

Figure 6.1

This chapter is an account of phase 2, where the software artefact produced during the
previous phase was continuously developed and tested in response to four studies:
1. Exhibition at University A, September 2015
2. Exhibition at University B, November 2015
3. Focus Group at School A, January 2016
4. Focus Group at School A, May 2016
Research focused on the development of the user interface, and its suitability in facilitating
experiential music composition. Note that all examples of pupil compositions referenced here
can be found in the digital appendix (see 1.1.5 for guidance on locating specific examples).

96
6.1 Exhibition at University A, September 2015

6.1.1 Context

Figure 6.2 – Graphick Score in September 2015

Following the initial development of Graphick Score during the first year of research the
programme was exhibited at University A as a pilot study. At this stage, the programme had
reached a reasonably stable version of the ‘sound painting’ interface (Figure 6.2). The
purpose of this study was to gather anecdotal feedback on the user interface, with the
intention of improving layout and navigation. The registration of new students at the music
department presented an opportune audience, as the programme could be set up for use by the
students as they waited for their appointment. Participants were all 18+ years of age, with a
background in music, and many had a sound knowledge of sequencing software such as
Ableton and Logic, as well as scoring software such as Sibelius. The benefits of this
demographic were the perspectives gathered from different musical backgrounds, and the
possibility of specific expectations prior to using the programme, and clear suggestions
following. This pilot study allowed for general feedback to be gathered in a less structured
setting than the later school-based research. However, it was also clear that there were
respects in which the data obtained might be deficient within the context of the study. Firstly,
the setting would only allow for a very brief period of use by each participant, so information

97
obtained with regard to functionality would necessarily be limited. Secondly, the participants
were much older than the target age group for the programme, so their reactions and
interactions would have to be weighed up against the pertinent pedagogical considerations.
To mitigate these concerns, the pilot study focused more on the layout and navigational
menus of the programme rather than its functionality, and participants were informed of the
target age range for the programme. Ultimately, it was decided that the study would gather a
useful set of reactions from the perspective of adult music students, as well as indicating the
extent to which a musical background might suggest certain expectations with regard to
layout and navigation.

6.1.2 Method
A basic workstation of computer, keyboard, mouse and speakers was set up in a teaching
room at the music department. As students waited to be registered, they could approach and
use the programme if they wished to do so. Others observed and passed comment, but did not
use the programme themselves. I explained the purpose of the study and the target
demographic to new arrivals, but did not go into any instructions about the operation of the
programme as I wished to observe the extent to which the interface could be used intuitively,
without prior directions. No instructions were given to the participants, and the study did not
follow a consistent structure. Being an initial study, it seemed appropriate to conduct in an
open-ended manner, allowing for the emergence of possibilities which I might not have
considered. Written memos were made, and, as with all later studies, a Zoom H2n portable
recording device was set to record the discussions for later reference and coding.

6.1.3 Results

Participants generally responded well to the canvas window, being able to compose musical
ideas with little difficulty. Upon interacting with it, the relationships between pitch and
position, as well as size and volume, became apparent. Some were asked to construct a
specific musical feature, such as a chord or simple melody, and all were able to comprehend
and complete this task within the context of the programme. However, few participants used
the graphics or instrument menus at all. When asked about this, many stated that they were
confused by the layout of these sections, and felt that they should be simplified. One

98
participant said that, being dyslexic, the canvas window interface might have helped them to
engage with music-making as a young child, which they found difficult. They were less
engaged by the graphics and instrument panels, however:

‘The controls are overwhelming. There are too many of them and they all look too
similar.’

Many participants were confused by the instrument menu. The fact that each channel
consisted of three unlabelled controls proved a barrier. In this version, the numbered box was
used to select the channel, the menu to the right of this was used to select a VST instrument
to load into this channel, and the rightmost button opened the front panel of the VST
instrument so that controls could be edited. The general response was that the opening of
separate hidden windows should be avoided. For example, the shortcuts window, consisting
of a list of keyboard commands, was seen as inefficient. Several participants suggested an
onscreen key to illustrate this, or a set of keyboard ‘skins’, as can be obtained for DAW
sequencers such as Logic or Ableton. Also, the mixer panel was seen as an unnecessary
addition of this version (see Figure 5.17). Some participants liked the idea of this feature, but
noted that it was inconsistent with the aesthetic and functionality of the programme, being
typical of conventional DAW sequencers.
A significant number of the participants expressed confusion at the use of separate
controls to select geos or switch between drawing single geos and lines. That this was
something that needed to be toggled on and off was generally seen as a barrier to creative
progress. However, the option to change the number of steps in the X axis, to quantise to
these steps, and to switch between major, minor and chromatic scales in the Y axis, were all
seen as useful features. Some participants suggested a wider range of scales, and a margin to
inform the user of the scale degrees, the note indicator in the upper-left corner being not
immediately noticeable according to some. A number of participants felt that the programme
was not presented in a way that invited the user to interact with it, and that certain defaults
should be in place to guide the user upon first use. Suggestions included onscreen
demonstrations and a default setup of VST instruments and samples. It was also generally
agreed that a wider range of graphics would make the programme more visually appealing.
Some suggested alternatives to the relationships between audio and image. For example, two
participants felt that colour would be a better indicator of velocity. One felt that it would
make more sense for the structure to be read up the Y axis, with pitch along the X axis.

99
 A working prototype of the MetroNotes variation of the programme (see 5.2.10) was
also shown to some participants. In this version, geos repeatedly send pitch and velocity
information to the audio engine at a metronomic rate. This means that there is no linear
timeline or playback function; melodies are instead generated by creating different pitches at
different rates. The quantise function still works within this version, as the vertical lines
represent rates which synchronise after a certain number of cycles, the equivalent of different
beat durations within a given tempo. Participants were intrigued by this novel mode of
composition, but felt that the use of a rhythm composer for each geo complicated the
interface. The general consensus was that the simplicity and directness of this approach,
without any playback controls, was the source of appeal.
This study was a source of useful anecdotal feedback. In general, responses to the
canvas window as a mode of composition were positive, but reliance on external controls
within the outer menus and hidden panels was suggested to be a barrier to experimentation;
this applied to both Graphick Score and the MetroNotes variation. The use of keyboard
‘skins’, or stickers to identify keyboard shortcuts, might be a promising resolution to this
issue. In addition, many of the controls were suggested to be of questionable usefulness and
efficiency going forward. This relates particularly to the mixer panel of Graphick Score,
which was perhaps inconsistent with the overall aesthetic of the programme. The response
suggested that developments should focus on varied interaction with the canvas window to
replace some toggles and other controls from the graphics menu.
From this study, I concluded certain changes which could be made prior to the next
exhibition: A wider range of graphics and scales could make the programme more visually
and musically appealing. General response to the relationships between sound and image was
broadly positive at this stage, though a few participants had alternative assumptions, such as
the use of colour to indicate velocity. I felt that this latter suggestion may be informed by
familiarity with DAW sequencers, the participant having referenced these in their suggestion.
In such interfaces, the note events of only one channel are displayed at a time. In an interface
where multiple voices need to be clearly shown, I suspect that changing colour may confuse
in terms of identifying separate voices. The suggestion that the timeline should occur on the
Y axis, effectively flipping the axes, would be unusual for a composing environment. This
layout may be more appropriate and intuitive for performance-based software, where the user
benefits from a sense of the musical material scrolling towards them, such as the keyboard
performance software Synthesia or the app Piano Tiles (see 3.2.3).

100
 Finally, response to MetroNotes as an alternative mode of interaction was generally
positive, but it was clear that the versatility of the programme should be increased while
maintaining the simplicity of interaction. In other words, further external controls should be
minimised. An analogue to this design is the app Bloom (see 3.2.1), which generates looping
patterns of notes and corresponding visual circles in response to minimal interaction by the
user, i.e. touching a single portion of the screen. This requires no understanding or
preparation by the user, and is consequently suitable for any age. It is by interacting with the
application that the user explores the relationships between the visual and musical material
and their kinaesthetic role in generating this material. It can be seen that, throughout the
responses, transparency between action and musical result, and the minimisation of settings
(menu options, etc.) was a key factor in whether participants responded positively during
their brief interaction with either of the programmes.

6.2 Exhibition at University B, November 2015

6.2.1 Context
Exhibition of Graphick Score at
University B followed much the
same format as the pilot study,
though the setup was more
expansive and took place over a
longer period. At this point, the
programme was being prepared for
Figure 6.3 – The canvas window in slendro tuning use in schools, so initial technical
and aesthetic issues were still being
resolved. This public exhibition of developing technologies provided an ideal opportunity to
gather more general feedback, and to observe how the average adult fared in using the
programme. This also provided an opportunity to examine some of the suggestions that
emerged from the pilot study, and to gather opinion on other variations, such as the
MetroNotes interface.

101
Figure 6.4 – Exhibiting Graphick Score at University B.

Following the pilot study, a wider range of graphics and scales were added to the
programme. This involved simple constructed geos such as stars and hearts; 2-dimensional
shapes consisting only of a basic outline and inner colour. Added scales now included the
whole tone scale, and a 5-tone equal tempered scale as an approximation of the Indonesian
slendro (Figure 6.3). This latter was included due to the inclusion of gamelan music in the
key stage 3 music curriculum, and in consideration of the inaccessibility of any
approximation of this tuning using a subset of the chromatic scale. In place of conventional
note names, the traditional Indonesian names for the degrees of the scale were displayed in
the canvas window; ji – ro – lu – ma – nam.

6.2.2 Method
Graphick Score was set up as a workstation in the exhibition space continuously from 9am to
6pm (Figure 6.4). During this time, visitors were invited to use the programme, or to ask
questions about its development. Notes were made of any feedback or observations for later
consultation, and an audio recording was also made.

102
Figure 6.5 – Keyboard commands printed onto a mouse-mat

Several changes were made to the setup following the

response generated from the pilot study. A set of
keyboard and mouse commands were presented on a
laminated mouse-mat display, which I hoped would
minimise the need for interaction with the graphics menu
(Figure 6.5). A second poster displayed the various
proposed developments for the programme, including a
brief plan for a proposed 3-dimensional equivalent of
Figure 6.6 – The proposed design for
SoundWorlds MetroNotes. This version, called SoundWorlds, used the
Z axis to control velocity, meaning that interaction with
external menus would be reduced by the greater level of control afforded by positioning
objects in a 3-dimensional space (Figure 6.6). Responses to these ideas were gathered during
the exhibition. Finally, the MIDI output channels were routed to instrument channels in
Ableton Live, meaning that the interface could be used to access the VST instruments in this
DAW.

103
6.2.3 Results

The improved appearance and functionality of the interface led to responses of a generally
more favourable nature than at the pilot study. Participants were more eager to use the
programme without instruction. This can be attributed in part to context, the participants
having attended an exhibition specifically to use such technologies, unlike the pilot study,
which was populated by passing participants. However, the keyboard and mouse controls
were clearly a factor, as participants consulted these during their interactions with the
programme. All participants agreed that it would be more efficient to use a set of keyboard
stickers to indicate controls. Also, the inclusion of separate modes for creating and selecting
geos was seen as a barrier to interaction. The suggestion that the right mouse button could
fulfil this function without the need for a separate mode was met with a positive response.
Participants were impressed by the capacity for interacting with external software
through the MIDI outputs. One participant suggested that this would be a better way of
accessing sound sources than the built-in VST functions, as it would improve the versatility
of the programme. Additionally, this generated discussion about possible other MIDI
capabilities. The participant in question – a former music student – wanted to know if it
would be possible to write MIDI files using the programme. I agreed that this possibility
would be worth pursuing in future.
SoundWorlds was met with intrigue, with many participants asking for more details
after reading the poster. I was keen to discuss this, as I had not yet devised a means of
interface. The mouse only allows movement on a 2-dimensional plane, whereas SoundWorlds
would require movement within a third axis. A number of solutions were discussed with
several participants, including a mouse with a jogwheel, and a joystick. Following the study,
a working prototype was developed using the Leap Motion controller (Figure 6.7; 6.8).
However, other than performances and conference presentations, this option was not explored
in further study for this project due to the complexity of the interface and lack of readily
available control surfaces16.

16
At this time, Graphick Score was being developed as a desktop-only application. Touchscreen interfaces may
offer a more accessible and readily-available control surface for the 3-dimensional layout of SoundWorlds, so I
intend to revise this option during later research.

104
 The exhibition at
University B generated a more
positive response to the
programme, and also several ideas
for further development. The use of
keyboard stickers to indicate
shortcuts, and removal of the
selection mode, were clear
improvements for the next study. I
also decided to further expand the
MIDI capabilities of the
programme, and investigate the
possibility of a write-to-MIDI

Figure 6.7 – SoundWorlds uses the Leap Motion to allow the function. The benefit of this would
user to ‘grab’ and ‘move’ 3-dimensional geos, altering their
musical parameters. The palm of the hand (large mesh sphere)
be that musically untrained users
and five fingers (surrounding smaller mesh spheres) can be could create MIDI files using
seen grabbing a geo here.
Graphick Score, and load these into
DAW sequencers or score editors, acting as a form of
scaffolding for generating musical scores and MIDI
projects as well as interacting with staff notation.

Figure 6.8 – The Leap Motion uses infra-

red to map hand motions to a XYZ axis

6.3 Focus Group at School A, January 2016

6.3.1 Context

The first study to take place in a school was at School A, a state comprehensive secondary
school. The purpose of the study was to evaluate use of Graphick Score by Year 7 students,

105
 i.e. those who had only recently begun secondary school education, and in many cases formal
music education. Consent was granted from the school, and sought in advance from the
pupils and their legal guardians through a letter distributed by the school a week previously.
This also contained information about the study, contact details, possible frequently asked
questions, and assurance that data would be stored anonymously. This format was used for all
subsequent studies, and all names used are not the real names of the participants.

Following the exhibition at University B, the select mode was removed, with all
select functions accessible using the right button of the mouse. The separate mode for
drawing lines was also removed; lines could be created by holding the left mouse button and
dragging. A default set of VST instruments was also loaded into the instrument menu. In the
interest of moving towards a possible 3-dimensional interface, the menu of available geos
was now expanded to include a library of .mod files (Figure 6.9). These files, often used in
3D printing and game design, are 3-dimensional virtual models of various items. A pack of
model files available for public domain use were downloaded and incorporated as geos. A
possible future development was that these models could be turned and rotated to affect some
parameter of the sound source, though this was not yet achieved at this stage.

6.3.2 Method

Five workstations were set up in an otherwise empty

classroom of School A. These consisted of laptops running
Graphick Score, with a mouse and set of headphones. Two
sessions of 30 minutes each were planned, with five Year 7
students in each session (though the second session had only
four participants due to absence). These participants were
chosen by the Head of Music at School A, and were a mix of
boys and girls with mixed levels of musical ability. As a
Figure 6.9 – Different geos; simple focus group, I wanted a sample representative of the range
geometric shapes, complex shapes,
3-dimensional .mod files of students in the class.
When participants entered the room, the purpose
of the session – to develop a programme for music composition – was explained. A short
survey was completed by each participant, which asked whether they had any instrumental
training or other musical experience. The purpose of this survey was to gather information

106
about the musical background of each participant for comparison of results. The participants
had 20 minutes to use the programme, after which followed a brief discussion. This was done
as an informal group discussion, as I felt that pupils would be less reticent about presenting
their opinions in this setting. Finally, participants were asked to rate how easy the programme
was to use, on a scale of 1 to 10, with 1 being very difficult and 10 being very easy.17

6.3.3 Results

All participants used the canvas window with no visible difficulty, and the menu of graphical
geos was also used extensively. This also included the use of the right mouse button to select
and drag geos, which appeared to pose no great difficulty. Few other controls were used
beyond the selection of different geos. Some participants tried to change the voice in the
instrument menu, but this was done by changing the VST rather than switching channels.
Similarly, the geo and colour was mostly changed within a single channel rather than
selecting a new one. Consequently, most of the results show different geos, some of which
have a different colour, but which are almost universally built through channel 1 on the
instrument menu.
The mode of interaction observed was very similar for many of the participants;
this typically involved ‘filling’ the screen with a variety of geos, clearing it, and then moving
on to carefully drawing a line or some other pattern. In the first session, the two girls began
this process with their headphones on, while the three boys did not, only putting these on
after having drawn on the canvas window to listen to their result. In the second session, more
instructions were given as to how the programme worked, with a few functions being
demonstrated to the group. This led to different results in the second session; all participants
used the headphones, and moved to new scenes rather than clearing the screen. Also, this led
to a different line of questioning. In the first session, with no direction or demonstration, one
participant asked ‘What should we do?’ Participants were also often looking at the screen of
their neighbour, as if looking for confirmation of process. In the second session, with a
demonstration of the programme, participants instead asked how to do specific functions.
This suggested that some demonstration of the working environment encouraged
experimentation and engagement.

17
Scanned copies of these surveys can be found in the digital appendix, within the folder for this study.

107
 Results for each session were similar. Participants generally used the programme to
draw pictures and create patterns, not paying attention to the ‘sound’ of their results. Alice
and Jessica (Figure 6.10; 6.11) both had formal musical training, for clarinet and piano
respectively, but were in separate sessions. Both have filled the screen with various geos, and
recognisable patterns, such as smiling faces, can be seen. However, as Jessica attended the
second session with its more detailed demonstration, she has made more use of different
functions, such as geo type and colour.

Figure 6.10 – Alice

Figure 6.11 – Jessica

108
Other participants produced tidier results, tending to draw a contour and listen back to it.
Examples of this are Bill from the first session (Figure 6.12), and Hattie from the second
session (Figure 6.13). Neither of these participants had any instrumental training.
Discussion was very much focused on the graphical elements of the programme.
Participants wanted to see more geo types, and were particularly eager about the prospect of
an emoji-like menu, typical of social media communications. This was suggested by one of
the participants:

‘I think it should have more like the smiley face. Like you get on… like Facebook…’

No mention was made of the sound sources used, though the suggestion of a specific sound
for each geo type received a positive response. In response to the question: ‘On a scale of 1 to
10, how easy was the programme to use?’ results were mixed but tended toward the higher
end of the scale, indicating that the participants found it fairly easy to interact with the
programme.

Figure 6.12 – Bill

109
Figure 6.13 – Hattie

The results of this study indicate that participants found the interface visually easy to
use, and even engaging, but that this did not prompt a mode of use that focused on musical
output. This can be seen from the screenshots and observations on mode of use, which mostly
involved visual engagement, and the discussion, which focused on the graphical elements.
None of the participants used the XY grid within the canvas window, even when shown how
to make it visible.
In summary, evidence of engagement with the interface was improved following a
demonstration of its functions, but this was mostly focused on the graphical elements. The
priorities for development in advance of the next study concerned improving the possibilities
for creative musical exploration; the sound sources, and the relationship that exists between
interactions with the canvas window and resulting compositional material. I concluded that,
where possible, this should utilise the already observed engagement with visual and graphical
explorations (painting, drawing, making patterns out of shapes).

110
6.4 Scaffolding with Simple Structures

6.4.1 Tune-blocks

The most significant change to Graphick

Score during this phase was influenced by
Bamberger’s concept of tune-blocks, which
are used in the pedagogic composing
software Impromptu (see 3.2.3). In this
programme, which is aimed at pre-school
age children, the user arranges tune-blocks
Figure 6.14 – A tune-block for the song Oh When
into a recognisable melody. The idea of the Saints
arranging recognisable structures is a
common feature in sandbox games, such as Minecraft and Lego (see 3.3), and I also
recognised the relevance to Papert’s objects-to-think-with (see 2.1.4). Therefore, I decided to
incorporate tune-blocks into Graphick Score, to see if pupils of my target age group could
arrange these scaffolding structures on the canvas window, and what influence this might
have on creative exploration compared to the previous study.
Tune-blocks were formed by dynamically generating a pattern of multiple geos at
positions relative to the cursor, so that the user could create a block of geos starting anywhere
on the canvas window (Figure 6.14). A list of melodic fragments was accessible via the
graphics menu, which the user could ‘stamp’ onto the canvas window. Upon selection of a
tune-block, the suggested line was highlighted; i.e. if the first note of the fragment began on
the tonic, this line became bolder than the others. The user could then choose to place the
fragment on this line, or on another line, allowing more complex structures to be formed from
these fragments (Figure 6.15). Various nursery rhymes, popular songs, folk songs, and
melodies from children’s films were included in the list at different stages. Block and broken
chords were also later added (Figure 6.16).

111
Figure 6.15 – A sequence generated using the tune-block shown in Figure 6.14

Figure 6.16 – Block chords and broken chords

112
Figure 6.17 – Fruit rhythms making the rhythmic sequence for ‘pat-a-
cake pat-a-cake baker’s man’; coconut coconut apple pear

Tune-blocks (see 3.2.4) were also added as a way of making

rhythmic patterns. Fruits with names corresponding to the rhythmic
pattern were used for the menu (Figure 6.17; 6.18). Initially, these
menus were text-based, but later became populated by cycling
images (see 7.1).

Figure 6.18 – Fruit rhythms

6.4.2 Editing Tune-blocks

Following these changes to the interface, any block of geos (including drawn lines) could be
selected by dragging across a region using the right mouse button. A white rectangle is
displayed around the selected geos, forming a new tune-block. This block can be edited in a
number of ways: moved, copied, deleted, resized (altering the dynamics), sent to a new MIDI
channel (changing the instrument), and flipped or rotated. The latter function is perhaps the
most musically interesting of these possibilities; when a region of geos is selected, all the
relative positions are calculated within a sub-patch, allowing them to be repositioned along
the Y or X axis, generating horizontal and vertical mirror images. By combining this with the
copying and moving functions, a single tune-block can produce several variations (Figure
6.19). These changes generate the possibility for motivic development; a segment of ‘drawn’
music can be isolated and manipulated, producing retrograde, inversion and sequence among
other gestural compositional devices; this is a feature lacking in other compositional
environments (see 3.2.4).

113
Figure 6.19 – Rotations of the tune-block shown in Figure 6.14

6.4.3 Other Developments

Later versions of the instrument menu further simplified the layout. The VST instruments
were removed to focus on the 16 MIDI output channels. These could be assigned to any
MIDI output device, though the general MIDI instrument set was given as a default (Figure
6.20). A basic structural sequencer was added at this stage, allowing a sequence of 16 scenes
to be programmed.

Figure 6.20 – The instrument menu

Various changes made at this stage were

aimed at making the interface more user-friendly,
following commonly-observed mistakes when
using the programme. For example, when the user
tries to draw in the canvas window without an
instrument selected, a message box directs them to
Figure 6.21 – A message directs the user to
the instrument menu instead (Figure 6.21). One select a channel

other significant change in this version was the

removal of a geo selector, as this had become too much of a focal point in the previous study,
and I hoped that a selector based upon musical content would provide different results.
Instead, a simple geo was assigned to each MIDI channel, along with a default colour, though
this could still be changed. The 3-dimensional model files were not included in this version,
as I found that rendering large quantities of these objects placed great demand on the CPU,
and often led to playback issues and errors. I therefore decided to return to more advanced

114
graphics at a later date, when some more efficient application could be found. At this time, it
was clear that to use these complex graphics for single note events was an inefficient use of
memory and processing power, but that their appeal demanded that they be reintroduced for
some other purpose at a later date.

6.5 Focus Group at School A, May 2016

6.5.1 Method

Figure 6.22 – Graphick Score in May 2016

This study involved a return visit to School A, and an almost identical setup and arrangement
to the previous study. Of the nine participants, eight attended again, with two new
participants being added to fill the ten possible spaces. The same survey was again used, and
the same structure. A brief demonstration was again given, but this focused on the
construction of melodic patterns. Again, the participants understood that the purpose of the
study was to develop the programme, and that their role was to evaluate it in terms of ease of
use.

115
6.5.2 Results

Figure 6.23 – Lara

Figure 6.24 – Lee

116
Figure 6.25 – Hattie

Figure 6.26 – Jessica

It was clear from observation that the participants were using the programme in what might
be described as a more linear fashion, with material conforming to the grid. For example,
Lara, who is musically trained and able to read standard notation, arranged rhythms in a
vertical and horizontal sequence (Figure 6.23). A similar pattern is seen in the work of Lee,
also musically trained, who followed a vertical line with a melodic contour by placing geos
on the grid (Figure 6.24). A sequence was also observed in the work of musically untrained
participants, but this tended to conform less rigidly to the grid. For example, Hattie follows a

117
repeating diagonal pattern across the canvas window (Figure 6.25). These participants were
observed listening to the results of their sketches as they made them. However, other
participants showed fewer signs of listening to the musical results. Jessica, who is musically
trained, filled the canvas window with geos in much the same manner as in the previous
study (Figure 6.26). However, Jessica also made use of the copying function to build her
patterns, which is more evident in the less busy scenes (Figure 6.27). As with the previous
study, few participants changed the instrument. No participants made use of the function to
select and rotate groups of geos.

Figure 6.27 - Jessica

Discussion focused more on the construction of musical material and less on graphical
material. As one participant who attended both studies stated:

‘It was easier to make something in time… to make it sound in time.’

This was a function of the grid, which had been ignored in the previous study. However, it is
notable that this was used more readily by trained musicians than those who had no
instrumental experience. When it was pointed out that few participants changed the menu of
phrases or instruments, it was generally agreed that these components did not look inviting.
Many of the untrained participants did not know what to expect from the language used,
especially with regard to instrument type, and menu options such as ‘chords’:

118
 Me: ‘When picking [a shape] did you know what it would be?’
Participant: ‘No, I just chose it.’ [With agreement from others]

In response to the question: ‘On a scale of 1 to 10, how easy was the programme to use?’
results were more consistent than in the previous study, being mostly scored at 7 or 8.
A much clearer impression of how the programme could be used for composition was
arguably gained from this study in comparison to the previous one. This is shown by the
more sequential nature of the content, the more musical focus of the discussion, and the
consistency of the response. There was a general consensus that the use of tune-block
structures and the canvas window grid made the programme easier to use, though some
participants felt that it was less aesthetically inviting than the previous version, due to the
temporary abandonment of complex graphics such as .mod files. This accounts for the
narrowing of results in response to the survey question. It is notable that, again, mode of
interaction was mostly confined to the canvas window. To further examine the usefulness of
the other functions, and therefore make the programme more versatile, I decided that I would
have to redesign the graphics and instrument menu, and to re-evaluate some of the language
used.

6.6 Summary

Phase 2 principally focused on the development of the user interface. The development of a
sandbox interface, with intuitive and improvised composition leading to meaningful musical
patterns, was the goal of this phase of research. Comparison of the concepts raised in the
discussion of each study led to the emergence of certain principal categories.
The first category is visual-oriented exploration, observed in the ‘sound-painting’
seen in the earliest studies. It can be seen from the first two studies that a concise interface,
with minimal reliance on the selection of options, was preferred. The two later studies
revealed a mode of exploration which tended toward the visual; participants seemed to focus
more on how the score looked than how it sounded, and the graphical options were related to
the emoji images used in social media platforms. Exploration of the programme was mostly
confined to the canvas window, i.e. the component which produces an instant visual result,

119
rather than exploring different menu options. However, the examples demonstrate a range of
different explorations, including patterns which form coherent musical gestures.
In producing a sandbox composing environment for experiential exploration of the
dimensions of music, one of the considerations for this project was to minimise the need for
verbal explanation prior to creative tasks (see 4.1.1). In the third study (see 6.3), more
engagement with Graphick Score was observed in the second group, who were given a brief
demonstration of the programme. This is likely due to the still relatively unclear layout of the
interface at this point. It is notable that the new participants in the fourth study (see 6.5) were
not confused about the mode of operation itself, though here a musical learning outcome of
sorts had been indicated in the demonstration of constructing a melody. This study showed
more evidence of listening by the participants, but was still predominantly based on the
formation of visual patterns.
Secondly, scaffolding with simple structures emerged as a further mode of interaction.
Bamberger (1996) argues that, from a pedagogical perspective, a fragment of melody, or
tune-block, is simpler than a single note in isolation, as it is a more identifiable structure, with
a greater potential for understanding. Similarly, Papert (1980, p.11) describes objects-to-
think-with as having the ‘capacity for personal identification’. In the third study (see 6.3),
such identification is apparent in the participants who recognise emoji-like geos, noting their
absence in the following study. I later incorporated Bamberger’s tune-blocks into Graphick
Score as a way of pursuing musical identification. The results from the subsequent study
indicate more musically-coherent patterns and sequences, arranged on the now-present metric
grid, as well as greater evidence of listening to these results. However, it is also observed that
the desired capacity for identification was not yet present here, as participants were still
making many choices at random, and did not know what to expect from the tune-block
options. This suggests that, while these structures can offer potential for scaffolding musical
behaviours, we need to further embed meaning-making structures which can support creative
choices.
While these form two distinct conceptual categories emerging from this early stage of
research, they are also clearly interconnected: If users of Graphick Score tend toward a more
direct and visual-oriented mode of interaction and exploration, scaffolding structures can
support this by allowing for more meaningful patterns and structures to arise from simple and
limited actions. However, by situating the stage of development outside of the classroom –
perhaps for too long – many features were explored which were admittedly unsuited to the
design objective, such as the mixer panel, and the use of VST instruments and effects. Also,

120
the lack of a clear learning objective acted as a barrier to exploration in the classroom-based
studies, as participants required a demonstration in order to embark upon a task. This was not
the case when used by adults during the exhibitions.
These outcomes suggest that the mode of use should be explored alongside refinement
of the current features in phase 3. The objective of experiential composition with the
dimensions of music appears to be reinforced by the canvas window to a point, in that
participants have been engaged by interacting with it. In chapter 3, I discussed how
interactive digital technology may clarify learning materials by making them visible and
tangible to the learner (see 3.3.1; 3.3.2). In the earliest versions of Graphick Score, the
interface offered basic interaction with pitch, dynamics and structure, but other dimensions,
such as timbre and tonality, were inaccessible to the novice user. The addition of tune-blocks,
and the means of altering them through repositioning and rotating, has enhanced these
possibilities by presenting a more complex interplay of, and interaction with, musical
dimensions than a single note can provide; as Bamberger (1996, p.34) notes, this increased
complexity generates a scaffold with which is simpler for the novice learner to engage:

We are asking students to begin with what we believe are the simplest kinds of
elements, but which for them may be the most difficult. In doing so, I think we are
confusing smallest elements – in music, isolated, de-contextualized pitch and duration
values – with what we assume are also the simplest elements.

Being aimed at the very young, Impromptu focused on sequential arrangement of tune-blocks
(see 3.2.3). Graphick Score attempts to harness this innovation for a school-age demographic
by offering further arrangement possibilities, facilitating more complex compositions.
However, this approach requires a clear layout and clear objective, something which should
be further refined.
Development up to this point raised some unexpected but interesting educational
possibilities. MetroNotes offers a format for generative music composition, which would
present an alternative to the timeline-based interfaces usually found in composition software.
Composition with non-Western scales has also been explored using Graphick Score, which
presents an opportunity to investigate the tunings of other cultures in the classroom. As
stated, the addition of tune-block structures presents an opportunity for enhanced engagement
with musical dimensions.

121
 Glaser (1992, p.76) argues that
grounded theory involves the cultivation of a
core category, which involves favourable
selection of the emergences which appear
most significant. As my research is practice-
led, these significant emergences should be
Figure 6.28 – Arranging a block of ‘sound
painting’ the focus of my practice going forward. The
addition of tune-blocks, and mode of
interacting with them, enhances the possibilities for the ‘sound painting’ interactions
observed in earlier studies. For example, improvised gestures may be arranged and edited
into more complex formations (Figure 6.28). Such possibilities, as well as other expected
outcomes, have the potential to emerge in the next research phase, as the classroom
application of Graphick Score comes under closer examination. However, some of the
features earlier described do not fit in with this design strategy: These modes of interaction
are timeline-based, and unsuited to the non-linear interfaces of MetroNotes and SoundWorlds.
Furthermore, implementation of tune-blocks meant the temporary removal of alternative
scales, as the melodic options are all based on recognisable popular tunes from Western
culture:

‘Through listening to music of our own culture, we have become most responsive to
structural functions such as stability and instability – whether a phrase sounds ended
or is still going on…’
(Bamberger 1996, p.42)

Consequently, an expansion of educational and creative possibilities led to a limitation, or at

least a restriction, as certain promising yet incompatible features of the software are not
investigated in the next phase of research.
Phase 3 is characterised by classroom research with objectives typical of music lessons,
with Graphick Score being applied and examined in the intended situational contexts. As the
focus is on learning objective, sensitivity to emergent possibilities which may extend beyond
that which can currently be addressed by Graphick Score is required. In particular, this
concerns the application of touchscreen tablets, which may present unique implications as
digital technologies in music classrooms (see 3.2.1). Though it is my intention to develop
Graphick Score as an app in future, this may not be addressed within the duration of this

122
project. Therefore, I resolved to remain open to opportunities to apply these technologies to
address the core activity of experiential composition in sandbox environments where
possible. My intentions for the following phase can be summarised as follows:
 Use and refine Graphick Score in the context of music lessons.
 Ensure the presence of clear learning objectives and outcomes.
 Enhance the possibilities for scaffolding structures in facilitating creative
composition.
 Work toward the construction of an intuitive, concise and accessible layout.
 Be continually sensitive to emergent possibilities which may support the existing
research aims, conducting studies with alternative bespoke and existing resources
where necessary.

123
 7. Phase 3: Application, Refinement, Expansion
June 2016 – July 2017

Figure 7.1

This chapter is an account of phase 3, where the software artefact is tested and refined in
music lesson contexts. Additionally, new emerging ideas are examined through the
application of other technologies, such as bespoke and commercial touchscreen tablet apps.
This phase involved seven studies with contrasting situational contexts:
1. School B, June 2016
2. School C, July 2016
3. School B, December 2016
4. School B, December 2016
5. School D, March 2017
6. School C, June 2017
7. School C, July 2017 (2 days)
The same methods of data collection and analysis were applied for all studies, and a
theoretical summary of the creative learning opportunities and barriers observed concludes
this chapter. Note that all examples of pupil compositions referenced here can be found in the
digital appendix (see 1.1.5 for guidance on locating specific examples).

124
7.1 Refinement of the Interface

Figure 7.2 – Graphick Score in June 2016

The start of phase 3 saw the programme

develop to a more user-friendly and
classroom appropriate state (Figure 7.2).
These changes were informed by the
research conducted in phase 2, and aimed
at developing an interface which
promoted exploration and experimentation
through concise and, where possible, self-
explanatory controls. Most of the menu
options are replaced by image buttons
which, when clicked, cycled through a
sequence of contrasting images relating to
a musical dimension. For example, major

Figure 7.3 – The tune-block for Somewhere over the

and minor tonality is represented by
Rainbow is changed from a major to a minor key. The sunshine and raincloud icons, displayed
resulting flattened note is circled.
on the graphics menu and canvas window
(Figure 7.3). Similarly, the previously used tempo slider was replaced by a sequence of
animal images representing different playback speeds (Figure 7.5). In this way, simple

125
cultural analogues were used in place of musical descriptors. Buttons displaying graphics
were used for other functions: The fruit-rhythms button, when pressed, cycled through
images of fruit, the list of tune-blocks was shown when the user clicked the rainbow button,
and the draw function could be activated by clicking the pencil. All of these functions could
be accessed from the keyboard, on which corresponding stickers were displayed (Figure 7.4).

Figure 7.4 – Keyboard stickers for operating the programme

The instrument menu also was also updated for a

more user-friendly layout. An image of the selected
instrument was displayed, and a brief flourish on the
Figure 7.5 – Animal tempos, chosen instrument and key was heard, providing a visual
moving from slow to fast; tortoise,
man, hare, cheetah and audible acknowledgment of the selection. Default
colours and shapes are also applied to each instrument
channel, with the corresponding colour shown behind the instrument name for quick
reference; this colour changes if altered by the user (Figure 7.6). An image of the selected
instrument was displayed in the instrument menu, though to make room for this, the structure
panel had to be moved to a hidden window. This temporary solution was undesirable, as
previous research had indicated that such options went unexplored, but I reasoned that this
part of the interface was not as well developed as other components, and not necessary for
immediate use, as structure could be directly altered by changing scene during playback.

Figure 7.6 – The instrument menu

126
7.2 Storyboard Composition at School B, June 2016

7.2.1 Context

Up to this point, studies had focused on the design of the programme, rather than the outputs
of the participants using the programme. Therefore, I decided to structure subsequent studies
more like a music lesson, and to focus on a more specific age demographic. I opted for key
stage 2, as the National Curriculum states two aims at this level which I felt Graphick Score
could practically address:

 ‘improvise and compose music for a range of purposes using the inter-related
dimensions of music’
 ‘use and understand staff and other musical notations’
(DfE 2013, p.2)

I held a meeting with the head of Year 6 at School B, a primary school. We agreed that I
could visit the school for a full day, from 9am to 3pm, holding sessions with five pupils at a
time, and hopefully working through the entire Year 6 class. With the assistance of my
contact at the school, I devised a lesson plan to fit in with their current scheme of work. The
class had been writing storyboards for a forthcoming school play, so we agreed that it would
be appropriate to ask the participants to create a storyboard using Graphick Score. This
would be the first time that the programme would be used for a functional purpose rather than
being objectively evaluated, and the proposed activity seemed promising in that there would
be much possibility of creative expression. I hoped that this storyboard format would
encourage the participants to imagine representations for the musical and graphical material,
and that this would reinforce links between these two components, perhaps lessening the
focus on graphical content alone observed earlier studies.

7.2.2 Method

Learning Objective: To compose a short story using sound

I was allocated a small classroom within which to set up five workstations as with prior
studies, with headphones and computer mice. This room, it was explained to me, was

127
 frequently used for alternative activities, so pupils
were used to being taken there in groups during class
time. Also, due to the artistic nature of the main
activity from which the pupils were being drawn –
rehearsing a play – I was confident that the pupils
would not feel a sense of disruption to the
educational setting. This would further support the
planned lesson structure. In total, 26 pupils used the
programme over the course of the day. One
participant, whose computer had unfortunately
frozen prior to showing their work, asked if they
could attend again for the final session, and was
allowed to do so. The work was later recovered.
When each group entered the room, I
engaged them in a discussion about what they had
done during the half-term break. This was to put
them at ease within the setting, and to quickly
establish a dynamic. It also served the purpose of

Figure 7.7 – The example story, Incy providing a frame of reference for the forthcoming
Wincy Spider, told with ‘sound painting’
storyboard activity, as their anecdotes could be used
as examples within the activity. I then posed a question to the group:

‘Words can be used to tell a story, and so can pictures. But can sounds be used to tell
a story?’

I then asked for examples, making sure that the difference between ‘sounds like’ and
‘represents’ were clarified. To illustrate this, I used Graphick Score to show how a xylophone
might sound like rain, and a sad melody might represent rain. As a demonstration, I then built
a storyboard using different scenes in Graphick Score, asking the group if they could guess
the story (Figure 7.7). In every session, this was eventually identified as Incy Wincy Spider,
though some prompting was occasionally required. I then asked the group to build their own
story with sounds.
I delivered these sessions alone, providing support to the students where requested.
Each session lasted for approximately 50 minutes in total. At the mid-point of the session, the

128
group were asked to play an excerpt of what they had composed, and discuss how this related
to their story. At the end of the session, this was repeated with the entire story. Data was
collected by observation and written memos, and an audio recording was made, which was
later transcribed. The pieces made by the participants were saved and later sent to the class
teacher as screen recordings.

7.2.3 Results

The sessions resulted in a very high level of engagement throughout the day. All participants
completed the task, and none expressed any boredom, confusion or dissatisfaction. A much
wider range of functions were used; Participants tried the different tune-blocks and rhythms,
altered the speed and tonality of their scenes, and listened to the various MIDI instruments.
Selecting, moving and rotating groups of geos were also commonly used functions. A range
of exploratory questions were asked, such as:

‘Can you have that bit fast and that bit slow?’ [Referring to sections within a single
scene]

This indicated that participants were imagining functions that they wished to apply to their
music, and that some creative impulse was being engaged.

Figure 7.8 – Carly

129
Figure 7.9 – Imogen

Figure 7.10 – Hayley

130
Figure 7.11 – Kody

Some participants mimicked the up-and-down gesture used in the demonstration of

Incy Wincy Spider, but reworked it to tell another story. In some instances, this was done by
drawing using the pencil tool (Figure 7.8; 7.9; 7.10) while others used metric rhythm blocks
to create a sequence (Figure 7.11). Both of these modes of interaction were seen in prior
studies (see 6.5.3). Pupils described their ideas of what this represented during the discussion:

Imogen: ‘… a rabbit being chased by something…’

Other Pupil: ‘You could have the piano for running fast, going down, then like, one
drum.’ [Makes action of something going down a hole with arm]

Here, the other pupil provides some instrumental suggestions to build upon this gestural idea,
a rapid piano ostinato and the single bang of a drum as the rabbit jumps down its hole. In the
later discussion as Imogen built upon this idea, it became apparent that the tempo animal
images were being used as a creative stimulus:

Imogen: ‘I did a rabbit trying to run up a hill… It was a rabbit trying to get up a hill,
so he got up but then it rained. So he had to go back down but then there was a
cheetah at the bottom because I saw that. [Points to tempo icon.] Then he had to run
back up, and he got away from the cheetah. Then he finds his family so he’s happy.’

131
Karla takes the nursery rhyme used in the demonstration as a starting point, before moving on
to variations on a fairy tale. Here, the discussion on musical features during the midpoint of
the session appears again to have led to a broader experimentation of musical gestures and
connected narrative, which appears loosely based on the indicated story:

Karla: ‘So I did Incy Wincy, but then for the last four scenes I did Little Red Riding
Hood.’
‘That’s where she’s walking along’ [linear drawing across screen]
‘She thinks someone’s following her’ [suspenseful, quiet tones]
’This is where he’s chasing her’ [busy, loud noises]
‘This is where he’s falling’ [crashing noises, drums]
‘This is where he’s upset’ [sad sounding music]
‘And then he tries to get up again’ [tinkling sounds, glockenspiel]
‘And then he’s celebrating’ [synthesizer, fanfare-like]

Another recurring feature was the way in which participants engaged with each other,
showing their ideas and asking questions:

Kathy: ‘So, what happens is… Have you ever watched Peppa Pig?’
Me: ‘I have. My niece really likes it’
Kathy: ‘Right, so basically, Peppa and George… it’s dead sunny outside so they want
to go and see what’s happening, and see if they can go and play outside. But then it
starts raining. Then it goes sunny, and they get to jump in muddy puddles so they’re
happy.’
‘so this is the happy bit where they think it’s sunny’
‘then it starts raining’
Other pupils: ‘what instrument is that?’ ‘that sounds good! play it again’
Kathy: ‘Then this is supposed to be the bit where they go outside.’

Here, as with other examples, the images of sunshine and rain, with the text ‘happy’ and ‘sad’
in the graphics menu, provide a wider narrative framework within which the pupils sketch out
details. Drawn gestures are again used for actions (jumping in puddles), with melodic ideas
made from tune-blocks, resulting with a major or minor tonality as required. Other examples
made more prominent use of tune-blocks to indicate motion or action:

132
 Anna: ‘Mine’s about like… this is waiting in a queue’
‘it went… (doo doo doo doo)’ [imitates it with up and down finger movement, like a
rollercoaster]
‘mine was… it was in Gulliver’s World. and we’re waiting on a queue, then going on a
ride, then waiting in a queue…’
Me: ‘so how did you represent the queue and the ride?’
Anna: ‘I went like… I put like four notes, like, 1 2 3 4, with’
Other pupil: ‘… with like strawberry?’
Anna: ‘yeah… and then with like pear… I did 1234, 1234’
Me: ‘right, okay’
Anna: ‘and then I did the same… and I did, erm, apple, and I did the upside down ride,
the hoop thing, where you go up then you spin’

This use of the language of the programme – fruits, animals etc. – to refer to musical features
present in the compositions was interesting, as it suggested that a frame of reference had been
established which enabled the participants to describe their ideas. Participants very quickly
built confidence in this setting, wanting to show their compositions, and being keen to show
others how they had utilised some function. This was especially prominent with the returning
participant, Anna, who was eager to show her existing knowledge of the programme. At one
point, when one participant asked for help as I was busy with another, Anna took off her
headphones and said ‘I’ll show you how.’ Another participant reconstructed melodies from
the tune fragments and, when invited to play their work back, had devised a game:

‘You have to guess the songs.’

The idea that participants might use the programme to build games and other activities was
unexpected, but has parallels with the problem-creation applications of sandbox games such
as Minecraft (see 3.3.2).

133
Figure 7.12 – Rob

134
Figure 7.13 – Renee

Figure 7.14 – Barry

Another interesting feature was the use of melodic contour and metronomic, repeating
sequences, which occurred much more frequently than in the previous studies. For example,
Rob (Figure 7.12) makes use of copied rising and falling patterns to achieve a polyphonic and
polyrhythmic complexity. Sometimes, these gestures emerged from simple representations of
the sun and rain images on the canvas window. Renee (Figure 7.13) uses the grid lines to
produce a metric series of blue ‘rain drops’, while Barry (Figure 7.14) represents the sun
through wavering orange trills. That these gestures appeared to emerge by discovery,

135
abstracted from the simple graphics, was an unexpected occurence. This often led to learning
opportunities which could be categorised as discovery learning. For example, when one
participant (Leighton) stated:

‘Mine sounds like Japan.’

It transpired that the participant had used a sequence of slow rhythms in a pentatonic scale,
and that this was a good imitation of Japanese music, the pentatonic scale being often used to
teach the music of Japan at key stage 3. The sound used, a guitar, was reminiscent of a koto
when heard in this manner. For other pupils, experimentation with basic musical dimensions
within the grid of the canvas window produced results of surprising regularity when played
back. An example of this is Emma (Figure 7.15) who uses the grid to produce blocks of
sound; intitially stacked thirds followed by three single note crotchets, gradually increasing in
texture and dynamics, until the notes ‘scatter’ about. This is an example of where the drawing
of visual patterns – what I have come to term ‘sound-painting’ – allows pupils to engage with
changing musical dimensions, and produce results with a clear musical coherence despite the
rudimentary mode of interaction.

Figure 7.15 – Emma

In other examples, sound-painting was used to characterise actions and events from a
simple story. Take, for example, Aisha’s Owl composition (Figure 7.16):

Aisha: ‘I’ve done… this is like an owl… then it flies over, and it drops, then that’s it in
its nest’

Though straightforward, the images tell a clear story in which there is a present musical
element. The owl, flying through the air, lands with a drop in pitch. Its sedentary position in

136
the next scene is represented by the circular ‘nest’ at one end of the window, as it waits out
the rain. Structure and timbre has been used in a simple but effective manner to tell the story.
A similar approach to ‘drawing’ movement is seen in Millie’s Bird (Figure 7.17), and
Ronan’s Humpty Dumpty, which depicts the motion of the egg falling off the wall with a
dropping pitch, and the visual and audible mess as it hits the ground (Figure 7.18). These
motions have been noticeably copied and pasted, effectively generating motivic gesture. A
clearer example of this is found in Harvey’s Ghost (Figure 7.19). Here, the visual and audible
characteristics were pointed out by other pupils; the single synthesizer timbre described as
sounding ‘sort of spooky’, and the vaguely ghostly-looking geo which by default
accompanied it. Harvey took these ideas to build several scenes of changing texture, from
tense sparse notes to ‘shock’ bursts of sound, such as the copied glissando (shown in the
figure) which is repeated at different intervals. Again here, these painted ideas become
compositional gestures which the pupils use to tell their stories.

Figure 7.16 – Aisha, Owl

Figure 7.17 – Millie, Bird

137
Figure 7.18 – Ronan, Humpty Dumpty

Figure 7.19 – Harvey, Ghost

This study proved to be much more successful than any previous attempts. The high
levels of engagement and improved mode of interaction point to the effectiveness of the new
user interface, as evidence by the participant’s willingness to assist one another in working
out functions, and in their referral to the onscreen icons when discussing their work. Also, I
believe that the educational setting was a powerful factor in influencing a creative output.
Pupils felt that there was a challenge and purpose for the activity, as well as an underlying
social influence to tell a story. This demonstrates the importance of considering not just what
technology is used, but how it is used to support a wider learning environment (see Fullan &
Langworthy 2014, p.30).
Evidence of creativity can be seen in the use of melodic and sequential material, and in
the impressionistic depictions of actions, characters and events. The pupils were quick to
attribute meaning to their creative explorations, drawing from the initial discussions on

138
holiday activities, the demonstrated example, and other ideas emerging from their ‘sound
painting’ in Graphick Score. This sometimes emerged from the presence of images linked to
musical dimensions, such as the use of sun or rain to accompany happy or sad scenes, or the
use of the animal tempo images. It is interesting that certain contexts are so quickly adopted
and altered. The demonstration of Incy Wincy Spider was chosen as a simple four-stage story
that anyone could identify, to model a possible outcome of using the programme. The up-
and-down motion and sunshine/rain polarity was then appropriated by some pupils for their
own purposes. Others took the idea of a nursery rhyme as a starting point, leading to
representations of Red Riding Hood and Humpty Dumpty; it was unexpected that the
demonstrated story would lead to this kind of influence, being aimed at a much younger
demographic. In some cases, it is perhaps unclear how the story, as told by the student, relates
to the musical gestures employed. Interestingly, those with perhaps the most clarity are
‘drawn’ gestures without the melodic coherence of the tune-blocks, but still presenting a
collation of musical dimensions to represent an action or event. Within the scaffolding format
of Graphick Score, this ‘painting’ or ‘drawing’ can lead to results of surprising musical
coherence, which is apparently then shaped to give some meaning (e.g. this is about a bird/an
owl/a ghost etc.) This is evidence of compositional thinking at a very fundamental level;
indeed, inspiration appears to be present here (see 2.2.1) as the pupils ‘shape the tools’ of the
environment (see 5.1.2) to generate meaning.
Many examples suggestive of the target activity of experiential composition have been
afforded by this digital resource. Given an open-ended objective to tell a story through sound,
pupils used the tools offered by Graphick Score for creative explorations, in many cases
leading to opportunities of discovery learning, and the development of expressive actions
which were not demonstrated; this includes the copying and arrangement of motivic gestures,
and the use of tune-blocks to develop harmonic and rhythmic sequences. In accordance with
the recommendations of Bassey (1998) I now decided to apply the same resource to a
different situational context in a new school environment.

139
7.3 Melodic Composition at School C, July 2016

7.3.1 Context

Following the storyboard format of the previous study, I wanted to further explore the
capacity for more traditional compositional approaches of melodic and harmonic writing
presented by Graphick Score. The study at School C focused on a more conventional
composing activity with a group of pupils who had prior musical experience. I arranged with
my contact at the school to conduct an afternoon session with pupils from a Year 5 class.
These were all instrumentalists, mostly consisting of members of the school brass band, and
were therefore used to reading and performing music, but had less experience with
composition. School C has a good reputation for musical practice, but my contact wanted to
implement more interactive technology, especially for composing activities, and was eager to
see how the pupils responded to new interfaces like Graphick Score. Therefore, seeing how
these ten young instrumentalists approached composition with a digital scaffolding tool was
very much a suggestion of the class teacher, which I agreed would make a valuable
situational contrast to the previous study. In some respects the two studies were similar, in
that the same resource was used, the pupils were roughly the same age (within one year), and
the same time frame and activity for the session. However, the pupils had more musical
experience than the previous group, and objective was different, focusing on harmonic and
melodic writing.

7.3.2 Method

Learning Objective: To compose a piece with varying rhythmic and melodic patterns

The study involved two successive groups of five pupils taken from years 5 and 6. As with
previous studies, five workstations were set up in a classroom. After engaging with each
group in an informal discussion about their musicality, a demonstration of the programme
was given. This involved comparison of a messy scene and a tidy scene, with the group being
asked ‘which sounds better, and why?’ (Figure 7.20). The group then had 15 minutes to build
some ideas and explore the programme, following which they would share with their
neighbours, and then another 20 minutes to complete their pieces. After this, we listened to
and discussed the pieces as a group.

140
Figure 7.20 – The demonstration from the session

7.3.3 Results
The pupils all decided that the first example from the demonstration sounded the better of the
two, being more ‘in time’ due to the notes conforming to the grid. After some prompting, the
idea of repeating sequences being ‘catchy’ was also established, and also the spacing of notes
to construct harmonies or chords. This latter concept was less readily understood. Although a
brief discussion, this was useful in that it gave me a sense of the skill level of the group,
perhaps more efficiently than the surveys used in phase 2.
This study yielded some interesting results in terms of music composed, but I felt that
engagement and depth of discussion did not match that of the previous study. I attribute this
to the learning materials not capturing the pupils’ imaginations, as they were composing
without the stimulus to tell a story. I did ask the groups to think of a title for their pieces, and
to discuss a possible title with their neighbour after the first 15 minutes, as I hoped that this
would provide an insight into compositional intentions. However, no participants had a title
in mind when presenting their work.

141
Figure 7.21 – Natasha

Figure 7.22 – Ethan

Figure 7.23 – Rickie

142
 Many of the pieces showed use of different phrases, rhythms, and voices, arranged in a
metronomic sequence. Several scenes composed by Natasha contain a bass ostinato and
repeating cross-rhythms (Figure 7.21). Similarly, Ethan (Figure 7.22) makes use of a
repeating motif composed of several voices. This use of tune-blocks for regular formation
upon the grid appears to have been very quickly ascertained from the demonstration by these
pupils. The example by Rickie (Figure 7.23) contains several conjunct melodic patterns
which are partially harmonised and imitated by other voices. Notice how the contour of the
blue triangle melody is harmonised a third above by another voice; this appears to quite
closely mimic the harmonic pattern used in the demonstration (Figure 7.20). The first
harmonic sequence in Rickie’s scene here appears to mimic the harmonic pattern of the
demonstration, but places the second voice a tone above, before build a harmony of thirds in
the second half of the scene, suggesting that the demonstration acted as something of a visual
scaffold which allowed her to attempt this kind of harmonic writing. This resonates with the
visual-oriented explorations seen in earlier studies (see 6.6) as well as the mimicking of the
visual up-and-down gesture demonstrated in the previous study (see 7.2.3).
The harmonic and melodic constructions from tune-blocks seen here are similar to
those which emerged from the previous study. Again here, the pupils appear to have quickly
established musical gestures, and developed them into complex textures. Sammy’s piece has
particularly interesting examples of polyphonic texture and musical gesture (Figure 7.24). It
contains a number of repeating rhythms and melodies spaced apart to form a consonant
harmony, and decorated by harp glissandos and flourishes as well as trills; as such, several
distinct voices combine to produce a musically and visually coherent piece. The work of
Lewis (Figure 7.25) is, by contrast, mainly constructed of percussion instruments and
rhythm-blocks, as well as blocks of sound less characteristic of diatonic harmony-oriented
writing. It transpired that Lewis was the drummer of the group, and it appears that he drew
from his existing experience and knowledge in approaching this piece; this is, of course, an
idea very much present in constructivist-oriented instruction (see 2.1).

143
Figure 7.24 – Sammy

144
Figure 7.25 – Lewis

In this study, the pupils’ use of Graphick Score appears to have again drawn heavily
from the demonstration. This resulted in instances of harmonic and rhythmic sequences, as
the objective required. The pupils were mostly able to compose with ease using the
programme, with some evidence of gesture and ornamentation, suggesting that Graphick
Score offered an accessible means of applying existing musical skills, and perhaps a scaffold
for more complex compositional writing. However, relatively little attention was given to
musical dimensions such as dynamics, with most geos at a uniform velocity. The earliest
studies (see chapter 6) suggested that menu options were rarely used for functions such as
tempo and velocity, as well as other variable settings. Options such as change of instrument
were more readily utilised. The suggestions from pupils of different instrument timbres in the
previous study (see 7.2.3) indicates that this is something they were actively considering
when composing; part of their creative intention. One of the musical dimensions that become
more frequently utilised was tempo. This was previously operated via a slider on the graphics
menu; when it changed to a series of animal images, it was not only more frequently utilised,
but also influenced discussion and narrative. Perhaps the mode of operation for dynamics
should be made more inviting to the user through some sort of conceptual representation, or
altered somehow via direct interaction with the canvas window.
There is, perhaps, less evidence of experimentation in this study, compared to the
previous study. Less evidence is seen of pupils ‘shaping’ the tools of the environment to
achieve creative outcomes. In part, this may be explained by the relative ‘disinclination to

145
experiment’ which has been observed in more experienced students in other research (see
2.2.2). However, I also feel that the task did not establish the forum for discussion observed
in the previous study, where the activity focused on telling a story rather than upon abstract
musical features. This provided a framework for discussion and reflection, with the pupils
able to give clear reasons and suggestions for creative choices, relating these to anecdotal
ideas in place of musical terminology. By focusing on musical ideas with no narrative
framework in this study, we were effectively dealing with these concepts in the abstract, and
the reflective component of the experiential learning cycle was consequently less noticeably
propelled; it is worth noting that these students had no titles for their pieces, though this was
given as an objective at the start. We also saw very little of the drawing or painting actions
which provided many imaginary narrative ideas to the previous study. The pupils were
following a more paint-by-numbers approach, and had no incentive to attribute meaning to
what they were doing18. Therefore, the sense of creative ownership, exploration and
reflection observed in the previous study was largely absent. This suggests that, in sandbox
environments such as Graphick Score, the facility for creation of meaning is as important as
the scaffolding tools which support creation of sounds; when dealing with abstract musical
concepts, we have seen more promising results, in terms of experiential learning, when pupils
shape these ideas within a narrative framework.

7.4 Performance and Composition with iPads at School B,

December 2016

7.4.1 Context 1

Following the previous study, I felt that the way tune-blocks were used within Graphick
Score should be addressed; results at the conclusions of the previous phase indicated that
pupils did not know what to expect from melodic tune fragments (see 6.5.2). Indeed, the idea
of arbitrary song choices for these materials seemed a limitation, and finding melodies
recognisable to the demographic from a single phrase was a challenge. I wanted these
structures to be something which could be harnessed and shaped for a creative intention, and
which perhaps presented more interplay of musical dimensions than the fragments of popular

18
In hindsight, this was initiated by the opening activity, which suggested that the grid-conforming scene was a
‘correct’ way of writing, compared to the freehand scene.

146
music hitherto used. For these reasons, I decided to focus on manipulating tune-blocks for a
single song in the next study, and to address how the narrative might change to results in
musical diversions. I also wanted to examine the possibility for a different social dynamic,
being aware that previous studies had mostly consisted of pupils working individually with
headphones, cut off from their peers for much of the session. This presented the opportunity
for a new situational context; developing bespoke resources for touchscreen tablet, but
continuing to focus on the core objective of experiential composition with digital technology.
The next two studies were conducted with the Year 5 class of a primary school in
December 2016. I decided to apply some of the software features used in Graphick Score,
such as tune-blocks and story-based mechanisms for engaging with musical elements, but to
focus specifically on the use of tablets. I therefore drew up a set of questions for these
particular studies, which relate to my existing questions and use much of the language of the
National Curriculum:
 How might the tablet function within the dynamic of the classroom, in terms of
facilitating ‘performance in solo and ensemble contexts’?
 How might the tablet be used to encourage ‘increasing accuracy, fluency, control and
expression’?
 How might the tablet allow pupils to engage with the ‘inter-related dimensions of
music’ in their performance and composition?
 How might the tablet utilise alternative modes of representation to lead pupils toward
a greater understanding of staff notation?
The studies took place in a primary school over two Friday afternoon sessions of 90 minutes,
taking place a week apart. The Year 5 class consisted of 27 students. No pupils were
identified as requiring special assistance, and none of the pupils took extracurricular music
lessons to learn an instrument or sing. While I led teaching of the lessons, the class teacher
was on hand to provide assistance. 15 iPads were available for use, meaning that most of the
students had one between two. One concern before the study was that the iPads would not be
sufficiently audible through the built-in speakers, while separate amplification for each iPad
was an unreasonable solution due to cost. Different methods were tried to resolve this issue
during the course of the study.
Prior to the session, I contacted the teacher by email to request a description of the
usual activities of a music lesson in Year 5:

147
 The school accesses the 'Charanga Musical School' website which provides a
breakdown of objectives for each year group along with a range of practical resources
for the children to take part in. The activities that the children take part in include
singing and using percussion instruments to help them perform a composition.

In music lessons at this time of the year, the class would usually use Charanga to learn and
sing Christmas songs, perhaps accompanied by percussion instruments. Being a Catholic
school, nativity plays were an annual feature, and the performance of Christmas carols a
popular music activity in December. The Year 5 teacher felt that this would present a good
opportunity to try some new approaches to performance of songs with which the class were
already familiar. Drawing from the National Curriculum, the main outcomes for this lesson
series were to facilitate solo and ensemble performance of Christmas songs using the tablets
alongside percussion instruments and voices, and to demonstrate an attention to expressive
functions, as well as the ‘inter-related dimensions of music’. In the second lesson, a similar
approach was used to facilitate performance and composition.
The purpose of the first lesson was to create variations in performance of the song,
changing the ‘mood’ with attention to musical dimensions, and providing opportunities for
discussion and reflection on how they function individually and in combination. Certain
dimensions could be expected to be addressed more readily on an individual basis, being
fairly instinctive to comprehend and manipulate. In particular, dynamics and timbre might fit
into this category. Other dimensions, by contrast, might require a degree of co-ordination and
communication between players to properly execute, such as tempo and texture. The
remaining dimensions of tonality and structure – and perhaps to some extent, pitch – are in
the context of this task more related to composition and arrangement, and could be expected
to require a greater degree of planning and creative exploration to address.

7.4.2 Method 1
Learning Objective: To perform a Christmas song in groups, and to change the mood using
the ‘inter-related dimensions of music’

The song chosen for this lesson was Little Donkey, a Christmas carol which the class knew
from its usual inclusion in their school nativity plays. This was a particularly suitable piece to

148
display in tune-blocks due to its regular phrasing pattern, which includes a recurring phrase
with an easily recognisable rhythmic and melodic pattern (Figure 7.26). I felt that the chorus
of the song would provide an opportunity to explore some contrasting dimensions, changing
from major to minor chords and having melodic contour and lyrics of a more rousing nature.
Also, the mid-bar chord changes might present more of a challenge than the steady changes
of the verse. I saw these as potential scaffolding activities between the performance of the
verse and a more creative arrangement-based exercise.

Figure 7.26 – The verse of Little Donkey

An interface was made in MobMuPlat which divided each bar into colour-coded tune-blocks,
with each colour corresponding to the accompanying chord. The notes were represented as
rectangles arranged into place according to their rhythmic position within the bar and pitch
relationship to the rest of the melody. When one of these buttons was touched, the
corresponding note of the melody sounded. Therefore, the player simply had to move
consecutively through these buttons following the rhythm of the song. A second window
showed the tune-blocks for the chorus. By swiping left or right, the player could move to the
previous or next screen (Figures 7.27-8). This presented a form of scaffolding for engaging
with musical dimensions, as I hoped pupils would be able to immediately play the piece using
the tune-block interface, and therefore engage with musical changes directed by a wider
narrative framework of what the song is about, and how the meaning might be altered.

149
Figure 7.27 – The verse and chorus of Little Donkey in iPad tune-blocks

Figure 7.28 – The chord interface

A third screen showed six coloured boxes representing the chords of the song (Figure 7.28).
These were colour-coded to match the corresponding tune-blocks, and each contained three
buttons which played the root note, third and fifth of the triad chord in question. I decided to
break the chord into separate notes rather than use a single button so that the pupils could
play broken chords and come up with their own accompanying patterns. I considered
arranging these in a vertical rather than horizontal line, to match the arrangement of pitch in
the tune-blocks, but felt that this would require the hand to be angled at an awkward position.
In my experience of teaching keyboards, novice learners often play chords using the three
middle fingers rather than the more conventional thumb, middle finger, little finger position,
so I decided to arrange the chords on the interface to match this triangular formation. This
more naturally lends itself to touchscreen multi-finger performance, where the thumb is less
likely to be used due to its different angle and shorter length compared to the other fingers.
The final screen allowed the player to choose an instrumental voice for each of the three

150
‘performance screens’, as well as setting the volume. A general MIDI sound set was used for
sound sources.
The plan for the lesson was to divide the class into five groups; three groups of 5 and
two groups of 6. Each group had three iPads and an assortment of percussion instruments,
and was allocated an area of the room in which to work. While planning the session, a
number of obstacles were faced with regard to amplification. The iPads would need to be
sufficiently audible over percussion instruments, and it was clear that the built-in speakers
would not transmit a loud enough signal. Transmitting the three iPads of each group to a
Bluetooth speaker was considered, though the models available could only accept one device
at a time. Ultimately, the setup had to be reconsidered. A laptop was connected to each of the
five speakers, containing the set of general MIDI sound sources. The iPad interfaces were
then edited so that, as well as playing notes directly, they could transmit Open Sound Control
(OSC) messages over a wireless connection to the laptop. This meant that, within each group,
the sounds triggered by any of the three iPads were played through their paired laptop and out
of the speaker. Though the school had a Wi-Fi connection, it was decided that a separate
router be brought to minimise any possible issues with connectivity. In this way, the setup
and equipment required to amplify the iPads in this particular way became quite complicated
and extensive.
This lesson was less open-ended in structure than sessions incorporating Graphick Score,
as the lack of a sandbox environment increased the dependence on a sequence of activities:
1. Students are put into pairs, with an iPad each, and are given time to explore the
interface and work out the song. Groups can then be chosen to identify and play part
of the song.
2. Questions: ‘What is the song about?’ ‘How fast should it be?’ The class are likely to
suggest a slow, plodding tempo to suit the pace of the tired donkey. This generates a
discussion about how musical dimensions can be used to improve a performance.
3. Half of the class clap the suggested tempo, while others play the melody. Question:
‘In each block of melody, how many times do we clap?’ This introduces the idea of a
bar of four beats.
4. Each group then plans and rehearses a performance of the song using iPads and
percussion instruments. The teacher moves around the room asking further questions
to each group to introduce other musical dimensions, e.g. ‘What kind of instrument
would suit this part of the song?’ ‘How might percussion instruments be used to fit
around the melody?’

151
 5. Groups then add a new verse which takes the song in a new direction. They are asked
to come up with a possible new lyric and to represent this with a change in musical
features; e.g. ‘little donkey, little donkey, time to go to sleep…’
6. Performances by each group take place at the end of the lesson, with other groups
asked to comment on how musical features were used to create a mood for the song.
This lesson was planned to allow as much time as possible for practical musical learning,
such as performance and composition, followed by reflection and discussion focusing on the
musical dimensions. Data was primarily collected by means of observation, and an audio
recording was made of the lesson. The teacher, who was present for the sessions but did not
participate, took some photographs.19

7.4.3 Results 1

It was clear from the start of the lesson that the students were enthusiastic about using the
iPads, and that they were generally confident about operating them. This was a characteristic
shared by both boys and girls. Both the teacher and I ensured that everyone had the
opportunity to use an iPad, and the students worked well in groups. No instances of arguing
over iPads or instruments were observed. The setup used for this lesson had, as discussed,
reached a more complicated level than originally intended. Equipment was set up during the
lunch break preceding the lesson, though it was agreed by both the teacher and I that these
requirements would be to demanding for regular use, and that a more efficient solution
should be found for the next lesson. The use of OSC communications proved highly stable, as
each iPad instantly connected to the host laptop and speaker and the transmission of wireless
communications remained unbroken. This allowed the students to rehearse and perform in
groups using a combination of iPads and percussion instruments. However, further issues
with this setup became apparent during the rehearsal session, as some students complained of
being unable to hear what they were doing as the noise level increased. Students were tending
toward turning the volume of their iPad as loud as possible to achieve clarity, and
consequently other students had to follow suit to maintain a consistent level. With three iPads
playing through each speaker, it also emerged that some students were struggling to identify
which sounds were being created by them. I attempted to resolve this by emphasising the

19
This was unexpected, but as the photographs show the activity taking place but not the faces of the pupils, I
later requested permission to use these in the thesis, and obtained consent to do so.

152
importance of communication in group performance, and demonstrating how a performer
should pay attention to the actions of their fellow performers in addition to their own. This
went some way toward resolving the issue, though it was still apparent that the setup used
was largely unsuitable for this kind of lesson.
The tune-blocks interface proved effective, as all students were able to play and
identify the carol. This allowed us to begin discussing performance features immediately
following the starter activity. Students were keen to answer the questions with practical
demonstrations using the iPad:

Me: So if the donkey is feeling tired in the song, how fast should we play it?
Student: It would be like, slow, like… [plays the first two bars at a deliberate,
plodding pace] … instead of… [plays at a franticly accelerated tempo]

This led to interesting developments during the group rehearsals, where students were
engaged with capturing the mood of the song through interpretation of musical dimensions,
or altering the composition in some way. I pointed out that percussion can be used to
augment the mood of the song by finding interesting rhythms, perhaps filling in gaps in the
melody, or taking a rhythm from a different song. This meant that the arrangements
developed interesting textures as the lesson unfolded. Likewise, performance of the main
melody was left to one iPad at a time in all groups, with others using the interface to provide
some form of accompaniment, again supported by the use of colour-coding in the tune-blocks
layout. In some cases, performance of the melody was divided between different players at
different points, allowing students to switch between different roles. All the groups were
engaged in discussion about these ideas, and were eager to perform when the opportunity
arose.
Most performances were characterised by the use of clopping, steady rhythms on
wooden percussion instruments during the verse, and tambourines or bell-like instruments
during the chorus. The chorus was also generally louder than the verse. Both of these features
were identified as being suitable due to the lyrics – the plodding, tired donkey described in
the verse, and the contrasting chorus of ‘ring out those bells…’. One group added a new set
of lyrics for an additional verse:

Little donkey, little donkey, sleep and dream tonight,

Little donkey, close your eyes and go to sleep so tight,

153
This was played using a low-pitched bass sound, with a quiet tinkling of bells and bell-sounds
from the other iPads. The low pitch was characteristic of the slumping, drowsy donkey,
contrasted by the high tinkling bells representing the stars above.
Three of the five groups used the whiteboards on their tables to make notes on the structure
of the performance, regarding who would be doing what, or the notes they wished to play.
This was undirected and emerged as they experimented with different ways of using the iPad
other than playing the melody. Two of the groups changed the rhythm of the melody, wanting
to achieve a more upbeat mood (Figure 7.29).

Figure 7.29 – Rhythmic variations

In the second group, this formed part of an intriguing structure. The verse had a tremolo-like
effect, achieved by playing each note multiple times in rapid succession, using a guitar
timbre. This was followed by the chorus, utilising bell sounds in the same manner as other
groups. Finally, this was followed by an improvisation between two iPads, selecting notes at
random within the colour-coded structure set out by the tune-blocks.

Figure 7.30 – Melodic variations with the same rhythmic structure

One group used notes from different tune-blocks to create a new melody to the same rhythm
(Figure 7.30). This was followed by the normal melody of the chorus, while the rhythm of the
verse continued to be played using the percussion instruments. This created an effective and
interesting textural contrast. Another group followed the tune-block structure, but used this to
create a new rhythmic and melodic pattern, while preserving harmonic structure (Figure
7.31).

154
Figure 7.31 – Melodic variations with the same harmonic structure

The final group had diverted the most from the original arrangement. A cello sound was used
on all iPads, and only the screen showing the triad chords was used. Performance consisted of
chords of long duration, alternating with broken chords moving in different directions,
accompanied by a steady rhythm. The other students were particularly enthusiastic in
commenting on this performance, and in finding comparisons or ways to describe it:

‘That’s definitely not Little Donkey no more!’

‘It’s like a drama… it sounded like the music to ‘Batman’ or something….’
‘It was like an organ weren’t it?’

This led to discussion on the qualities of dramatic music, and of the suitability of instruments
like the cello and organ to creating this kind of mood, being capable of low pitches, and long
sustained durations.

7.4.4 Context 2
The purpose of this lesson was to explore how a Christmas carol is constructed, and to
compose a new one based upon these guidelines. It took place exactly a week after the
previous study, with the same class of students. As such, it provided a valuable opportunity to
build upon the learning outcomes of the previous session.

7.4.5 Method 2
Learning Objective: To compose and perform a Christmas carol

The interface used in the previous session had produced interesting variations on musical
dimensions, but had perhaps proved restrictive, as it limited pupils to the notes of the song

155
Little Donkey. I wanted to continue to use tune-blocks and tablets, but within the context of a
more open-ended task.
To avoid the increasingly complex setup that was eventually used for the previous
lesson, the class was divided into pairs rather than groups, with each having one iPad.
Headphone splitters were used to send the output to two pairs of headphones, while a single
Bluetooth speaker was brought for the purposes of performance, as only one iPad would need
to be used at a time. Two spare speakers were brought, however, in case the opportunity for
multiple iPads to perform together arose. Use of iPads was simplified to a single interface,
the Rainbow Keyboard, again made in MobMuPlat. This consisted of eight numbered and
coloured keys, corresponding to an octave of the major diatonic scale (Figure 7.32). These
keys illuminated when pressed, and played high-quality instrument samples. Clip-art images
were used to show the selected instrument, and volume could again be altered using a dial.
The screen was divided into two identical sections, allowing students to select different
instruments or volume settings and easily switch between the two. On a second screen, the
upper section was flipped upside-down, so that students had the option to play the iPad in
pairs, by laying it flat on a table and sitting on either side of it.

Figure 7.32 – The Rainbow Keyboard

Having found the tune-blocks approach to be effective in breaking down a melody

into manageable sections, and also finding that the students could follow this with minimal
direction, I decided to make a game out of the idea that could serve as a starter activity.

156
Christmas carols were again broken into bar-long tune-blocks, but would be mixed up. The
class would then have to work out how to play them, identify the song, and arrange the tune-
blocks into their correct structures. To use carols which would contrast in mood to Little
Donkey, and which prove more challenging to work out and play, I decided on Jingle Bells
and Rudolph the Red Nosed Reindeer. This would allow the students to use musical
dimensions more characteristic of a sprightly mood, and the more complex rhythmic
groupings and melodic contours would prove more of a challenge. The notes of the melody
were numbered and colour-coded to match the iPad interface.
My original intention was to make a new tablet interface in which these tune-blocks
could be arranged, but ultimately decided to print these off onto pieces of card, to be put back
together like a jigsaw puzzle. The students could later turn the cards over, and compose their
own tune-blocks on the blank side, arranging these into a larger composition.
The plan for this lesson was as follows:
1. Students work in pairs, with a set of cards and an iPad, to put an entire tune together.
Those who complete this task quickly are given percussion instruments and asked to
plan a performance using the musical dimensions, in the same manner as in the
previous lesson.
2. Selected pairs are then asked to perform the melody, followed by a discussion on what
musical characteristics are commonly found in Christmas carols.
3. The class writes a short carol together. They are asked for a sentence that has
something to do with Christmas, which is then chanted to fit a four-beat bar. A
student is then asked to improvise a melody to this rhythm using the iPad. The class
write another three tune-blocks in the same fashion to complete the song.
4. Each pair then writes their own carol. The teacher moves around the room to assist,
and remind how musical dimensions might be used to create a mood.
5. The carols are performed at the end of the lesson, with the rest of the class
commenting on how a Christmas mood was created using the musical dimensions.
This was the last lesson before the Christmas holiday, so the ways in which musical
dimensions might be used to create a seasonal mood were a prominent feature of the lesson.
As with the previous lesson, the class start with a practical music-making activity, and the
lesson follows a cycle of practical activities followed by reflective discussion.

157
7.4.6 Results 2

The use of headphones rather than the complex setup used for the previous lesson avoided the
issues previously encountered, though it restricted performance and composition to groups of
two. Again, the students responded well to the task, were observed to share use of the iPad,
and were keen to perform. The starter activity of arranging the tune-blocks was met with
great enthusiasm, and the students were able to work out the connection between the interface
and the tune-blocks with no guidance (Figure 7.33). This may be a result of playing tune-
blocks directly from the iPad in the previous lesson, or it may be due to the intuitive nature of
this as a mode of representation. During the performance, students typically played at a slow
pace, following the tune-block cards carefully, except for one pair, who tried to maintain a
faster pace to keep up with the accompaniment of percussion bells they had decided upon. It
was notable that one of the four pairs to perform following this activity had at least two tune-
blocks in the wrong place, but performed the carol following this structure. This suggests that
they were relying on the tune-blocks in the task and performance, rather than relying on their
memory of the familiar tune.

Figure 7.33 – Arranging tune-blocks into a complete structure

158
 After a discussion about the lyrical themes of Christmas carols, a student suggested a
first line for our carol: ‘snow is falling all around’. Students were then asked to chant this
over a four-beat bar to suggest a rhythm, with an interesting pattern eventually being decided
upon by varying the durations of syllables. Another student suggested ‘snow is falling on the
ground’ as a second line, following the same rhythmic pattern and ending with a rhyme. A
number of improvised melodies were then suggested for the first line. The students soon
realised, after reviewing the completed tune-block carols, that certain notes, particularly the
red root notes at either end of the Rainbow Keyboard, would make the melody sound finished
while others would make it sound unfinished. When a melody which generally rose in steps
was selected for the first line, another student made an interesting observation:

Student: ‘Could you do like, for “the ground”, could you go a bit lower on the bottom
one though…’
Me: ‘Yeah, so go lower on the last note? Why would we go lower on the last note?
Other student: ‘Because it makes variation.’
First student: ‘Because the ground is below.’

The student had seen the opportunity to use the musical dimension of pitch to represent the
lyrical theme, though the interrupting student’s observation that this would give variation to
the melody is also valid reasoning.

159
Figure 7.34 – Composing their own carols

The class were then able to compose their own carols following this method. Interestingly,
pupils improvised their own notational systems from the numbers given on the Rainbow
Keyboard; this was undirected and unexpected. Many of the lyrics were borrowed from
existing carols, and altered to fit a new structure or theme. In a continuation of the pattern
seen in previous studies, this also frequently involved a rhythmic structure borrowed from
one of the carols used in the starter activity (Figures 7.34). Others had composed original
melodies, often paying attention to melodic contour and the resolution of the tune by ending
on an appropriate note (Figure 7.35). There was evidence of the iPad interface leading to
more varied compositions, as students attempted to use all of the notes on screen by
following ascending and descending patterns (Figure 7.36).

160
Figure 7.35 – Original composition by group

Figure 7.36 – Original composition by group

The composition of one student was characteristic of pealing bells, and followed a clear
pattern, whereby the hands started at opposite ends of the keyboard and worked towards one
another. In this way, the student had devised a memorable kinaesthetic pattern (Figure 7.37).
When I attempted to repeat this performance to demonstrate the memorable nature of the
pattern, the student was eager to again show how he had done it:

Me: ‘That was very memorable… I think I even remember how you played it…’
[attempting to play the piece]
Student: [correcting me] ‘I’ll show you. It went…’

Figure 7.37 – Original composition by pupil

After the lesson, one student was copying their melody from their whiteboard onto a piece of
paper, saying they wanted to work out how to play it on a recorder. I opened another app with
a standard piano keyboard layout, and showed the student how the numbered notes matched
up with the white keys, so they were able to write down the note names for their melody.
They then told me that their parents had a keyboard and they would play the tune on the
instrument when they got home. The tablet can be seen acting here as a scaffolding tool, and

161
leading to independent learning on the part of the student. In retrospect, this approach might
have been used at the end of the lesson so that all students had a notated version of their
composition.
Examples of deep learning are present in this study, defined by Fullan and
Langworthy as instances of learning which ‘[develop] the learning, creating and “doing”
dispositions that young people need to thrive now and in their futures’ (2014, p.i), as well as
what the authors term ‘new pedagogies’, ‘models of teaching and learning that are both
enabled and accelerated by digital technology and resources, and that take place in
environments that support “deep learning”’ (Clarke & Svaneas, 2015, p.7). The use of the
tablet to facilitate performance through accessible modes of interaction and representation
allowed the students to move further up the ‘pyramid of processing’ (see 2.2.1) and engage
with higher order skills such as improvisation, and the manipulation of musical dimensions.
We were able to conduct a highly practical lesson, where no students became stuck on lower
order musical skills such as repetition or memorisation, and instead focus on highly social,
creative and exploratory musical activities. As well as meeting the intended lesson outcomes,
several advanced aims from the National Curriculum were also met.

How might the tablet function within the dynamic of the classroom, in terms of facilitating
‘performance in solo and ensemble contexts’?
The students successfully and confidently performed in larger groups, pairs and solo contexts.
Whether playing iPads or percussion, the students moved rhythmically to keep in time, and
displayed other kinaesthetic responses to musical dimensions, such as hunching when playing
quietly. Kerchner (2000) observes that many children of the same age group participating in
this study are ‘kinaesthetic listeners’, responding to music with physical actions. That they
respond in this way when playing a tablet is an indication of their readiness to employ it as a
musical instrument, and also helped to co-ordinate the ensemble when students struggled to
connect what they were playing with what they were hearing in the first lesson. Clearly, the
setup used in the first lesson was not a sustainable solution, and other technical options
should be explored in further research employing the tablet as an ensemble performance
instrument.

162
How might the tablet be used to encourage ‘increasing accuracy, fluency, control and
expression’?
As discussed, the accessible modes of representation and interaction employed allowed the
students gave the students a greater degree of control. The tune-blocks interface became a
scaffold, or framework, from which the students could deviate whilst still maintaining sight
of the original melody. This allowed them a further degree of control and accuracy in their
invented patterns. We can also see evidence of this in the second lesson, where the simplicity
of the interface led to fluent and confident performances. In this case, however, the tablet was
relatively limited in terms of dynamic control, or the capacity for varied articulation. There
are many functions of the touchscreen tablet which could be utilised to allow students a
greater depth of control, such as touching the screen in different place for different dynamic
levels, tilting to control certain parameters, or using a variety of gestures.

How might the tablet allow pupils to engage with the ‘inter-related dimensions of music’ in
their performance and composition?
These lessons focused heavily on the use of musical dimensions to create mood or represent
lyrical themes. Students were able to show their understanding of musical dimensions, but
also how these relating to their performances and compositions, or those of others. This was,
for the most part, something that the students were able to ‘jump ahead to’, as they were able
to perform the pieces following the starter activities and able to focus their attention on the
creative development of their performance. However, as above, I feel that there is far greater
capacity to use a range of tablet functions to communicate musical dimensions, especially if
we wish to explore the potential of the tablet as an expressive instrument in its own right and
not merely a surrogate or scaffold for ‘real’ instruments.

How might the tablet utilise alternative modes of representation to lead pupils toward a
greater understanding of staff notation?
The use of tune-blocks was a highly successful component of both lessons, as was the use of
game activities, such as the jigsaw puzzle starter of the second lesson, in promoting
understanding of pitch and rhythmic relationships. This led toward a greater understanding of
staff notation in at least one instance, as the student at the end of the second lesson learned
the corresponding note names so that they could play their composition on other instruments.
Simplified layouts of this kind, especially when interactive, can help children to understand
the basic functions of musical notation, and the touchscreen is an ideal interface for

163
communicating this meaning, as it can effectively combine the functions of score and
instrument. In the first lesson, the students literally played the score, by following the
onscreen notations, later deviating from this score to produce original musical ideas. By
contrast, in the second lesson, the students used the tune-block cards as scores, or wrote their
own on whiteboards.

7.5 Sound Objects and Keynote Sounds

Figure 7.38 – Graphick Score in 2017. A version with a white canvas window was requested by the London
Review of Education due to printing considerations.

The previous studies highlighted the significance of kinaesthetic connections between

meaning and musical gestures; thought experiments, such as asking ‘how does the donkey
feel in the song?’, led to experimentations with musical dimensions based upon these ideas of
motion and action, while simple kinaesthetic patterns were composed and memorised using
the Rainbow Keyboard. In these examples a result which is musically complex, or variable, is
represented, understood and reproduced according to a simple kinaesthetic action. A
precedent for this exists in pedagogic composition, for example, in Kurtág’s Alapelemek from
Játékok20 vol. 6 (Figure 7.39). Notations exist as kinaesthetic actions which are easy to
understand: raking up or down the piano, bashing a group of keys with the fist, etc. However,
20
Incidentally, this translates to ‘games’

164
these identifiable structures act as a scaffold for other musical dimensions, such as reading
differences in pitch or dynamics on a score. The novice player therefore has an accessible
mode of interaction for engaging with these skills, circumventing the more complex demands
of reading specific pitch and durational values.

Figure 7.39 – Kurtág’s Alapelemek from Játékok vol. 6

The most successful studies, in terms of creative outcome, had so far focused on telling a
story as a scaffold for musical composition. I had observed how simple ‘sound painting’
gestures had been used as actions or events, and images or ideas had been taken as wider
contexts, characters or settings (see 7.2). Drawing from Schaeffer’s (1977) description of
soundscapes, these images and ideas are like the ‘keynote sounds’ which provide a backdrop
to the ‘sound objects’ which represent specific actions and events.
The menu of melodic tune-blocks in Graphick Score was replaced by a menu of
kinaesthetic sound-objects (Table 7.1). This was aimed at facilitating accessible descriptions
of musical gestures, devices which novice composers could understand and identify, and use
as a creative vocabulary for telling a ‘sound story’.

165
Name Shape Musical Name Shape Musical Description
Description
Burst Crescendo Shrink Decrescendo

Crunch Dissonant cluster, Slide Descending arch over

sfz an octave

Float Ascending arch over Splat Descending

an octave glissando, p to fff

Glide Long duration across Step Four steps down

a bar down

Hop Three quick steps Step up Four steps up

up, p to f

Jump Four steps down in Tickle Trill

down thirds

Jump up Four steps up in Trickle Descending scale

thirds

Launch Ascending Tumble Long descending

crescendo glissando

Leap Three quick leaps up Turn Turn

(thirds), p to f

Shimmer Tremolo a third Zigzag A third up, a step

apart down, a third up

Table 7.1 – Table of sound objects

Many of these sound objects are recognisable musical ornaments given simple kinaesthetic
descriptions such as burst and tumble. As with tune-blocks, these sound objects can be placed
anywhere in the canvas window, copied, resized and rotated to form complex larger

166
structures and sequences. This also addressed the pressing need for dimensions such as
velocity to be changed through interaction with the canvas window (see 7.3.3); a variety of
musical dimensional interplay could be presented by these scaffolding structures, created and
manipulated through simple interactions.
To accompany the sound objects, and to support storyboard composition with
Graphick Score, a menu of images was also added to the graphics menu. Each of these
corresponds to an audio file, which loops when the scene is played (Table 7.2).

Name Image File Audio File (description)

Balloons Cheering, party horns etc.

Brass Band Marching band, brass instruments, percussion

Birds Tweeting, birdsong

Car Car horns, traffic sounds

167
Traffic Cone Roadworks, diggers, drills

Cow Cows mooing, pastoral sounds

Crowd Crowd chatter

Fire Fire crackling

Football Unison crowd noises (disappointment,

celebration)

Helicopter Rotating helicopter blades

168
Lightning Thunder clap

Moon Owl hooting, crickets

Phone Chirping ringtone

Plane Jet engine passing

Raincloud Rain falling

Ship Ship horn, seagulls

169
Siren Alarm, siren

Sun Parents chatting, children playing in park,

swings etc.

Snow Footsteps crunching in snow

Tower Clock tower striking the hour

Train Train passing

Tree Dog barking, footsteps in leaves, distant sounds

of people talking etc.

170
Wind Wind blowing leaves

Table 7.2 – Table of storyboard images and sound files

These audio files are all different lengths, meaning that they overlap when combined; this
prevents the result from becoming too repetitive. Up to three can be added to each scene.
Resizing the image alters the volume of the corresponding audio file, so the user can create a
simple mix of sounds to tell a story, augmented by sound objects.

171
7.6 Storyboard Composition at School D, March 2017

7.6.1 Context

I was invited to run an afternoon class on music with a Year 5 class at School D, a primary
school. The teacher had informed me that the class would be using graphic notation later in
the term to tell the story of the RMS Titanic, so I suggested a lesson focusing on the link
between sound and story. The purpose of the study was to further examine how Graphick
Score may be used to facilitate music composition, following changes made to the interface.
Previous studies highlighted the importance of a creative stimulus, such as a story, image or
idea, in producing musical decisions. In an earlier study (see 7.2), I observed how the pupils
used the story from the demonstration as a starting point, and many took the image of a sun
or raincloud (indicating tonality) as a reference point for building narrative ideas. I therefore
reasoned that further images and sounds of a recognisable nature could be used to enable or
accelerate creative explorations, and that the examples of what I term sound-painting or
sound-sketching might be shaped by this wider meaning, or conversely, may shape the
meaning of the story. Because I wanted to promote reflection and discussion, I felt that these
stories should be conducted in groups rather than individually.
A library of images and corresponding audio files were downloaded from royalty-free
sources, and a new section of the graphics menu was added. This allowed the user to add
images to a scene, which played looped audio files corresponding to the chosen image when
the scene was played. These images could be resized and moved around the canvas window.
As with geos, size was linked to volume, so larger images increased the volume of the linked
audio file. Changes to the audio when images were moved were also initially planned but
were not implemented in time; these included changes in panning of the audio files
depending on the position of images in the X axis. Because I wanted the pupils to be selective
about the combination of images, and to not focus exclusively on this feature, the number of
image/audio files per scene was limited to three. Smaller changes included a re-arrangement
of the layout to make room for this new feature, and an indication of the chosen MIDI
instrument; a broken tonic chord of the chosen key was triggered when an instrument was
selected, to give the user an idea of the timbre.

172
7.6.2 Method

Learning Objective: To compose a musical story using graphic scores

8 laptop computers running Graphick Score were set up in the classroom, each with a
headphone splitter which allowed the audio output to be divided between up to 6 pairs of
headphones. The programme was also installed on the teacher’s desktop computer at the front
of the class, which was connected to an interactive whiteboard (or smart board) and speakers.
This allowed Graphick Score to be operated via touchscreen, as well as onscreen drawing
using the smart board pen tools. During the lunch break, I set this up and gave a brief
demonstration of the programme to the teacher, showing the basic operations and functions.
We agreed that, though I would lead the session, the teacher would provide support, and that
they would intervene at any time if they recognised learning opportunities or barriers.
The 27 pupils were divided into 8 groups of 3 or 4 per group. After being introduced
by the teacher, I asked one student to draw a picture of a fish on the board, and another to
draw an elephant. I then told the class that would play them two pieces of music; one about a
fish and one about an elephant. These were The Aquarium and The Elephants from The
Carnival of the Animals by Saint-Saëns (the titles were not given). The class were asked to
identify which piece of music matched to which picture, and to explain why, linking
characteristics of the music to the drawing.
The class were then shown a number of scenes using Graphick Score on the smart
board, and were asked how the sounds and images communicated a story. The first involved
thunder and rain sounds, as well as descending blue patterns of xylophone and glockenspiel
notes, in a minor key. The second kept the same sound files and images, but was
accompanied by a happy melody in a major key, with yellow and orange geos. The class were
asked to suggest changes. Finally, two pupils were asked to come to the front of the class to
select a sound file and then draw a pattern in the canvas window, to demonstrate how a scene
could be started.
Each group was given 5 minutes to look and listen to the options in the programme,
and then was given a piece of paper and a set of coloured pens, and asked to sketch out their
chosen story. The rest of the lesson was spent using Graphick Score, with me and the teacher
circulating the room to discuss the work of the group, and to ensure that all participants
shared opportunities to use the programme. At the end of the lesson, the teacher directed the

173
groups to play their pieces, and asked the rest of the class to guess the story, and to give
reasons for their decision.

7.6.3 Results

The discussion of The Carnival of the Animals provided an engaging start to the lesson. The
class immediately identified The Aquarium as being about fish, with appropriate observations
on the melodic contour of the piece, such as ‘it flows and glides.’ These discussions could be
linked to musical elements:

Pupil: ‘It’s like magical… like the sea and… oceany!’

Me: ‘Oceany? That’s an interesting verb. So what makes it sound like that?’
Pupil: ‘Because it twinkles.’
Me: ‘Okay. So what instrument gives it this twinkling sound?’
Pupil: ‘Was it like… you know, like, chimes?’

Following this discussion, I played The Elephant. There was an outbreak of laughter from the
class from the introductory clumsy waltz pattern of the piano, and again from the low melody
of the tuba. Kinaesthetic reactions were also observed; stomping and slumping motions in
time to the music, imitating the lumbering movement of a heavy animal. When asked what
animal the piece represented, there was a resounding shout of ‘elephant!’ A discussion
followed:

Pupil: ‘It was like a stomp.’ ‘There was a trumpet.’

Me: ‘Why would that suit an elephant?’
Pupil: ‘Because of the trunk.’
Me: ‘Yeah, and they make that noise.’ [I unsuccessfully attempt an elephant noise, responded
by attempts by the rest of the class]

We then discussed how it was a similar instrument to the trumpet, also made of brass, called
the tuba. I explained that this was a big instrument with lower notes (deepening my voice for
effect). I then told the class that the music of these pieces told a story, and that we would
make a musical story using a computer programme (indicating Graphick Score on the smart
board); the class were engaged, and noticeably enthusiastic at the prospect.

174
 The examples in Graphick Score followed a similar discussion to The Carnival of the
Animals. Pupils noted that the downward motion of the first example ‘sounds like rain, going
down’, and that a sun would be more suitable than a raincloud for the second example
because the tune sounded happy. Moving on to a demonstration of a new scene, a pupil chose
birds as the image and sound file, and another was asked to go through the instrument options
for a suitable instrument, settling (as anticipated) on a flute. The class then identified this as
suitable due to the similarity with tweeting and birdsong. By this point, the class were
noticeably restless; many were chatting or otherwise distracted, and responses were less
engaged. I therefore decided to let the class put on their headphones and try out the
programme, while the teacher handed out paper for the graphic scores. This was met with a
more enthusiastic response; the entire class were seen to be exploring the different
possibilities, and none appeared confused or disengaged.
The teacher was notable very capable in providing support to the pupils on how to
perform various features, remembering much of the details from the few minutes in which I
demonstrated the programme before the lesson. The pedagogical training of the teacher was
valuable in directing the class, having more classroom experience than me and, of course,
knowing the group well. At the end of the lesson, we played all of the pieces through the
smart board and speakers, and I asked the first group what the piece was about; the teacher
instead suggested that we ask the other groups to guess. They then led the plenary, asking
other pupils to guess the story, and to provide reasons. Their classroom management skills
mitigated any issues here that I may have struggled to address, for example:

‘I think if you’d have done a bit more, we’d have been able to realise your story’ [moving on,
when the students were unable to guess one of the least substantial pieces]
‘You aren’t ready to play yours, so we’ll move on’ [as a group were not paying attention]

Their encouraging attitude and rapport with the class also helped to facilitate a productive
plenary session:

‘Wow, well done! Give them a clap!’

‘I’ve definitely got an image of what I think it means.’

The teacher later (after the lesson) told me that Graphick Score would be useful for the
module on the HMS Titanic:

175
 ‘Maybe we could download the programme… I don’t know how much it costs…’

I told them that I was not trying to market the programme (it being an open-source Pd project
and not a commercial software application) but that I would send them a copy along with the
exported video clips of the pupils’ work.
The stories made with the programme have some interesting features. Some focused
mostly on keynote sounds, while others made use of sound objects and ‘painting’
interactions. The written scores were also useful in clarifying the aims. Looking at the work
of Group 1 (Figure 7.40), the two scenes are clearly discernible from the score. ‘Launch’ and
‘tumble’ sound objects have been used to represent the movement of birds, with melodic
patterns on brass instruments, giving the first scene a tuneful and happy character. This
contrasts with the droning metallic sounds in the second scene, when the rain starts. The
dynamic changes within these sound objects appear to have directed the pupils to make
dynamic variations of their own, as the overall velocity of these objects has been varied.
Group 2 (Figure 7.41) demonstrate similar contrast; the first and third scenes (sun and snow)
make use of a slow texture of overlapping tuneful patterns, while the second and third (rain
and transport) are fast, with repetitions of the same pitch, suggesting duration and monotony.
Cymbals and drums are also used here, indicating a focus on rhythm and timbre over melody.
Also, it is interesting to note that each scene appears to end with a precursor of what will
happen in the next, heralding not only the motivic gestures, but also the ‘colour key’ to be
used in each scene.
Group 5 (Figure 7.42) made one graphic score to indicate the wider structure, before
dividing this into smaller sections. The drawn flute line, representing the flight of birds,
shows evidence of copying and rotating to create symmetrical patterns, and is clearly
represented in the graphic score; interestingly, this has been copied and rotated to form a
countermelody. A contrasting texture and timbre of busy drums and alarms in the following
scenes is also clearly indicated. These ‘bird flight lines’ are also seen in the work of Group 7
(Figure 7.43), who supplement the images with other drawn additions, such as the ‘lemon
tree’ in their second scene.

176
Figure 7.40 – Group 1

177
Figure 7.41 – Group 2

178
Figure 7.42 – Group 5

179
Figure 7.43 – Group 7

While some groups relied almost exclusively on keynote sounds, others focused on
sound objects and drawing. The changing seasons score made by Group 6 (Figure 7.44) tells
a clear story which is represented in the drawn materials. Bird flight and the business of
crowds in the park are identifiably through patterns of movement and texture. The winter
scene, with no keynote sounds but a texture of blue celesta tune-blocks, is surprisingly
consonant and melodically coherent, though clearly built through an improvisatory
arrangement of blocks.

180
Figure 7.44 – Group 6

181
 A particularly interesting example is Group 8 (Figure 7.45), who used musical
terminology in their scores, suggesting that at least one member of the group has musical
training, although the distinction of ff for the aeroplane and mf for the crowd appears to have
been swapped around in the eventual realisation of the scene. In this example, the students
wanted to tell a humorous story of going on holiday on a sunny day and arriving to a storm:

 Scene 1 – Getting to the airport

o Samples – Traffic and birdsong
o Music – A clutter of brass instruments (car horns) at a low pitch (on the
ground) moving across the screen (traffic)
 Scene 2 – On the flight
o Samples – Aeroplane taking off, chatter of a crowd
o Music – A rising sequence of notes of increasing volume (take-off) and a high-
pitched happy melody (flying)
 Scene 3 – Arriving to bad weather
o Samples – Rain, wind and thunder
o Music – A descending sequence of notes (plane landing) and descending
patterns played on xylophone and glockenspiel sounds (rainfall)
 Scene 4 – Night time
o Samples – Owl noises, crickets
o Music – Descending patterns cease (rain stops, sun goes down) then twinkling
bell-like sounds (stars)

182
Figure 7.45 – Group 8

This was the last study conducted with Graphick Score for this research project. It
was similar in structure to the study at School B (see 7.2) in that stories were constructed with
sound. However, the addition of ‘sound objects’ and ‘keynote sounds’ appears to have
provided a more substantial framework for creative exploration and reflection. It is more
apparent, in these examples, how the intended stories relate to sound gestures, as
demonstrated by the guessing game organised by the teacher. A more social application was

183
engineered through the use of headphone splitters, though further collaborative approaches
should continue to be explored.
It is notable that the pupils and teacher were able to use the programme with very
little explanation; in hindsight, the starter activities preceding the project were perhaps
excessive, as pupils were visibly impatient to start the main activity. It is also promising that
the teacher expressed an interest in using the programme for an additional module. Though
discussions often involved face value commentary upon the keynote sounds, the teacher was
able to relate these discussions to musical dimensions:

[playing a piece with owl keynote sounds]

Pupil: ‘Was it at night when you hear the owls?’
Teacher: ‘What else made it sound like night?’
Pupil: ‘It was quiet…’

This suggests that further educational opportunities may be presented by manipulation of

these keynote sounds. The possibility to record audio and directly interact with it in the
canvas window is an avenue which should be explored in future versions of Graphick Score.
Migrating the design to a touchscreen app, and implementing further controls for recording
and editing, as well as ‘sound painting’, will be a focus for future development.

7.7 Performance and Composition with iPads at School C,

June – July 2017

7.7.1 Context 1

Brown (2007a) notes that the SoDaR researcher should be sensitive to unexpected outcomes
emerging from research, especially where these present new educational opportunities.
Though this project focused on interface design, and principally the development of Graphick
Score, the study undertaken with iPads (see 7.4) had prompted interesting questions regarding
the use of such devices for performance and composition. I felt that this avenue had been
hitherto underexplored in my research, and decided to undertake a further series of studies to
examine how the touchscreen tablet might be used for group music activities.

184
 Admittedly, it may have been
preferable to conduct this kind of
research earlier, to examine how
pupils responded to existing digital
resources. However, sensitivity to
emergence is at the core of my
methodology, and the importance of
the touchscreen tablet as a classroom
digital instrument only emerged
during the course of my design
Figure 7.46 – Smart keyboard in GarageBand
process; indeed, when this project
started in 2014, the tablet was still a fairly recent invention, and had not yet achieved the
level of influence upon classroom activity seen today. However, the question of how it can be
harnessed effectively for musical learning has received relatively little attention (see 3.2.1).
Williams (2014, p.49) has argued that the iPad can function as an expressive musical
instrument:

When used to make music, the iPad is a musical instrument. It can be performed well
or poorly. It takes practice to build performance technique on it. It will do nothing
without musicianship, creativity, and imagination supplied by a person. It has musical
limitations just like any instrument, but in the right circumstances, it can be used to
make amazing music.

This resonates with the key finding of Fullan and Langworthy (2014, p.30) who report that,
when using digital technologies it is ‘the pedagogy of the application of technology in the
classroom which is important: the how rather than the what.’ As portable tools with a tactile
mode of interaction, tablets might function like any other musical instrument. The National
Curriculum tells us that children should ‘play tuned and untuned instruments musically’ from
key stage 1 (DfE 2013, p.2). Instrument stocks in primary schools have generally consisted of
miscellaneous inexpensive percussion instruments, along with rudimentary melodic
instruments such as the recorder. Policies such as the Wider Opportunities and First Access
schemes have had some success in bringing a greater range of instrumentation to the
classroom, though this has been beset by various logistical and financial barriers, mainly
attributed to the transportation and setting up of instrument stocks (MU 2014, p.3). Access to

185
tuned instrumentation in the classroom must then be a more frequent opportunity, and while
policies like First Access should continue, schools cannot rely completely on this for
instrumental provision. The tablet can provide a scaffolding opportunity in the classroom by
emulating a range of tuned and untuned instruments. In addition to many keyboard and piano
interfaces, free apps to emulate the playing interface and timbres of instruments such as
guitars, concertinas, drum kits, gamelans and Theremins are immediately accessible and
plentiful. The capacity for customisation and communication between devices also opens up
possibilities for musical performance unique to digital instruments.
I chose an existing app, GarageBand, due to the range of performative scaffolding
tools; these include ‘smart keyboards’, where chords can be programmed into an assistive
interface, and drum machine pads (Figure 7.46). The app also contains a sequencer layout
similar to Logic Pro or Ableton live, making it suitable for both performance and
composition. I also developed a sampling interface, Sample-It, using MobMuPlat (Figure
7.47). This enabled the recording and playback of audio clips, with basic editing functions
such as pitch control and audio reverse.
To conduct this research, I returned to School C during the summer term. I proposed
to deliver these sessions as an ‘iPad ensemble’, with each pupil ‘playing’ a tablet to produce a
rendition of a popular song, Happy by Pharrell Williams. This song was chosen because it is
well-known and popular among children, receiving a lot of airplay and appearing in the film
Despicable Me 2. It also has a consistent lyrical theme of happiness that I felt could be clearly
related to musical dimensions.

Figure 7.47 – Sample-It

186
7.7.2 Method 1

Learning Objective: To perform ‘Happy’ as a class using GarageBand

I was allocated the Year 3 class for an afternoon lesson. This group were younger than the
other pupils I had worked with, but saw it as an opportunity to explore whether we could
scaffold to some clear outcomes using digital technology. The class consistent of 26 pupils,
sat in 4 groups. 30 iPads had been loaned, running GarageBand and Sample-It. Prior to the
session, the class teacher forwarded some music targets for the group:

To be able to recognise sounds that move in steps or leaps.

To use musical vocabulary to describe a piece of music, discussing likes and dislikes -
and why!
To state how their work has been improved.
Compose repeated patterns.
Choose and combine different sounds to create an intended effect.

I decided to plan a session which would incorporate several of these targets within a simple,
performance-based activity:
1. The class introduces themselves by ‘playing’ their names on the drum pad interface,
i.e. following the syllable pattern with drum hits.
2. Play ‘Happy’ and discuss likes and dislikes of the song; what makes it sound happy?
3. Compose a basic 4-beat drum loop, counting to 4, to establish the pulse
4. Sing the verse as a repeating call-and-response over this pulse, where each student
answers the vocal line by ‘playing’ their name as before
5. Work in groups to add in the chords using the ‘smart keyboard’ – this can be followed
from a simplified chord sheet
6. Perform the piece at the end, and discuss what went well, and what could be added

7.7.3 Results 1

This session started promisingly, with pupils able to follow activities such as ‘playing’ their
name using the drum pads, and some engaging discussion of the song. This discussion led to
two features of the song being highlighted; the lively beat and the happy tune. We then tried

187
to set the pulse of the tune with a kick drum pad so we could perform as a call and response,
between the vocal of the verse and drum fills. This ultimately proved very difficult; the class
were unable to maintain a consistent pulse. This was also attempted with clapping instead of
iPads, but with similar results; the pulse very quickly fell apart. With the assistance of the
teacher, we were eventually able to establish this coherently enough to continue with the call
and response activity.
Having drawn a line under this task, I moved on to the chord accompaniment using
the smart keyboard interface. Turning to the lyric and chord sheets, I asked the class what the
symbols meant:

Pupil: ‘Are they notes?’

Me: ‘Sort of. The letters actually tell us a group of notes that we can play that will go
with the tune.’

None of the group had heard the term ‘chord’ before, but this explanation seemed sufficient. I
had planned to show the pupils how to set the chords up on the smart keyboard, reasoning
that this would provide an opportunity for them to learn something which they might apply
outside of the classroom to play other songs. This too proved far more complicated, in
practice, than expected. It was possibly due to the unfamiliar nature of the symbols, plus the
complications of sharps, flats and tonality, that many of the pupils struggled to navigate the
menus and assign the chords listed on the lyric sheets. Upon trying to direct the pupils to
follow the structure of the chord sheet, the issue of pulse again became a barrier. Eventually,
I decided to set the class to working in groups on their tables to put the beat and chords
together and to listen to these performances at the end, and circulated the room with the
teacher providing support. The level of noise as four individual groups attempted this made
the task exceptionally difficult, and the performances were ultimately more a case of singing
with raucous and unsynchronised iPad accompaniment, though enthusiasm and engagement
was, at least, a present factor here.
A great deal of disruption was encountered as the class became restless. This was
particularly evident when structured activities became laborious or complicated, such as
establishing a pulse and assigning the chords. Increasingly, the teacher intervened for
classroom management as pupils took to pressing each other’s screens, and other instances of
misbehaviour. It was difficult to ascertain where interruptions came from; at several points,
as one pupil was saying or demonstrating something, it was interrupted by a loud noise

188
coming from an iPad at another part of the room, with the teacher trying to identify the
responsible individual and an argument of denial and accusation ensuing among the pupils.
Other issues were less clearly matters of behaviour; pupils frequently clicked the wrong
button accessing some other part of the app, or turned the iPad off by mistake. This led to
shouts of ‘mine’s gone off!’ and similar protests, which required a lot of maintenance. Due to
these issues, I decided not to use the Sample-It interface for the lesson as planned.
After the lesson, I gave my apologies to the teacher that the session had not run more
smoothly. I explained that, in retrospect, the activities planned may not have been appropriate
to the age level, and the whole-class dynamic had not been supported by the scaffolding
structures in GarageBand as expected. The teacher offered a solution:

Teacher: ‘If I was a teacher and I was going to use that app, I would have probably,
before that, told them about the chords, so it wasn’t new for them…’

I then explained that my intentions for these sessions had been to encourage improvisatory
creative behaviour, and to minimise reliance on explanation.

Teacher: ‘That wasn’t as improvised because you were giving them specific chords to
follow. They weren’t really creating their own music as such… For these, it’s
understanding what the beat is, what the pulse is. That’s very hard for a lot of
children… As a teacher, I’d break it down into smaller parts, then put it together
afterwards’

The teacher then noted that Charanga was useful for breaking down and explaining these
concepts. It was clear that, in this instance, the successful outcomes I had observed with this
approach in previous studies had not been replicated. There were a number of possible factors
which contributed to this:
 The younger age of the group, and consequent change of ability level
 The use of commercial software, with a more complex layout
 The whole-class performance structure
 The reliance on non-exploratory features which had to be accomplished correctly by
everyone, such as maintenance of a pulse and following the chords

189
7.7.4 Context 2

I returned to the school for two afternoon sessions the following week, working with a Year 6
class. Though this group were 3 years older than the previous class, I decided to abandon the
‘iPad Ensemble’ approach, wanting to avoid the same kinds of disruption where possible. To
plan a new session, I requested the music targets for this year group:

Evaluate their work and suggest improvements to the work of others.

Use appropriate musical vocabulary stating what was successful/unsuccessful and
why.
Compare and contrast a variety of music indicating preferences.
Identify different musical devices in a variety of musical genres and show awareness
of the influence and place music has had in society over time.
Use digital technologies to aid with the creation and recording of ideas

I decided to continue to use GarageBand and the song ‘Happy’, but instead to focus on how
digital technology can be used to make a ‘remix’. This would provide an opportunity to
address all of these learning aims, with a particular focus on the last one.

7.7.5 Method 2
Learning Objective: To make a remix of ‘Happy’ using GarageBand

For both of these sessions, I was allocated a room and sent a group of students from the class.
This was a different group of students for each session; 7 took part in the first session, and 10
took part in the second. Data was collected by note taking and from an audio recording,
which was later transcribed. The GarageBand projects were saved, and were later sent to the
class teacher to distribute back to the pupils. These could then be reopened in the app, or
shared on iTunes.
We started by discussing what a remix is, how it involves copying some parts of a
song and changing others. We then listened to ‘Happy’, and discussed which parts we might
keep or change. We then turned to the iPads and GarageBand. I had set up each iPad with a
new GarageBand project, setting the tempo and assigning the correct chords to the smart
keyboard interface. I demonstrated how a looping beat could be written by counting to 4 and

190
choosing drum pads for each number, then ‘filled in’ the gaps by adding drum pads at some
of the half-beat measures (‘1 and 2 and 3 and 4 and…’). I also quickly demonstrated
recording a pattern using the chords and recording a vocal part. For the most of the rest of the
lesson, the pupils worked at their own pace with an iPad and headphones each, supported by
me or supporting each other. As we neared the end of the session, I directed them to the
Sample-It app, and encouraged them to use it in their compositions. We listened to the pieces
and discussed them at the end.

7.7.6 Results 2
These sessions21 proved to be more productive than the previous lesson with Year 3. Levels
of engagement were high, and there was no disruptive behaviour. Because of the small group
sizes, and the less sequential and formulaic structure, a lot of discussion took place
throughout the lesson, all of which focused on the task or programme.
The demonstration at the start of the lesson was met with an unexpected level of
interest; the comments of the pupils indicate that they had not used a sequencing workstation
like GarageBand before, and were not familiar with the kind of results that could be
relatively easily engineered:

‘Did you make that?’

‘Are you an expert at making stuff like that?’

Consequently, the groups were eager to start using the app. In many cases, the exploratory
mode of use led to the pupils happening upon something that intrigued them:

‘It sounds good even if you just click on random ones’ [using the drum pads]
‘Mine sounds weird’ [finding an arpeggiator]
‘It’s a banjo’ [followed by rapid succession of notes, possibly imitating Bluegrass
playing]

This format meant that I could respond to interesting occurrences as they happened.
However, it also meant that a lot of time was spent experimenting with different presets and

21
Due to the similar structure and outcomes of the two sessions, these results will be presented as one account.

191
instruments. Though this engaged the pupils, and led to a lot of discussion, they often had to
be encouraged to record something with these options, or to make any gradual changes which
would cause them to reflect on how a particular sound has been shaped. These directed
instances were often stalled by the interface, which though simpler in layout than audio
workstations such as Logic Pro or Ableton Live, was still fairly advanced for the novice user.
I often had to make quite unclear references to the graphical user interface:

‘Now we go to section B, so we click this button up here to add a new section…’

‘Do you see this button that looks like three squares?’
‘It’s that button that looks like a bunch of lines.’

This led to a lot of uncertainty from the pupils, who frequently asked for help at times:

‘How do you go onto that?’

‘How do you get different sounds?’
‘Something’s happened to mine. It’s gone bad.’

These less productive sections of the lesson were reminiscent of the barriers faced in the
previous Year 3 study, though the open format and smaller group size meant that they were
shorter in duration and less of a barrier to progress.
More positive outcomes emerged as the pupils began to hear their remixes take shape,
often after the drum loops had been recorded. Musical features were discovered through
exploration, and came to the attention of the group through demonstration:

Pupil: ‘It says tempo’ [finding and altering the control for beats-per-minute]
Me: ‘That means speed. You can change the speed from there.’

Rather than stopping the class to demonstrate chords, this too emerged from exploration:

Pupil 1: ‘What are these symbols?’

Pupil 2: ‘The notes?’
Pupil 1: ‘Oh…’ [as if something suddenly became clear]
Me: ‘These tell us a group of notes we can use.’

192
The use of the smart keyboard then led to some comments on the features of different genres.
In this way, the explorations of presets had some reflective significance to the target on
musical devices in genres:

‘I’m being a rocker’ ‘I’m going to be hard rock’

‘I’ve found dubstep’
‘It’s the 80s all over again!’ [when playing back a synthesizer arpeggiator]

Pupils were visibly surprised at possibilities such as the smart guitar interface, allowing them
to ‘strum’ or pick out a guitar pattern; this encouraged more recording. Also, this led to other
new instances of discovery learning:

‘I didn’t know there was such a thing as an electric piano.’

‘What does dry mean?’ [followed by a discussion on dry and wet as a metaphor for
the amount of effects processing; ‘think of the effect like a liquid’]

Though not all of the pupils were able to follow the chords to match the song exactly, the
remix task gave license to experiment and improvise:

Pupil 1: ‘But that’s not going to sound like that… Mine doesn’t sound anything like
‘Happy’’
Pupil 2: ‘But you’re making music’
Pupil 1: ‘I’ve got something good!’ [after a few moments]

This kind of discussion and social interaction was observed throughout the lesson. Far from
being a disruptive influence, it appeared to provide a source of intrinsic motivation, as the
pupils showed each other their pieces, and swapped techniques for getting different results.
There was a lot of good-natured discussion on the respective differences between their pieces,
often with reference to musical dimensions:

‘What songs do you listen to? You must have, like, fifty songs playing at once’
[remarking on the busy texture of their friend’s remix]

193
On occasion, pupils took it upon themselves to demonstrate something to their peers, and
help them to get the same result:

‘I’ll come round and do it for you now… Has everyone got it?’

This emergent sense of organisation was also seen when the pupils decided it would be better
if they all sang and recorded the vocals together:

‘We should all sing it together’ ‘We’ll all sing it at the same time’
‘Everybody get ready to sing’

It appears that, in such instances, the open format of the session generated intrinsic
motivation – to share an idea, to do something together, to demonstrate something – that was
grounded in the possibilities afforded by the technology.
The Sample-It interface, being used later in the session, was met with interest. As we
had more than enough iPads, this could be set up on a second device, allowing pupils to
record from one to another. In a similar fashion to the earlier studies with Graphick Score,
links were pointed out to functions found on social media platforms:

‘Is that like a Snapchat filter? Like, it makes your voice sound like a chipmunk?’

We were then able to discuss how the ‘chipmunk’ effect is created (increasing the pitch,
speeding up) and the consequent prevalence of this effect in electronic dance music. The
pupils were able to cut up, loop and trigger their vocal samples, leading to various
experimentations:

Pupil 1: ‘Listen, this is it high… this is it slow…’

Pupil 2: ‘We sound like minions!’ [referencing the characters from Despicable Me]

One pupil worked out how to imitate the sound of an EDM ‘build-up’, where the number of
repetitions in a cycle increases to build tension. The tilt-filter function was also discovered,
which prompted instances of ‘jamming’ between the pupils, as they tried to stay in time with
each other. This record and edit function, particularly the capacity to reverse audio and
change speed, led to a lot of experimentation with recording and editing different sounds. The

194
pupils were noticeably listening to, and communicating with, one another here, rather than
being solely focused on what they were doing themselves. As such, this section of both
lessons became a very collaborative and reflective activity, which may have benefitted from
further explorations outside of the allocated classroom, looking for interesting sounds. One
pupil, Jonah, made extensive use of the Sample-It interface to edit vocal sounds and record
them into his GarageBand project. This emerged from experimenting with different vocal
noises, with alterations in pitch, and then alternately playing them backwards and forwards
(‘it’s like a helicopter’). This sampling and resampling of improvised vocal sounds was
carefully recorded into GarageBand by Jonah to produce unusual effects which punctuated
his remix at timely intervals.
Toward the end of the second session, the class teacher entered to take some
photographs, and the pupils were keen to show their work. As with the earlier study at School
D, the rapport between the teacher and the pupils offered opportunities for engagement and
reflection:

Teacher: ‘Jonah’s really into technology. He wants to be a YouTuber when he’s

older.’ [as Jonah demonstrated his interest in Sample-It]
Teacher: ‘That’s a Taylor Swift country version.’ [listening to a smart guitar based
remix by another pupil]

We ended with a discussion on how different the remixes were from the original song, but
how this had meant that we created something original of our own. The pupils indicated that
they wanted to continue with these activities:

‘When will we be doing this again?’

‘My mum won’t let me download anything like this on my iPad’
‘Can you get this on tablet… on Samsung?’

They wrote down the name of the app, and I suggested some other sampling apps that they
might use at home. I also told them I would send the GarageBand projects, as they could
continue working on these in the app, or stream to iTunes to show friends and family.
The barriers faced in the ‘iPad Ensemble’ session with Year 3 were largely mitigated
by the open format of the later sessions. As the class teacher pointed out, the first session did
not involve much opportunity for improvisation; there was a formulaic structure and certain

195
materials which had to be learned in a very specific way. My approach of minimising
explanation therefore acted against the flow of the lesson. The later sessions were, no doubt,
improved by the smaller group size and older participants, but were lacking certain other
supportive structures, such as the presence of the teacher. The open format here, offering
more opportunities for improvisation, allowed pupils to work at their own pace, to discover
opportunities, and to support each other. The flexible nature of the activity, and capacity for
independent work, meant that pupils had the option of following the chord structure or
experimenting with different possibilities; this emerged from an inability to follow the
chords, a disinterest in doing so, or simply from mistakes in attempting to do so. Crucially,
this freedom allowed for a differentiation that maintained the flow of the lesson for all pupils,
as they were ‘given musical and technological information as they needed it’ (Pitts & Kwami
2002, p.69).
Demonstration played an important role in clarifying the objective. The pupils of the
later sessions were unexpectedly impressed by my demonstration, which they considered an
expert level. Collins et al. (1991, p.13) note how modelling helps students to ‘observe and
build a conceptual model of the processes that are required to accomplish it’. Similar results
are seen in the research with Graphick Score; the appropriation of features from the
demonstration to individual creative outcomes in each of the studies. As with earlier studies,
pupils listened to sounds and attributed meaning, though this was mostly confined to
reflections on genre. However, with Sample-It, discussions took on much more
impressionistic form similar to that which has been observed with Graphick Score (‘it’s like a
helicopter’). This is perhaps due to the gradual changes to the sounds presented by this
interface, which were less preset-based than the genre-oriented GarageBand. It is also
notable how this interface directed the focus away from screen-based interactions and toward
imaginative discussion and exploration of the sounds made by each other. This occurrence
was unexpected, but it was clear that the pupils were reflecting more on the musical results of
their actions here than when, to paraphrase one of the pupils, just clicking random buttons.
While the range of options in GarageBand promoted engagement, a lot of ‘teaching
the software’ was still required. Even some of the scaffolding tools, such as the smart
keyboard interface, proved challenging for some pupils. In many cases, this seemed to be due
to reliance on musical representations and language that was not intuitive for the pupils. By
contrast, when using the Rainbow Keyboard, the pupils were able to devise kinaesthetic
patterns, and even came up with their own notations (see 7.4.5). It seems that the openness of
format, and the lack of a sense of ‘right and wrong’ ways of completing the task, afforded the

196
pupils the opportunity to come up with their own methods and ‘influence the problem
manipulation space’ (see 4.1). The open format may also assist the generalist teacher by
focusing on the creative activity rather than the concepts involved, as we saw in the study at
School D (see 7.6).

7.8 Theoretical Summary

In my attempts to develop a sandbox composing environment which enables pupils to interact

creatively with the inter-related dimensions of music, certain core concepts have emerged.
This has been grounded in the process of my literature review, formulating and addressing
the research questions (the conclusions to which will be provided in Chapter 8), my
development and testing of Graphick Score, and the application of these considerations to
other bespoke and existing applications (notably utilising the iPad for performance and
composition). We have observed instances of what may be termed sound-painting, sound-
sculpting or sound-sketching (depending upon the mode of interaction involved) whereby the
user interacts directly with musical structures, benefitting from real-time visual and
kinaesthetic (as well as auditory) feedback. Crucially, digital technology offers a means for
transparency between these manifestations of musical ideas; they become objects-to-think-
with, with the potential for embedded meaning (see 3.1). In this way, this approach is an
example of what Fullan and Langworthy (2013, p.4) term ‘new pedagogies… enabled and
accelerated by digital technology’ (Fullan & Langworthy 2013, p.4).
The approach detailed by this theoretical summary relates to Bamberger’s theory of
tune-blocks (see 3.2.4) which has been used extensively in this research (see 6.4). For the
purposes of early-years music education, a familiar musical phrase, as a structure of multiple
notes and durational values, is simpler than the first note in isolation, because the child has a
context for understanding the phrase as a whole. They can recognize it, replicate it in a
variety of possible ways, and possibly modify it for creative and expressive purposes. In this
way, the musical language derived from culture becomes the starting point, the knowledge
that children bring to the learning process. This offers a creative framework within which
students can explore musical ideas with a degree of freedom; we observed the potential
benefits on lesson structure of this freedom in this research (see, for example, 7.7). Where
musical outcomes are intertwined with the modes of interaction and representation employed,
and generally easy to comprehend for both student and teacher, these structures become

197
effective resources for the ‘Bloom’s 21’ model of learning (see 2.2.1). This was observed
with Graphick Score as the user interface developed to the point where it could be used for
experiential music composition, by both students and teacher, with minimal instruction (see
7.6).
This research has highlighted that to ‘compose… using the inter-related dimensions of
music’ (DfE 2013, p.2) is not, in itself, an effective starting point or approach. The reason
why these dimensions are inter-related is because they manifest through larger meaningful
structures. We can explore our own idea of these concepts when we can utilise these
scaffolding structures for creative purposes. The original application of this idea in my
research has been to investigate how digital technology can enable us to utilise such
structures in open-ended ‘sandbox’ environments; to arrange and mould digital embodiments
of inter-connected musical dimensions into something that demands only imagination, or a
willingness to experiment. By moving away from melodic and harmonic ideas and toward a
more open framework of ‘sound objects’, Graphick Score has opened up a new mode of
digital interaction which has had promising results, and which bears further exploration,
especially to examine the musical dimensions which have perhaps been less thoroughly
applied (see 8.3.3). This open-format exploration of sound and attributing of meaning has
also been seen through digital recording and manipulation with Sample-It (see 7.7.6) which
presented further examples of social and kinaesthetic interaction using iPads; this too should
be examined further.

198
 8. Conclusion

Figure 8.1

In the concluding chapter to this document, the outcomes are presented as answers to the
wider questions, and directions for further research are highlighted. I will briefly summarise
the outputs of this project, before answering each of the four research questions, and
discussing the implications and recommendations for future research. I will then critically
evaluate the project, focusing on any limitations and alternative possibilities which have
emerged.

199
8.1 Summary of Outcomes

8.1.1 Practical Outcomes

This research project produced a number of software artefacts through an iterative and
continuous process of classroom application and development. For the purposes of this thesis,
these artefacts are thought of as the research tools; the means by which the research questions
are addressed. However, they have also been shown to lead to instances of creative
engagement and learning in the classroom, and will continue to be developed and utilised
beyond the completion of this project. These artefacts are submitted in the electronic
appendix, which accompanies this thesis, alongside instructions for installation. Updated
versions will be continually available at my Wordpress account:
www.adammatthewhart.wordpress.com

8.1.2 Theoretical Outcomes

The application of grounded theory to the literature review, practice and research, has led to a
central theoretical outcome, describing how abstract musical dimensions may be learned
experientially through digital scaffolding structures in sandbox environments. This is
summarised at the end of the previous chapter (see 7.9) and elaborated upon in the answer to
the research questions (see 8.2). The implications of these findings for new research and
development follow these answers (see 8.2.5). Also, by undertaking a new practice-led
model, this project serves as an examination of how software development may be utilised
for educational research.
Through the methodological approach of SoDaR supported by grounded theory, the
resultant practical and emergent method has reached a point of theoretical saturation (see
Glaser & Strauss 2017, pp.224-8). The various codes emerging from the application of this
method, with respect to our definition of musical learning (see 1.1.1) within digitally-
supported environments, have, through the process of data collection and analysis, developed
into wider concepts and ultimately a summative theory (see 7.8). This emergent process of
theoretical development and saturation is described by Figure 8.2. I propose that this model,
being a newly adapted method for researching creative musical learning through technology
and a theoretical outcome in its own right, is an appropriate approach for future research in
this particular field (see 8.3).

200
Figure 8.2 – The theoretical codes emerging from the studies and analytical process, and their relevance
to the four research questions.

201
8.1.3 Other Outcomes

Besides the practical and theoretical outcomes documented in this thesis, there are other
achievements of this project which can be acknowledged here. Firstly, the research conducted
has had a substantial impact in terms of musical learning and creativity in many cases, as the
evidence documents. Of course, this occurs more prominently in some studies than others,
but all were planned for positive impact, and the findings (even from those where the session
did not go exactly as intended) have informed my future practice and led to the wider
outcomes of this thesis. We have seen impact on musical creativity in the classroom, as well
as influence on teachers and school practices in some cases. Furthermore, this has led to the
definition of new projects, with expressions of interest obtained from some of the schools
who were involved in this research. Finally, this project has so far produced two new
publications (Hart 2017a; 2017b) as well as dissemination at several major international
conferences.

8.2 Research Questions

8.2.1 Question 1: Interactive Digital Technologies and ‘Musical Behaviour’

To what extent have interactive digital technologies hitherto succeeded in encouraging

musical behaviour in the classroom?

It is clear from our survey of existing software resources that many options are available for
classroom use, though the range of software actually used is, by contrast, very limited (see
3.2). Software packages such as Charanga, offering sequential guidance for students and
teachers, help to bridge the skills gap in music education by offering a digital, interactive
curriculum which breaks down concepts into accessible activities. Digital audio workstations
such as GarageBand and Logic are less frequently used in primary education, but more
common in secondary education where pupils are more likely to work on long-term
compositional projects; the construction of a meaningful public entity. Research has
highlighted the scaffolding benefits of music interfaces following a simplified design, such as
eJay and jam2jam. A wider systematic review is requred, on a nationwide level, to examine
the range of technologies utilised by teachers, both specialist and non-specialist, to achieve

202
different musical outcomes. Even the wide-ranging reports of OFSTED are limited in their
attention to this issue, noting ‘insufficient improvements’ (see 2.2.3). These reports pre-date
the relatively recent emergence of Charanga, and increased affordability and implementation
of digital devices such as the touchscreen tablet in recent years, as well as recognition of the
‘new pedagogies’ (Fullan & Langworthy 2014) suggested by such technologies, points to the
current need for a review of digital resources in music education and the wider creative
curriculum.
In terms of the extent to which this question is addressed by this research project, we
must first consider what we have interpreted as ‘real musical behaviour in the classroom’.
Current models for classroom practice focus on creating behaviours, followed by reflection
and synthesis of new knowledge leading to new creative interactions. This is reflected in the
experiential learning cycle (Kolb 1984) as well as more recent proposals such as ‘Bloom’s
21’ (Wright 2012). By promoting these ‘doing dispositions’ in supportive, collaborative and
reflexive environments, we are enabling students to extend their learning and understanding
outside of the classroom, facilitating ‘deep learning’ (Fullan & Langworthy 2014, p.i). Such
creating behaviours are important to musical learning, as the literature review demonstrates
(see 2.2). We therefore must ask whether interactive digital technology has hitherto supported
such behaviours.
One of the benefits of interactive digital technology is the capacity to present abstract
concepts in a visually and tangibly accessible format, as seen in the research of Papert (see
4.1.4; 3.1.1). Where software is not oriented toward a pedagogic design, much time has to be
directed toward learning the software, to the detriment of the learning outcomes. This would
only further exacerbate the problem identified by OFSTED concerning the prevalence of
explanatory behaviours in music lessons (see 2.2.3). The research with GarageBand
highlighted how pupils require more direction if the interface is not designed for this kind of
accessibility, and may require guidance to relate their explorations to creative musical
decisions. By contrast, the research with Graphick Score saw the need for such explanations
decrease as the interface developed toward a more open-ended environment (this will be
addressed in the response to question 2). Though limitations are still present with this
resource, this change indicates the importance of pedagogic interface design in minimising
the time spent ‘learning the software’. However, this must also be balanced against the
stability of professionally developed commercial software, which is often marketed to a
wider demographic, and not built with pedagogic outcomes in mind. This will be addressed in
the answer to question 3.

203
 Early in this research project, the importance of the social element of musical learning
became apparent. Integrating these aims with digital technology has admittedly been a
challenge. In jam2jam, Brown and Dillon address this issue by implementing jamming and
text messaging functions over WiFi networks (see 2.2.3). In GarageBand, users have the
option to ‘share’ the song by sending directly to iTunes (see 3.2.4). In this project, I explored
a range of approaches, such as splitting headphone signals (see 7.4; 7.6), using touchscreen
tablets (see 7.4; 7.7), and even streaming performance commands over wireless networks (see
7.4.2). These were attempted with varying success, though it ultimately became apparent that
considering how the software fits into the wider environment of the classroom, and planning
sessions that incorporated digital technology while not being encapsulated by it, was the most
effective approach. In the best of these examples, the digital technology is seen to accelerate
and augment the socially creative as well as individually creative aspects of learning (see
7.2.3; 7.4.6; 7.6.3; 7.7.6). A focus of future research should be the consideration of how
interactive digital technology can facilitate a range of social musical learning environments,
away from computer screen and headphone based workstations.
It is noteworthy that the digital resources available for classroom music-making are
almost exclusively screen based in their mode of interaction. While touchscreen tablets have
a higher capacity for portability and collaborative learning, making them more versatile
classroom tools in many respects, very little research exists on their suitability for active
musical learning. This research explored some approaches, with positive results where small
groups composed and performed with iPads and accessible music interfaces. However, the
visually-dependent mode of interaction, lack of haptic feedback, and lack of clarity
concerning the source of individual sounds, led to less promising results in larger groups,
with less evidence of listening and communication within the ensemble. To make use of the
scaffolding and processing potential of these tools within a context of real musical behaviour,
we must consider how they might be utilised in a manner that also guides attention away
from the device in our hands and toward the social and sound environments around us. This
will be addressed in the response to question 4.

204
8.2.2 Question 2: Sandbox Environments

What conditions prompt musical learning in ‘sandbox’ environments?

The term ‘sandbox’ is applied in a variety of recreational and educational contexts. Within
the constructivist paradigm, learning and play are linked or, within some interpretations,
indistinguishable from one another (Piaget 1951). This project has offered a specific
interpretation of the term that recognises a model for digitally-augmented, play-based
discovery learning: A sandbox environment is one where the learning materials, whatever
they may be, are at hand, and shaped by the imagination of the learner to construct something
meaningful, through which the intended learning outcomes are ultimately achieved. A key
feature is the use of scaffolding structures as creative triggers, just as moulds are used to
build sandcastles and other recognisable shapes, and combined or manipulated to construct
larger patterns or structures.
Research has demonstrated how sandbox computer games such as Minecraft and Sim
City have been used in the classroom for the creative, interactive and collaborative learning
of abstract topics. In these examples, the blocks, structures and sprites become objects-to-
think-with, infused with the abstract learning concepts in question. In the research with
Graphick Score, the learning materials – the inter-related dimensions of music – are
represented by the shapes and structures arranged and manipulated by the user. A simple
action, such as clicking or drawing in the canvas window, produces not a single note but a
recognisable structure, which becomes the trigger for formation of more complex patterns
and structures. It is the capacity for the learners to ascribe their own meaning to this, to
exercise their creative imagination, reflect on their decisions, and use this as an impetus for
further decisions, which enables them to explore and discover these abstract concepts for
themselves.
Constructivism tells us that interplay between new experience and existing ideas leads
to learning. Supportive structures which enable learners to attribute meaning to their creative
explorations are therefore vital in sandbox environments. In Graphick Score, the addition of
keynote sounds and images, as storyboarding tools, allow the user to construct a scenario,
which they then augment through sound objects. We also see that children are quick to adopt
and alter suggestions, such as those given in demonstrations. For this reason, the possibility
to directly change the objects-to-think-with as a bridge between creative ideas and new

205
occurrences is a valuable feature. These functions can be extended further in Graphick Score;
the possibility to record and edit sound files (harnessing the social and creative engagement
we saw with the Sample-It app) and to further manipulate keynote sounds by directly
interacting with images (perhaps with rotating functions such as those used with sound-
objects) should be explored.
In some cases, the presence of an unexpected change may be required to further
stimulate creative outcomes and challenge existing intentions and goals. In Minecraft and Sim
City, such occurrences are present in the features of the game over which the player has no
control; the actions of autonomous sprites and events, such as changes in weather.
Implementing such wider processes into Graphick Score, perhaps whereby the user discovers
or ‘unlocks’ control over such processes, is another possible avenue for investigation. It may
be that generative interfaces, such as MetroNotes and SoundWorlds, offer a means of
addressing this possibility as this research continues.
Finally, we should not assume that sandbox environments, and the processes
involved, must be limited to digital interfaces. This research has shown that these
environments can be supported by digital technology, but that this is most effective when the
relevant skills can be applied to non-virtual contexts. Commercially-produced digital
resources tend toward a wholly software-oriented mode of interaction, and are consequently
screen-based. Conversely, educational research in this area may consider the wider classroom
environment as part of the sandbox, augmented by the pedagogical instincts of the teacher,
and the creative imagination of the pupils. This was evidenced by the introduction of games
(see 7.2.3; 7.6.3) and collaborative activities (see 7.7.6) in my research. In retrospect,
focusing entirely on computerised playback in Graphick Score missed an opportunity to
challenge creative decisions and stimulate further reflection on the learning materials.
Building upon the game of the class teacher in School D, the graphic score made by one
group could be given to another to be performed with rudimentary instrumentation or voice,
and then compared with the computer playback, generating opportunities for comparison and
reflection on musical interpretations, as well as the realisation of shared conceptual ideas.
This will be explored in further research.

206
8.2.3 Question 3: Constructivism and Design

How can constructivist educational theory be applied to software design as a means of

facilitating creative musical learning?

This question relates not just to the musical learning of the pupils as a result of constructivist
design, but also to my developing practice of interface design, and how constructivism helps
me to understand this progress. Jonassen (2006) argues that constructivism is not a design
strategy, but a lens through which to analyse our methods. If we mistake ontology for
strategy, we risk perceiving outcomes that would have occurred anyway as fruits of design;
this is factual relativism (see 2.1.4). However, we can adopt a design strategy that is informed
by constructivism. The sandbox environment is an example of this, as it presents an
environment for exploring new possibilities while facilitating the meaning-making processes
that constructivism tells us leads to new knowledge. Similarly, the bottom-up design process
acknowledges constructivist theory by emphasising responsiveness to emergent meanings
and possibilities which only become apparent to the developer when engaged in the design
activity.
It has been demonstrated that research and innovation is needed to help classrooms
keep pace with technological change, and harness new learning opportunities (see 2.2.3).
Research in investigative pedagogic interface design, informed by educational theory and
grounded in classroom research, is therefore a worthwhile pursuit. In music, precedents to
this kind of research, where educators use software development as a tool to investigate new
creative pedaogiges, is to be found in the work of Bamberger and Hernandez (2000), as well
as Brown (2007a) and Dillon (2004) among others. This project has investigated a new
practice-led model for combining teaching experience with software design; the benefits of
this approach include a sensitivity to pedagogic considerations and wider learning outcomes
and environments that may now be found in commercial software development, though this
necessarily also demands a recognition of the limitations of the researcher-programmer
compared to that of a professional developer (see 6.2.3). Where areas of skill and experience
are combined in pursuit of a new practical model, this is perhaps inevitable.
In this project, grounded theory was identified as a methodology compatible with the
constructivist ontology (see 4.1.4). This allowed me to apply this lens to the practical
methods of SoDaR, and to critically reflect on my own practice as it occurred. Sensitivity to

207
emergence, emphasised by both Glaser (1992) and Brown (2007a) with respect to these
methodologies, meant that SoDaR was not followed as a sequential formula but adapted to
meet the requirements of emerging learning opportunities. This is particularly seen in the
shift away from melodic scaffolding structures and toward kinaesthetic ‘sound-objects’ (see
7.4; 7.5; 7.6) and the later examination of iPads as performance tools (see 7.7; 7.8). In
addressing these emerging considerations and opportunities, maintaining equilibrium
between these new ideas and those already established, and adapting the trajectory of
research to accommodate this change, I have developed a working interface which will
continue to be applied and updated in response to educational outcomes. This is a reflexive
and dynamic cycle that has seen both the software artefacts of my practice, and my practice
itself, continually improve in response to the opportunities of the music classroom.
Constructivism accounts for understanding; it tells us that this comes only from the
learner, it is their creation, and it is the role of the teacher to provide an environment which
facilitates this process. Similarly, inspiration can be understood as only part of an experiential
process; a function of the creative imagination, which can be exercised. Thinking leads to
more thinking, and creative behaviour leads to more creative behaviour. If we take a creative
exercise that children can already do – telling a story – and use this as a scaffold for what we
want them to learn – composing a piece of music – we have to help them to understand a
creative connection between the materials they are used to and the new materials we are
introducing. In Graphick Score, this led to the development of sound-painting/sound-
sculpting structures which facilitate these creative ‘doing’ behaviours. Inspiration is merely
part of this process, where exploration leads to realisation and further creative exploration.

8.2.4 Question 4: Creative Expression and the Musically Untrained

How might interactive digital technology be harnessed to facilitate creative expression in the
musically untrained?

In many creative subjects, the learning materials are very much ‘at hand’; we sculpt clay,
splash paint, and sketch with pencil and pen. Music is different, and altogether less tangible.
The learning materials, too often, are abstract concepts like the inter-related dimensions of
music, apparent only in the results of creative expression. In this project, I have used

208
interactive digital technology to make these materials altogether more visible, tangible and
accessible. I have examined how structural scaffolds may be made of interconnected musical
concepts, to act as an interactive framework for expressive activities which children
understand from an early age, such as telling a story or painting a picture. The interface of
Graphick Score can make these interactions more musically coherent (see 7.2.3) while also
providing a format for reflection. We have seen that children are quick to attribute meaning to
creative outcomes. Digital technology may be used to engineer transparency between action
and result, as well as the consequent change in musical dimensions; a simple action can
produce a complex result (see 7.4; 7.5; 7.6). Children are therefore more readily able to
engage in creative activities without having to first understand the smaller mechanisms
involved, promoting a more free and exploratory environment in which they can form these
concepts for themselves.
Papert describes how children explore concepts where they can produce a meaningful
public artefact (see 3.1.1). It is notable that not all the creative explorations of children yield
such tangible artefacts. Drawings and paintings are produced from a young age, and parents
will often display these on the fridge, and ask the children what they represent; the children
reflect upon these visual-kinaesthetic creations, attach meaning to them etc. Children also
indulge in sound-based creative expressions, make up or alter songs, or something as
rudimentary as exploring unusual sounds. However, we less frequently make artefacts of this
due to the more ephemeral nature of these activities. Without a piece of paper or a sculpture
to act as a record that these activities took place, they often go unnoticed and uncommented
upon, and are consequently less of a focus for reflection. It is therefore perhaps no surprise
that visual-kinaesthetic interactions, as seen in the experimentations with Graphick Score,
provide a useful scaffold for attributing meaning to sound. Therefore, we may ask how digital
technological can promote creative and imaginative attitudes to sound, in terms of both
making and listening.
Though the benefits of making a sound-artefact were explored through Graphick
Score, it was admittedly a challenge to make these screen-based interactions part of a social
learning dynamic. This was achieved in some respect through group work and discussion.
When using iPads, a different dynamic was often achieved, due to the more kinaesthetic and
portative modes of engagement. However, the ensemble musicianship and group work seen
in study 7.4 could ultimately not be replicated with a larger group (although the unsuitability
of the task to sandbox learning no doubt played a factor here). One of the most prominent
examples of social creative expression through digital technology within this research was

209
seen when pupils edited sounds using the Sample-It interface. This led to an unexpected
degree of listening and creative exploration, even ‘jamming’ interactions (see 7.7). We also
saw how this made a creative and changing artefact of actions such as making sounds with
voice; the kinds of rudimentary, or ‘childish’, actions which would typically have little
significance to music curriculum targets. In this case, by manipulating and sequencing these
sounds, the pupil in question did something of significant creative expression, engaging with
changing musical dimensions through voice and digital technology (see 7.7.6).
The findings from Sample-It, though brief, present an opportunity to use digital
technology to move away from screen based interaction and turn focus to the sounds around
us, or those which we are capable of making. For example, how can we capture and change
these sounds to express a story? These then become sound objects to think with, much like
the scaffolding structures employed in Graphick Score. Changes to these sounds can also be
made with non-screen based functions, such as tilt, while learning can extend outside of the
classroom, into the playground or at home. I therefore feel that this presents an application of
digital technology which has tremendous potential for music education; to get children
interested in sound, to listen with imagination, change sound to express creative ideas, and
reflect on this process.

8.2.5 Recommendations
This research has implications for the design of new technologies, but also for curriculum
change. We took, as our design strategy, ‘composition with the inter-related dimensions of
music’. However, these ideas only arise from creative efforts, in how they combine to form
complex musical structures, with the potential for embedded meaning. They are inherently
inter-related, and we therefore cannot simply ‘compose with’ such concepts. It is therefore
illogical to teach these concepts in the isolated abstract. Garnett (2013) argues that these are
the components of a constructivist curriculum which are too often taught in a behaviourist
way. The 1992 curriculum framed these ideas within a range of experiential strategies, while
1999 revision focused on mood, meaning and feelings as a way of engaging with these
concepts (see 2.2.1). As of 2013, these concepts are presented in the isolated abstract. This is
problematic, as these musical dimensions only emerge from creative and practical musical
experience, and are only ‘used’ in the sense that they are present in the results of our creative
efforts. By themselves, they are arguably meaningless, from a perspective of musical activity.

210
This suggests that instructional design in music education should focus on making
opportunities for creative musical experience available. In this project, I have explored how
this can be made accessible through interactive digital technologies.
Digital resources like Charanga have been useful in breaking down concepts for
pupil, and in supporting generalist teachers who may otherwise struggle to teach an active
music curriculum (see 2.2). The findings of this research suggest that musical experiences
can be made accessible without prior engagement of musical concepts, through digitally-
augmented sandbox environments. Further research should be undertaken on how the
recording and editing of sounds with simple digital apps can provide an accessible,
exploratory and experiential format for engaging with musical concepts. This is akin to the
traditions of tape music; digital technology brings these possibilities more readily into our
classrooms, as evidenced by the simple handheld interface of Sample-It (see 7.7). Graphick
Score will continue to be developed to accommodate this. Also, it should be noted that this is
an open-source and freeware application, and I encourage these findings being used for
further non-commercial development for educational purposes. I hope that further practice-
led projects such as this can forge closer links between research and practice in this area, to
make a positive impact on promoting music in the classroom and at home.

8.3 Critical Evaluation

8.3.1 Scope

This has been an exploratory research project, and as such, has been characterised by an
emerging focus on some areas while others might bear closer observation. In particular, more
research is required to investigate how gender, age and other factors impact creative use of
digital technology. As indicated, the findings generated may have significance in other areas
beyond music composition, presenting other avenues which could have been explored in this
project, but certainly should be considered going forward. This includes the potential benefits
of these creative experiential approaches to digital technology in music therapy, social
inclusion and assistive musical environments for pupils with special educational needs. The
principle of sound-painting or –sculpting sandbox environments may also be applied in a
cross-curricular context, by forging links between visual-kinaesthetic and auditory arts. The

211
general findings and hypotheses of this research could be directed toward any of these further
investigations.

8.3.2 Methodology

The wider methodological model of this project was grounded theory; this informed the
research aims and questions, directed the data collection and analysis process, and led to the
emergence of a theoretical outcome. This encompassing application is favoured by Glaser
(1992, p.32) when he argues that ‘grounded theory must be free from the idea of working on
someone else’s work or problems’. A secondary model was provided by SoDaR, though the
step-by-step process outlined by Brown underwent some changes due to sensitivity to
emergence and unexpected outcome, an attitude whose importance is emphasised in both of
the methodologies applied.
This approach, perhaps inevitably, led to some changes in direction. Early data
collection methods, such as the use of surveys, were abandoned as the nature of the studies
became more resembling of music lessons, though ultimately this led to more opportunities
for evidence of creative learning. Other qualitative methods might have been utilised more
consistently and exhaustively, in hindsight. For example, the interviewing of teachers, though
only used on several brief occasions (notably in 7.8) proved a useful resource, the
significance of these perspectives only becoming apparent as the focus moved away from the
interface itself and toward the mode of application. When building upon these findings, I
intend to make much more prominent use of teacher perspectives.
A potential criticism of the methodological approach is that it relies too heavily on my
own practice and perspective as a musician, programmer and teacher, especially considering
that my experience across these three areas is not in equal proportion. However, one of the
advantages of the methodology used in this research project is that, as the researcher-
programmer, I am in possession of the domain knowledge, the evolving research
considerations and the means to pursue a software solution. This enabled me to take an
exploratory and experiential approach, being able to respond directly to findings with
immediate and appropriate changes in my programming practice. Schön (2017) argues that
practitioners respond to problems and adapt their practice in real-time, a theory supported by
Kolb’s (1984) experiential learning model. The key, then, is to adopt an approach which
allows the practitioner to step back and view the process, to be able to document and analyse

212
it as a component of research. By using grounded theory, and by viewing my own developing
practice through the constructivist lens, I have sought to document this process as part of this
project.
In addition to the principal practical and theoretical outcomes, this research has
provided a case study example of SoDaR, which, as a new methodology, bears further
experimentation. When applying Brown’s methods, and asking his suggested questions, the
resulting model was actually much less structurally formulaic, with reflexive and cyclical
characteristics. These changes are, in part, due to the combination with grounded theory, but
it is this mixed methodology approach that has allowed a clear commentary upon the process.
Crucially, the methodology is reproducible (Glaser 1992, pp.116-7); the same software
artefact may be used, and the context of each study may be replicated (as close as possible)
with a similar group. Certain variables may be altered for comparison, such as the age of the
group, changes to the software, and different approaches to lesson structure and activity.
Conversely, some variables may not be accounted for, such as unexpected outcomes from
pupils, and the question of how effective or ineffective I might have been in providing
support, or leading the lesson. Though I have attempted to clearly document and analyse any
significant occurrences of this nature through audio transcription, it may be that more
exhaustive methods of documentation and analysis are required. In particular, video
recording of the lessons may be implemented, and transcription may account for reactions
and other behaviours besides speech; specifically, I feel that the system of qualitative
presentation proposed by Roth (2017, p.11) could address this deficiency in future research.

8.3.3 Design and Development

By developing this myself, I have faced advantages and disadvantages. The design process
has been more thoroughly embedded in the other considerations of research, such as the
literature review, methodology and testing. Another advantage is that this process has
allowed me to develop my practice, with substantial improvements in my programming
ability, and in recognising opportunities for learning and creative expression in this practice.
These advantages are noted by Brown is his description of SoDaR (see 4.3.2). However, this
approach also presented disadvantages, and there are certain functions which perhaps should
have been explored differently.

213
 Graphick Score itself has presented some original modes of interaction, yielding
promising results with respect to creative pedagogy. The core design features will continue to
be developed, while perhaps addressing some limitations. I feel that the drawing function,
being an original feature, should remain, but perhaps this should be achieved by generating
single complex geos rather than multiple individual ones. This would save processing power,
but it should be ensured that the same flexibility and malleability of the resulting sound
object is maintained. I also feel that this function could relate to shaping other musical
parameters, with promising results; the idea of visualising, interacting with musical ideas
presented by sound painting need not be restricted to making note events. There are certain
restrictions to the 4-bar grid structure utilised in later versions; we saw this when pupils asked
whether the speed of the canvas window could be varied (7.2.3). The aforementioned
function of ‘drawing’ envelopes for dimensions such as tempo could resolve this issue.
Perhaps the non-linear and real-time responsive designs of MetroNotes and SoundWorlds
present a more open-ended interface. These possibilities will be reviewed in future research.
I feel that the domain knowledge of musical dimensions was well represented in
Graphick Score. Some of these functions were in place from early versions (see 5.3) while
others emerged following later refinement. Dimensions such as timbre, the shape and
character of a sound, as well as dynamic variation, emerged from the implementation of
sound-objects, consisting of multiple geos, as single sound events (see 7.5). This enables
users to see these meaningful structures – manifestations of interrelated dimensions – rather
than thinking in terms of single musical dimensions. Though the polarity of major and minor
tonality was a useful concept, I feel that this could be extended further. In particular, I am
keen to reintroduce non-Western scales such as the Indonesian slendro. Perhaps such
functions could be ‘unlocked’ through use (see 8.2.2).
Velocity and size seemed to present an intuitive connection during initial design (see
5.3), but this ultimately generated a conflict with the concept of duration, as the size of a
shape is related to its length. I incorporated certain solutions, such as drawing a long line with
the draw function, and longer joined structures of sound-objects. Early studies indicated the
need to avoid, where possible, convoluted navigation of menus to change note events already
created (see 6.2). It may be that icons of different tools could be used: For example, painting
a line can create a note of certain length, while moving up and down produces a a
portamento. A different tool, such as a stamp or felt pen, could be used for strings of single
note events. An alternative to velocity is found in many digital audio workstations, which

214
tend to use changes in colour to indicate dynamics (see 3.2.4). However, this presents an
issue with my design, as overlapping textures of different voices would become unclear.
Pd proved an appropriate environment to use for this project. The deprecation of Pd-
extended (see 4.2.1) proved an issue, but Graphick Score has now been successfully
transferred to Pd-vanilla for submission and future development. A new alternative version
of Pd, with an extended library and user friendly interface, Purr Data, is now also available.
This raises the question of whether Graphick Score could be transferred to this platform for
development by less experienced programmers. In the spirit of open-source programming, I
would like to continue to work on Graphick Score in such a way that it invites development
as well as utilisation by others22.

8.3.4 Application

The process of development was augmented and accelerated by application in the classroom.
The most valuable of these sessions were conducted during phase 3, where the focus shifted
away from the design of the interface and toward learning objectives and outcomes, as
promising developments were defined by measurable added value. I do feel that this approach
could have been applied earlier, perhaps by testing the prototype interface for smaller
activities. However, this is an exploratory project, and the significance of such an approach
only emerged through the application of my methodology, and the findings which emerged in
phase 2. It is for this reason that phase 3 constitutes the most substantial presentation of
results in this document. This also presented an opportunity to examine further applications
of interactive digital technology, such as the touchscreen tablet.
In all cases, I delivered and lead these sessions, as was necessitated by the
developmental model. To have teachers use these prototype interfaces was never an aim of
this project, due to the focus on development, refinement and application of bespoke
technology, and application of comparative technology. I am keen, however, to address the
consideration of class teachers using these technologies in future research.

22
As of July 2018, ongoing development of the interface, for educational and research purposes, has been
transferred to the game development platform Unity. Project files for iOS and Android platforms will be made
available from www.adammatthewhart.wordpress.com for the purposes of free non-commercial use and
development.

215
8.3.5 Originality

This project has yielded an original composing environment, as well as a series of variations
and additional resources. The mode of interaction employed from the early stages of
Graphick Score, sound-painting (see 5.2.4) is both original and highly suited to pedagogical
creative expression. Some analogues to this mode of interaction exist (see 3.2.4) although this
is the first instance where this kind visual-kinaesthetic mode of drawing and manipulating
gestures to produce graphic scores has been observed (see 7.2.3). In some respects, the non-
linear approach, similar to canvas painting or sketching, demands other modes of playback
beyond the standard left-to-right format. For this reason, variations such as MetroNotes and
SoundWorlds will be revisited in future research. As these presented their own issues (see
6.6) I must consider how these designs could now be implemented within the developed
framework of Graphick Score. It may be that multiple modes of playback are required.
The concept of tune-blocks is drawn from Bamberger (1996; 2000; see 3.2.4). I built
upon this idea by moving away from the idea of melodic composition and toward
impressionistic sound objects, which can be manipulated to tell a story. In a sense, this
approach is partly rooted in the experimental electroacoustic music tradition of soundscapes
and tape music. Applying these forms to the classroom, and recognising the pedagogical
value, has a precedent in the work of Paynter and Savage (see 2.2). My application opens up
new questions about how these musical forms might be suited to the classroom, and argues
that the curriculum could further accommodate these forms (see 8.2.5).
Finally, I have offered a constructivist perspective on musical learning with digital
technology, but also the practice of developing such technologies as a research project. As an
educational community, we are still looking for approaches which are informed by these
ideas to augment our practice, not because of slavish adherence to an ideology, but because
the lens of constructivism can bring to light learning processes that may otherwise go
unobserved. One of the outcomes of this research is that I have gained an improved
understanding of what this ontology tells us about learning, and how this can support
practice. This has included an application and examination of Brown’s methodology (see
4.1), examining the role of the researcher/teacher/developer. I hope that this model can offer
some insight to work in future projects of this atypical yet valuable form of research.

216
 Bibliography

Ackermann, E. (2001). Piaget’s constructivism, Papert’s constructionism: What’s the

difference? Future of learning group publication, 5(3), 438.

Anderson, T. and C. Smith (1996). “Composability”: widening participation in music making

for people with disabilities via music software and controller solutions. Proceedings of the

second annual ACM conference on Assistive technologies, ACM.

Arsenault, D. (2008). Guitar Hero:" Not like playing guitar at all"?. Loading..., 2(2).

Backchannel, T. L. (2017 March 22) “A professor built an AI teaching assistant for his

courses — and it could shape the future of education”. Retrieved 03 April 2017 from

www.businessinsider.com/a-professor-built-an-ai-teaching-assistant-for-his-courses-and-it-

could-shape-the-future-of-education-2017-3?IR=T

Bamberger, J. (1996). Turning music theory on its ear Do we hear what we see; Do we see

what we say?. International Journal of Computers for Mathematical Learning, 1(1), 33-55.

Bamberger, J. (2010). "Noting time." Min-Ad: Israel studies in musicology online 8(1&2).

Bamberger, J. (2003). Music as embodied mathematics: A study of a mutually informing

affinity. International Journal of Computers for Mathematical Learning, 8(2), 123-160.

Bamberger, J. S., & Hernandez, A. (2000). Developing musical intuitions: A project-based

introduction to making and understanding music. Oxford University Press, USA.

Barrett, M. (2006). "‘Creative collaboration’: an ‘eminence’ study of teaching and learning in

music composition." Psychology of Music 34(2): 195-218.

Barrett, E., & Bolt, B. (Eds.). (2014). Practice as research: Approaches to creative arts

enquiry. Ib Tauris.

Bassey, M. (1998). Fuzzy generalisation: An approach to building educational theory.

Retrieved 17 November 2015 from www.leeds.ac.uk/educol/documents/000000801.htm.

217
Bassey, M. (2001). A solution to the problem of generalisation in educational research: Fuzzy

prediction. Oxford Review of Education, 27(1), 5-22.

Bean, A., Siddiqi, S., Chowdhury, A., Whited, B., Shaer, O., & Jacob, R. J. (2008, February).

Marble track audio manipulator (MTAM): a tangible user interface for audio composition. In

Proceedings of the 2nd international conference on Tangible and embedded interaction (pp.

27-30). ACM.

Beckstead, D. (2001). "Will Technology Transform Music Education?" Music Educators

Journal 87(6): 44-49.

BERA (British Educational Research Association) (2014). Revised ethical guidelines for

educational research (2014). British Educational Research Association. Retrieved 19

December 2014 from

content.yudu.com/Library/A2xnp5/Bera/resources/index.htm?referrerUrl=http://free.yudu.co

m/item/details/2023387/Bera

Berkley, R. (2004). "Teaching composing as creative problem solving: Conceptualising

composing pedagogy." British Journal of Music Education 21(03): 239-263.

Bevilacqua, F., Guédy, F., Schnell, N., Fléty, E., & Leroy, N. (2007, June). Wireless sensor

interface and gesture-follower for music pedagogy. In Proceedings of the 7th international

conference on New interfaces for musical expression (pp. 124-129). ACM.

Biasutti, M. (2010). Investigating trainee music teachers' beliefs on musical abilities and

learning: a quantitative study. Music Education Research, 12(1), 47-69.

Biasutti, M., Hennessy, S., & de Vugt-Jansen, E. (2015). Confidence development in non-

music specialist trainee primary teachers after an intensive programme. British Journal of

Music Education, 32(2), 143-161.

Blaine, T. and S. Fels (2003). "Collaborative musical experiences for novices." Journal of

new music research 32(4): 411-428.

218
Blaine, T. and T. Perkis (2000). The Jam-O-Drum interactive music system: a study in

interaction design. Proceedings of the 3rd conference on Designing interactive systems:

processes, practices, methods, and techniques, ACM.

Boghossian, P. (2007). Fear of knowledge: Against relativism and constructivism. Clarendon

Press.

Bolden, B. (2009). "Teaching composing in secondary school: a case study analysis." British

Journal of Music Education 26(02): 137-152.

Borgo, D. (2007). "Free jazz in the classroom: An ecological approach to music education."

Jazz perspectives 1(1): 61-88.

Bos, N. (2001). "What do game designers know about scaffolding? Borrowing SimCity

design principles for education." Ann Arbor, MI: University of Michigan and CILT,

PlaySpace Project, documento digital en http://playspace. concord.org/papers.html# report.

Bos, B., Wilder, L., Cook, M., & O'Donnell, R. (2014). Learning mathematics through

Minecraft. Teaching Children Mathematics, 21(1), 56-59.

Bower, D. N. (2008). Constructivism in music education technology: Creating an

environment for choral composition in the fourth and fifth grades. New York University.

Brand, J. and S. Kinash (2013). "Crafting minds in minecraft." Education Technology

Solutions, 55, pp.56-58.

Breckenridge, J., Jones, D., Elliott, I., & Nicol, M. (2012). Choosing a methodological path:

Reflections on the constructivist turn. Grounded Theory Review, 11(1), 64-71.

Bridges, D. (2003). A philosopher in the classroom. Educational Action Research, 11(2),

181-196.

Brophy, T. S. (1996). "Building Music Literacy with Guided Composition." Music Educators

Journal 83(3): 15-18.

219
Brown, A. R. (2007a). Software development as music education research. International

Journal of Education & the Arts, 8(6), 1-14.

Brown, A. R. (2007b). Computers in music education: Amplifying musicality, Routledge.

Brown, A. R. (2012). "Experience design and interactive software in music education

research." Visions of Research in Music Education 20.

Brown, J. K. (2008). "Student-Centered Instruction: Involving Students in Their Own

Education." Music Educators Journal 94(5): 30-35.

Bruner, J. S. (1961). "The act of discovery." Harvard educational review.

Burnard, P. (2000). "How children ascribe meaning to improvisation and composition:

rethinking pedagogy in music education." Music Education Research 2(1): 7-23.

Burnard, P. (2012). Musical creativities in practice. Oxford University Press.

Burnard, P. and B. A. Younker (2002). "Mapping pathways: Fostering creativity in

composition." Music Education Research 4(2): 245-261.

Cain, T. (2004). "Theory, technology and the music curriculum." British Journal of Music

Education 21(02): 215-221.

Carr, W., & Kemmis, S. (2003). Becoming critical: education knowledge and action research.

Routledge.

Catala, A., Jaen, J., Martinez-Villaronga, A. A., & Mocholi, J. A. (2011, July). AGORAS:

Exploring creative learning on tangible user interfaces. In Computer Software and

Applications Conference (COMPSAC), 2011 IEEE 35th Annual (pp. 326-335). IEEE.

Charmaz, K. (2000). Grounded Theory: Objectivist and Constructivist Methods. In Norman

K. Denzin & Yvonna S. Lincoln (Eds.), Handbook of Qualitative Research, 2nd edition

(pp.509-535). Thousand Oaks, Ca.: Sage.

Charmaz, K. (2014). Constructing grounded theory. Sage.

Churches, A. (2008). Bloom's taxonomy blooms digitally. Tech & Learning, 1, 1-6.

220
Churches, A. (2009). Bloom’s digital taxonomy. Educational Origami, 4.

Churchill, D. (2009). Educational applications of Web 2.0: Using blogs to support teaching

and learning. British journal of educational technology, 40(1), 179-183.

Clarke, B., Svanaes, S., & Zimmermann, S. (2013). One-to-one tablets in secondary schools:

an evaluation study. Tablets for schools.

Clarke, B., & Svanaes, S. (2014). An updated literature review on the use of tablets in

education. Tablets for Schools. UK: Family Kids & Youth.

Clarke, B., & Svanaes, S. (2015). Updated review of the global use of mobile technology in

education. Techknowledge for Schools, London.

Cobb, P. (1996). Where is the mind? A coordination of sociocultural and cognitive

constructivist perspectives. Constructivism: Theory, perspectives and practice. C. T. Fosnot.

New York, NY, Teachers College Press, Columbia University: 39-57.

Collins, D. (1992). "Creativity and special needs: a suggested framework for technology

applications." British Journal of Music Education 9(02): 103-110.

Collins, A., Brown, J. S., & Newman, S. E. (1988). Cognitive apprenticeship: Teaching the

craft of reading, writing and mathematics. Thinking: The Journal of Philosophy for Children,

8(1), 2-10.

Collins, A., Brown, J. S., & Holum, A. (1991). Cognitive apprenticeship: Making thinking

visible. American educator, 15(3), 6-11.

Cousin, G. (2016). Reflexivity: the new reflective practice. International Journal of Practice-

based Learning in Health and Social Care, 1(2), 3-7.

Costanza, E., Shelley, S. B., & Robinson, J. (2003). Introducing audio d-touch: A tangible

user interface for music composition and performance.

Criswell, C. (2011). Technology on the Horizon. Teaching Music, 18(5), 30.

221
Crow, B. (2006). "Musical creativity and the new technology." Music Education Research

8(01): 121-130.

Cunningham, D., & Duffy, T. (1996). Constructivism: Implications for the design and

delivery of instruction. Handbook of research for educational communications and

technology, 51, 170-198.

Custodero, L. A. (2002). "Seeking challenge, finding skill: Flow experience and music

education." Arts Education Policy Review 103(3): 3-9.

Csikszentmihalyi, M. (1996). Flow and the psychology of discovery and invention. New

York: Harper Collins.

Daubney, A., & Mackrill, D. (2013). Music technologies in education–playing the home

advantage. Music Education UK and Music Education Asia magazines.

DeLorenzo, L. C. (1989). "A Field Study of Sixth-Grade Students' Creative Music Problem-

Solving Processes." Journal of Research in Music Education 37(3): 188-200.

Denis, G. and P. Jouvelot (2004). Building the case for video games in music education.

Second International Computer Game and Technology Workshop.

De Corte, E. (1990). Towards powerful learning environments for the acquisition of problem-

solving skills. European Journal of Psychology of Education, 5(1), 5-19.

DfE (Department for Education) (2013) Music programmes of study: key stages 1 and 2.

National Curriculum in England. Retrieved 07 September 2014 from

www.gov.uk/government/uploads/system/uploads/attachment_data/file/239037/PRIMARY_n

ational_curriculum_-_Music.pdf

DfE (Department for Education) and DCMS (Department for Culture, Media and Sport)

(2011) The Importance of Music: A national plan for music education. Retrieved 09

September 2016 from

222
www.gov.uk/government/uploads/system/uploads/attachment_data/file/180973/DFE-00086-

2011.pdf

DfEE (Department for Education and Employment) (1999) The National Curriculum.

Handbook for primary teachers in England. Key stages 1 and 2. Retrieved 07 September 2014

from dera.ioe.ac.uk/18150/7/QCA-99-457_Redacted.pdf

Department for Education and Skills (1992) Music in the National Curriculum, England.

Retrieved 07 September 2014 from http://www.jsavage.org.uk/jsorg/wp-

content/uploads/2013/12/Music_NC-_1992.pdf

De Vries, P. (2013). Generalist teachers' self-efficacy in primary school music teaching.

Music Education Research, 15(4), 375-391.

Dewey, J. (1916). Education and democracy. New York: Macmillan.

Dewey, J. (1938). Experience and Education. New York. Touchstone, 18.

Dillon, S. (2009). "Examining meaningful engagement: Musicology and virtual music

making environments." Proceedings of the Musicological Society of Australasia, Newcastle

upon Tyne: Cambridge Scholars Publishing: 297-310.

Dillon, S. (2004). jam2jam-Meaningful music making with computers. Artistic Practice as

Research: Proceedings of the XXVth Annual Conference, Australian Association for

Research in Music Education.

Dillon, T. (2003). "Collaborating and creating on music technologies." International Journal

of Educational Research 39(8): 893-897.

Downes, S. (2008). Places to go: Connectivism & connective knowledge. Innovate: Journal

of Online Education, 5(1), 6.

Downes, S. (2012). Connectivism and connective knowledge. Essays on meaning and

learning networks. National Research Council Canada.

223
Duffy, P. (2008). Engaging the YouTube Google-eyed generation: Strategies for using Web

2.0 in teaching and learning. Electronic Journal of E-learning, 6(2), 119-130.

Eaglestone, B., Ford, N., Nuhn, R., Moore, A., & Brown, G. (2001, November). Composition

systems requirements for creativity: what research methodology. In Proc. MOSART

Workshop (pp. 7-16).

Ekaputra, G., et al. (2013). "Minecraft: A Game as an Education and Scientific Learning

Tool." Information Systems 2: 4.

English, A. R. (2013) Discontinuity in learning: Dewey, Herbart and education as

transformation. New York: Cambridge University Press.

Farbood, M. M., Pasztor, E., & Jennings, K. (2004). Hyperscore: a graphical sketchpad for

novice composers. IEEE Computer Graphics and Applications, 24(1), 50-54.

Farbood, M. M., Kaufman, H., & Jennings, K. (2007). Composing with Hyperscore: an

Intuitive Interface for Visualizing Musical Structure. In ICMC.

Fels, S., Gadd, A., & Mulder, A. (2002). Mapping transparency through metaphor: towards

more expressive musical instruments. Organised Sound, 7(2), 109-126.

Fischman, R. (2013). "A Manual Actions Expressive System (MAES)." Organised Sound

10(3): 328-345.

Fober, D., Letz, S., Orlarey, Y., Askenfelt, A., Hansen, K. F., & Schoonderwaldt, E. (2004).

IMUTUS: An interactive music tuition system. In the Sound and Music Computing

Conference (SMC 04), October 20-22, 2004, IRCAM, Paris, France (pp. 97-103).

Fober, D., Letz, S., & Orlarey, Y. (2007). VEMUS-Feedback and groupware technologies for

music instrument learning. In Proceedings of the 4th Sound and Music Computing

Conference SMC (Vol. 7, pp. 117-123). Retrieved 04 November 2014 from smc.afim-

asso.org/smc07/SMC07%20Proceedings/SMC07%20Paper%2018.pdf

224
Folkestad, G., Hargreaves, D. J., & Lindström, B. (1998). Compositional strategies in

computer-based music-making. British Journal of Music Education, 15(1), 83-97.

Foreman‐Peck, L., & Murray, J. (2008). Action research and policy. Journal of Philosophy of

Education, 42(s1), 145-163.

Fosnot, C. T., & Perry, R. S. (1996). Constructivism: A psychological theory of learning.

Constructivism: Theory, perspectives, and practice, 2, 8-33.

Fox, R. (2001). "Constructivism examined." Oxford review of education 27(1): 23-35.

Freeman, J., et al. (2011). "Soundscape composition and field recording as a platform for

collaborative creativity." Organised Sound 16(03): 272-281.

Fullan, M., & Langworthy, M. (2013). Towards a new end: New pedagogies for deep

learning. Seattle: Creative Commons.

Fullan, M., & Langworthy, M. (2014). A rich seam: How new pedagogies find deep learning.

MaRS Discovery District. Retrieved 03 November 2015 from michaelfullan.ca/wp-

content/uploads/2014/01/3897.Rich_Seam_web.pdf

Gaber, J. (2007). "Simulating planning simcity as a pedagogical tool." Journal of Planning

Education and Research 27(2): 113-121.

Gadd, A. and S. Fels (2002). MetaMuse: metaphors for expressive instruments. Proceedings

of the 2002 conference on New interfaces for musical expression, National University of

Singapore.

Gall, M. and N. Breeze (2005). "Music composition lessons: the multimodal affordances of

technology." Educational Review 57(4): 415-433.

Gall, M. and N. Breeze (2008). "Music and eJay: An opportunity for creative collaborations

in the classroom." International Journal of Educational Research 47(1): 27-40.

Garnett, J. (2013). Beyond a constructivist curriculum: a critique of competing paradigms in

music education. British Journal of Music Education, 30(2), 161-175.

225
Garvis, S. (2013). Beginning generalist teacher self-efficacy for music compared with maths

and English. British Journal of Music Education, 30(1), 85-101.

Glaser, B. G. (1992). Basics of grounded theory analysis: Emergence vs. forcing. Sociology

Press.

Glaser, B. G. (2002). Constructivist grounded theory?. In Forum qualitative

sozialforschung/forum: Qualitative social research (Vol. 3, No. 3). Retrieved 17 October

2016 from www.qualitative-research.net/index.php/fqs/article/viewArticle/825/1792

Glaser, B. G., & Strauss, A. L. (2017). Discovery of grounded theory: Strategies for

qualitative research. Routledge.

Gouzouasis, P. and D. Bakan (2011). "The future of music making and music education in a

transformative digital world." The University of Melbourne refereed e-journal 2.

Greenhow, C., Robelia, B., & Hughes, J. E. (2009). Learning, teaching, and scholarship in a

digital age: Web 2.0 and classroom research: What path should we take now?. Educational

researcher, 38(4), 246-259.

Hallam, S., Burnard, P., Robertson, A., Saleh, C., Davies, V., Rogers, L., & Kokatsaki, D.

(2009). Trainee primary-school teachers’ perceptions of their effectiveness in teaching music.

Music Education Research, 11(2), 221-240.

Hart, A. (2017a). Towards an effective freeware resource for music composition in the

primary classroom. London Review of Education, 15(3), 407-424.

Hart, A. (2017b). “The Tablet as a Classroom Musical Instrument", Athens: ATINER'S

Conference Paper Series, No: EDU2017-2254. Retrieved 07 November 2017 from

www.atiner.gr/papers/EDU2017-2254.pdf

Hattie, J. A. (2009). Visible learning: A synthesis of 800+ meta-analyses on achievement.

Abingdon: Routledge.

Hattie, J. (2012). Visible learning for teachers: Maximizing impact on learning. Routledge.

226
Hattie, J., & Yates, G. (2013a). Visible learning and the science of how we learn. Routledge.

Hattie, J., & Yates, G. (2013b). Plenary 4-understanding learning: lessons for learning,

teaching and research. Retrieved 14 March 2016 from

research.acer.edu.au/cgi/viewcontent.cgi?article=1207&context=research_conference

Helle, L., Tynjälä, P., & Olkinuora, E. (2006). Project-based learning in post-secondary

education–theory, practice and rubber sling shots. Higher Education, 51(2), 287-314.

Hennessy, S. (2000). Overcoming the red-feeling: The development of confidence to teach

music in primary school amongst student teachers. British Journal of Music Education, 17(2),

183-196.

Hewitt, A. (2002). "A comparative analysis of process and product with specialist and

generalist pre-service teachers involved in a group composition activity." Music Education

Research 4(1): 25-36.

Hewitt, A. (2008). "Children's creative collaboration during a computer-based music task."

International Journal of Educational Research 47(1): 11-26.

Hewitt, A. (2009). "Some features of children's composing in a computer-based environment:

the influence of age, task familiarity and formal instrumental music instruction." Journal of

Music, Technology & Education 2(1): 5-24.

Higgins, W. (1992). Technology. Handbook of research on music teaching and learning: A

project of the Music Educators National Conference. R. Colwell.

Hodges, R. (1996). The New Technology. Music Education: Trends and Issues. C.

Plummeridge. London, Institute of Education.

Holt-Reynolds, D. (2000). "What does the teacher do?: Constructivist pedagogies and

prospective teachers’ beliefs about the role of a teacher." Teaching and teacher education

16(1): 21-32.

227
Honebein, P. C. (1996). "Seven goals for the design of constructivist learning environments."

Constructivist learning environments: Case studies in instructional design: 11-24.

Ip, H. H. S., Law, K. C., & Kwong, B. (2005, January). Cyber composer: Hand gesture-

driven intelligent music composition and generation. In Multimedia Modelling Conference,

2005. MMM 2005. Proceedings of the 11th International (pp. 46-52). IEEE.

James, D. and J. Stanton (2011). Beyond being (t) here: the social and personal implications

of making music at a distance. Proceedings of the 2011 iConference, ACM.

Jennings, K. (2005). "Hyperscore: A case study in computer mediated music composition."

Education and Information Technologies 10(3): 225-238.

Jennings, K. (2006). "Computer Graphical Interfaces, Reflection and Music Composition, A

Holistic Study." Department of Computer Science, Trinity College Dublin, Phd: 1-300.

Jennings, K. and B. Tangney (2001). DrumSteps-A constructionist approach to music

learning. 9th Technological Directions in Music Learning Conference, San Antonio, Texas.

Jonassen, D. H. (1994). Thinking technology: Toward a constructivist design model.

Educational technology, 34(4), 34-37.

Jonassen, D. H. (1999). Designing constructivist learning environments. Instructional design

theories and models: A new paradigm of instructional theory, 2, 215-239.

Jonassen, D. H. (2006). A constructivist's perspective on functional contextualism.

Educational Technology Research and Development, 54(1), 43-47.

Kardos, L. (2012). "How music technology can make sound and music worlds accessible to

student composers in Further Education colleges." British Journal of Music Education

29(02): 143-151.

Kelly, E. (2011). Gemnotes: a realtime music notation system for pure data. Proceedings of

IV International Conference of Pure data–Weimar.

228
Kerchner, J. L. (2000). Children's verbal, visual, and kinesthetic responses: Insight into their

music listening experience. Bulletin of the Council for Research in Music Education, 31-50.

Kolb, D. A. (1984). Experiential learning: Experience as the source of learning and

development (Vol. 1). Englewood Cliffs, NJ: Prentice-Hall.

Krathwohl, D. R. (2002). A revision of Bloom's taxonomy: An overview. Theory into

practice, 41(4), 212-218.

Lehrman, P. D. and T. M. Ryan (2005). Bridging the gap between art and science education

through teaching Electronic Musical Instrument Design. Proceedings of the 2005 conference

on new interfaces for musical expression, National University of Singapore.

Liu, C. H. and R. Matthews (2005). "Vygotsky's Philosophy: Constructivism and Its

Criticisms Examined." International Education Journal 6(3): 386-399.

Maag, M. (2006). "Podcasting and MP3 players: Emerging education technologies."

Computers Informatics Nursing 24(1): 9-13.

Manzo, V. (2010). Computer-aided Composition with High-school Non-music Students,

Philadelphia: TI: ME Action Research.

Marsh, J., Plowman, L., Yamada-Rice, D., Bishop, J. C., Lahmar, J., Scott, F., Davenport, A.,

Davis, S., French, K., Piras, M. & Thornhill, S. (2015). Exploring play and creativity in pre-

schoolers’ use of apps: Final project report. Technology and Play.

Mayer, R. E. (1992). Cognition and instruction: Their historic meeting within educational

psychology. Journal of Educational Psychology, 84(4), 405.

Marshall, P. (2007). Do tangible interfaces enhance learning? Proceedings of the 1st

international conference on Tangible and embedded interaction, ACM.

McCarthy, C., Bligh, J., Jennings, K., & Tangney, B. (2005). Virtual collaborative learning

environments for music: networked drumsteps. Computers & Education, 44(2), 173-195.

229
McKnight, L. and C. Davies (2013). Current Perspectives on Assistive Learning

Technologies: 2012 review of research and challenges within the field. University of Oxford,

The Kellogg College Centre for Research into Assistive Learning Technologies.

Mejía, A. (2009). On why there is generalisation in research on single cases. In Annual

Conference of the Philosophy of Education Society of Great Britain, New College, Oxford,

3rd to 5th April. Retrieved 09 December 2016 from

www.researchgate.net/profile/Andres_Mejia3/publication/228358734_On_why_there_is_gen

eralisation_in_research_on_single_cases/links/0deec5227aaf192b91000000/On-why-there-is-

generalisation-in-research-on-single-cases.pdf

Mellor, L. (2008). "Creativity, originality, identity: investigating computer-based

composition in the secondary school." Music Education Research 10(4): 451-472.

Merriam, S. B. (1998). Qualitative Research and Case Study Applications in Education.

Revised and Expanded from" Case Study Research in Education.". Jossey-Bass Publishers,

350 Sansome St, San Francisco, CA 94104.

Mills, J. (1989). The generalist primary teacher of music: A problem of confidence. British

Journal of Music Education, 6(2), 125-138.

Mills, G. E. (2000). Action research: A guide for the teacher researcher. Prentice-Hall, Inc.,

One Lake Street, Upper Saddle River, New Jersey 07458.

Mills, J., Bonner, A., & Francis, K. (2006). The development of constructivist grounded

theory. International journal of qualitative methods, 5(1), 25-35.

Mills, J. and A. Murray (2000). "Music technology inspected: good teaching in Key Stage 3."

British Journal of Music Education 17(02): 129-156.

Musicians’ Union (2014). Music education hubs: The real story so far. Retrieved 30

November 2016 from: www.musiciansunion.org.uk/Files/Reports/Industry/Music-Education-

Hubs-the-real-picture-so-far-Feb-2.aspx

230
Nelson, R. (Ed.). (2013). Practice as research in the arts: Principles, protocols, pedagogies,

resistances. Springer.

Norris, C., Hossain, A., & Soloway, E. (2011). Using smartphones as essential tools for

learning. Educational Technology, 51(3), 18-25.

Odam, G. (2000). "Teaching composing in secondary schools: the creative dream." British

Journal of Music Education 17(02): 109-127.

O'Donoghue, T. (2006). Planning your qualitative research project: An introduction to

interpretivist research in education. Routledge.

OFSTED (Office for Standards in Education) (2012) Music in schools: wider still, and wider.

Retrieved 14 November 2014 from

www.gov.uk/government/uploads/system/uploads/attachment_data/file/413347/Music_in_sc

hools_wider_still__and_wider.pdf

OFSTED (Office for Standards in Education) (2013) Music in schools: what hubs must do.

Retrieved 14 November 2014 from

www.gov.uk/government/uploads/system/uploads/attachment_data/file/413107/Music_in_sc

hools_what_hubs_must_do.pdf

Overby, A., & Jones, B. L. (2015). Virtual LEGOs: Incorporating Minecraft into the art

education curriculum. Art Education, 68(1), 21-27.

Palincsar, A. S. (1998). Social constructivist perspectives on teaching and learning. Annual

review of psychology, 49(1), 345-375.

Papert, S. (1980). Mindstorms: Children, computers, and powerful ideas, Basic Books, Inc.

Papert, S. (1987). Microworlds: transforming education. Artificial intelligence and education.

Papert, S., & Harel, I. (1991). Situating constructionism. Constructionism, 36, 1-11.

Partti, H. and S. Karlsen (2010). "Reconceptualising musical learning: New media, identity

and community in music education." Music Education Research 12(4): 369-382.

231
Paynter, J. (1982). Music in the Secondary School Curriculum: Trends and developments in

class music teaching. Cambridge university press.

Pegrum, M., Oakley, G., & Faulkner, R. (2013). Schools going mobile: A study of the

adoption of mobile handheld technologies in Western Australian independent schools.

Australasian Journal of Educational Technology, 29(1).

Phillips, D. C. (1995). "The good, the bad, and the ugly: The many faces of constructivism."

Educational researcher: 5-12.

Piaget, J., & Cook, M. (1952). The origins of intelligence in children (Vol. 8, No. 5, p. 18).

New York: International Universities Press.

Piaget, J. (1962) Play, dreams and imitation in childhood. London:Heinemann.

Pitts, A. and R. M. Kwami (2002). "Raising students' performance in music composition

through the use of information and communications technology (ICT): a survey of secondary

schools in England." British Journal of Music Education 19(01): 61-71.

Powell, K. C., & Kalina, C. J. (2009). Cognitive and social constructivism: Developing tools

for an effective classroom. Education, 130(2), 241-251.

Prensky, M. (2001). "Digital natives, digital immigrants part 1." On the horizon 9(5): 1-6.

Prensky, M. (2009). "H. sapiens digital: From digital immigrants and digital natives to digital

wisdom." Innovate: journal of online education 5(3): 1.

Pring, R. (2000). Philosophy of educational research. London: Continuum.

Puhlmann, K. (2014) Introduction to mathematics teaching and learning in the 1970s. In E. J.

Chernoff & G. Repenning, A., Webb, D. C., Brand, C., Gluck, F., Grover, R., Miller, S.,

Nickerson, H., & Song, M. (2014). Beyond minecraft: Facilitating computational thinking

through modeling and programming in 3d. IEEE Computer Graphics and Applications,

34(3), 68-71.

232
QCA (Qualifications and Curriculum Authority) (2000). The Arts, Creativity and Cultural

Education: An international perspective. London, QCA/NFER.

Raffle, H., Vaucelle, C., Wang, R., & Ishii, H. (2007, June). Jabberstamp: embedding sound

and voice in traditional drawings. In Proceedings of the 6th international conference on

Interaction design and children (pp. 137-144). ACM.

Reese, S. (2003). Responding to student compositions. Why and how to teach music

composition: A new horizon for music education. M. Hickey, Rowman & Littlefield

Education: 211-232.

Richardson, V. (1997). "Constructivist teaching and teacher education: Theory and practice."

Constructivist teacher education: Building a world of new understandings: 3-14.

Rieber, L. P. (1992). "Computer-based microworlds: A bridge between constructivism and

direct instruction." Educational technology research and development 40(1): 93-106.

Riley, P. (2013). Teaching, learning, and living with iPads. Music Educators Journal, 100(1),

81-86.

Riley, P. (2016). iPad apps for creating in your general music classroom. General Music

Today, 29(2), 4-13.

Rinaldo, V. and S. Denig (2009). "A Constructivist Approach to Learning Music: What Role,

if any, does active engagement play in the learning process?".

Robinson, M. (2017 November 21) “Tech billionaires spent $170 million on a new kind of

school — now classrooms are shrinking and some parents say their kids are 'guinea pigs'”

Retrieved 22 November 2017 from uk.businessinsider.com/altschool-why-parents-leaving-

2017-11

Roscoe, J. F., Fearn, S., & Posey, E. (2014, October). Teaching Computational Thinking by

Playing Games and Building Robots. In Interactive Technologies and Games (iTAG), 2014

International Conference on (pp. 9-12). IEEE.

233
Rosenbaum, E., & Silver, J. (2010, June). Singing Fingers: fingerpainting with sound. In

Proceedings of the 9th International Conference on Interaction Design and Children (pp.

308-310). ACM.

Roth, W. M. (2017). The invisible subject in educational science. Journal of Curriculum

Studies, 1-18.

Rouvroy, A., & Berns, T. (2013). Gouvernementalité algorithmique et perspectives

d'émancipation. Réseaux, (1), 163-196.

Rudi, J. (2007). "Computer Music Composition for Children [DSP Education]." Signal

Processing Magazine, IEEE 24(2): 140-143.

Ruismäki, H., Juvonen, A., & Lehtonen, K. (2013). The iPad and music in the new learning

environment. The European Journal of Social & Behavioural Sciences, 6(3), 1084-1096.

Russell-Bowie, D. (2009). What me? Teach music to my primary class? Challenges to

teaching music in primary schools in five countries. Music Education Research, 11(1), 23-36.

Salavuo, M. (2008). "Social media as an opportunity for pedagogical change in music

education." Journal of Music, Technology & Education 1(2-3): 121-136.

Sangani, K. (2013). BYOD to the classroom [bring your own device]. Engineering &

Technology, 8(3), 42-45.

Santos, I. M. (2013). Use of students’ personal mobile devices in the classroom: Overview of

key challenges. In E-Learn: World Conference on E-Learning in Corporate, Government,

Healthcare, and Higher Education (pp. 1585-1590). Association for the Advancement of

Computing in Education (AACE). Retrieved 06 July 2015 from

www.learntechlib.org/noaccess/115102/

Savage, J. (2005). "Working towards a theory for music technologies in the classroom: how

pupils engage with and organise sounds with new technologies." British Journal of Music

Education 22(02): 167.

234
Savage, J. (2007). "Reconstructing music education through ICT." Research in Education

78(1): 65-77.

Savage, J. and M. Challis (2001). "Dunwich revisited: Collaborative composition and

performance with new technologies." British Journal of Music Education 18(02): 139-149.

Savage, J. and M. Challis (2002). "Electroacoustic composition: Practical models of

composition with new technologies." Journal of the Sonic Arts Network 14: 8-13.

Sawyer, R. K. (2008). "Learning music from collaboration." International Journal of

Educational Research 47(1): 50-59.

Schafer, R. M. (1977). The tuning of the world: Toward a theory of soundscape design. New

York: Alfed A. Knopf.

Schifter, C., & Cipollone, M. (2013). Minecraft as a teaching tool: One case study. In Society

for Information Technology & Teacher Education International Conference (pp. 2951-2955).

Association for the Advancement of Computing in Education (AACE). Retrieved 04 July

2015 from www.learntechlib.org/p/48540/

Schifter, C. C. and M. Cipollone (2015). Constructivism vs Constructionism: Implications for

minecraft and classroom implementation. E-Learning Systems, Environments and

Approaches, Springer: 213-227.

Schön, D. A. (1987). Educating the reflective practitioner: Toward a new design for teaching

and learning in the professions. Jossey-Bass.

Schön, D. A. (2017). The reflective practitioner: How professionals think in action.

Routledge.

Scott, S. (2006). "A Constructivist View of Music Education: Perspectives for Deep

Learning." General Music Today 19(2): 17-21.

Scott, S. (2011). "Contemplating a constructivist stance for active learning within music

education." Arts Education Policy Review 112(4): 191-198.

235
Scripp, L., Meyaard, J., & Richardson, L. (1988). "Discerning musical development: Using

computers to discover what we know." Journal of Aesthetic Education 22: 75-88.

Seddon, F., & Biasutti, M. (2008). Non-music specialist trainee primary school teachers’

confidence in teaching music in the classroom. Music Education Research, 10(3), 403-421.

Seddon, F. A. and S. A. O'Neill (2001). "An evaluation study of computer-based

compositions by children with and without prior experience of formal instrumental music

tuition." Psychology of Music 29(1): 4-19.

Seddon, F. A. and S. A. O’Neill (2003). "Creative thinking processes in adolescent computer-

based composition: An analysis of strategies adopted and the influence of instrumental music

training." Music Education Research 5(2): 125-137.

Seddon, F. A. and S. A. O’Neill (2006). "How does formal instrumental music tuition

(FIMT) impact on self-and teacher-evaluations of adolescents’ computer-based

compositions?" Psychology of Music 34(1): 27-45.

Shively, J. (2015). Constructivism in music education. Arts Education Policy Review, 116(3),

128-136.

Short, D. (2012). Teaching scientific concepts using a virtual world—Minecraft. Teaching

Science-the Journal of the Australian Science Teachers Association, 58(3), 55.

Siemens, G. (2006). Connectivism: Learning theory or pastime of the self-amused. Retrieved

17 March 2016 from www.elearnspace.org/Articles/Connectivism_response.doc

Siemens, G. (2014). Connectivism: A learning theory for the digital age. Retrieved 04

January 2016 from

pdfs.semanticscholar.org/f5a6/d010046e4da2ef00e59730633ec0422b236b.pdf

Smith, H., & Dean, R. (Eds.). (2009). Practice-led research, research-led practice in the

creative arts. Edinburgh University Press.

236
Smith, N., Sutcliffe, C., & Sandvik, L. (2014). Code club: bringing programming to UK

primary schools through scratch. In Proceedings of the 45th ACM technical symposium on

Computer science education (pp. 517-522). ACM.

Song, Y. (2014). “Bring Your Own Device (BYOD)” for seamless science inquiry in a

primary school. Computers & Education, 74, 50-60.

Song, Y. (2016). “We found the ‘black spots’ on campus on our own”: development of

inquiry skills in primary science learning with BYOD (Bring Your Own Device). Interactive

Learning Environments, 24(2), 291-305.

Sterenberg (Eds.) Selected writings from the Journal of the Mathematics Council of the

Alberta Teacher’s Association. (pp. 75-80). Charlotte, NC: Information Age Publishing Inc.

Strauss, A. L. (1988). Basics of Qualitative Research. Sage.

Swanwick, K. (1979). A basis for music education. Routledge.

Swanwick, K., & Tillman, J. (1986). The sequence of musical development: A study of

children's composition. British journal of music education, 3(3), 305-339.

Tierney, S. (2012). Bring your own device to school. Microsoft Education.

Théberge, P. (1997). Any sound you can imagine: Making music/consuming technology,

Wesleyan University Press.

Thorndike, E. L. (2013). The principles of teaching: Based on psychology (Vol. 32).

Routledge.

Tobias, E. S. (2012). "Hybrid spaces and hyphenated musicians: secondary students' musical

engagement in a songwriting and technology course." Music Education Research 14(3): 329-

346.

Trouche, L. (2004). Managing the complexity of human/machine interactions in

computerized learning environments: Guiding students’ command process through

237
instrumental orchestrations. International Journal of Computers for mathematical learning,

9(3), 281.

Ufer, C. (1894). Introduction to the Pedagogy of Herbart (Vol. 23). DC Heath & Company.

Von Glasersfeld, E. (1995). "A constructivist approach to teaching." Constructivism in

education 3: 15.

Vygotsky, L. S. (1978). Mind and society: The development of higher mental processes,

Cambridge, MA: Harvard University Press.

Webster, P. R. (2002). Creative thinking and music education: Encouraging students to make

aesthetic decisions. u 10e Anniversaire, Actes de la Conference, European Society for the

Cognitive Sciences of Music, Liege, Belgium.

Webster, P. R. (2011). "Construction of music learning." Menc handbook of research on

music learning: 35-83.

Webster, P. R. (2012). "Key research in music technology and music teaching and learning."

Journal of Music, Technology & Education 4(2-3): 115-130.

Welch, G., Hallam, S., Lamont, A., Swanwick, K., Green, L., Hennessy, S., Cox, G., O’Neill,

S. & Farrell, G. (2004). Mapping music education research in the UK. Psychology of Music,

32(3), 239-290.

Wiggins, J. H. (1994). "Children's Strategies for Solving Compositional Problems with

Peers." Journal of Research in Music Education 42(3): 232-252.

Williams, D. A. (2014). Another perspective: The iPad is a REAL musical instrument. Music

Educators Journal, 101(1), 93-98.

Wilson, A., Hainey, T., & Connolly, T. (2012). Evaluation of computer games developed by

primary school children to gauge understanding of programming concepts. In European

Conference on Games Based Learning (p. 549). Academic Conferences International

Limited.

238
Wilson, A., Hainey, T., & Connolly, T. M. (2013). Using Scratch with primary school

children: an evaluation of games constructed to gauge understanding of programming

concepts. International Journal of Game-Based Learning (IJGBL), 3(1), 93-109.

Winters, M. (2012). "The challenges of teaching composing." British Journal of Music

Education 29(01): 19-24.

Wise, S., Greenwood, J., & Davis, N. (2011). Teachers' use of digital technology in

secondary music education: illustrations of changing classrooms. British Journal of Music

Education, 28(2), 117-134.

Wood, D., Bruner, J. S., & Ross, G. (1976). The role of tutoring in problem solving. Journal

of child psychology and psychiatry, 17(2), 89-100.

Wright, S. (2012) Flipping Bloom’s Taxonomy. Retrieved 18 November 2014 from

http://plpnetwork.com/2012/05/15/flipping-blooms-taxonomy/

Yin, J., Wang, Y., & Hsu, D. (2005). Digital violin tutor: an integrated system for beginning

violin learners. In Proceedings of the 13th annual ACM international conference on

Multimedia (pp. 976-985). ACM. Retrieved 18 November 2014 from

iiia.csic.es/praise/files/articles/p976-yin.pdf

Zhou, Y., Percival, G., Wang, X., Wang, Y., & Zhao, S. (2011). MOGCLASS: evaluation of

a collaborative system of mobile devices for classroom music education of young children. In

Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (pp. 523-

532). ACM. Retrieved 18 November 2014 from

comp.nus.edu.sg/~zhaosd/paper/chi2011_mogclass.pdf

Zuckerman, O., Arida, S., & Resnick, M. (2005). Extending tangible interfaces for education:

digital montessori-inspired manipulatives. In Proceedings of the SIGCHI conference on

Human factors in computing systems (pp. 859-868). ACM. Retrieved 09 November 2014

from kingkong.cc.gatech.edu/hci-seminar/uploads/29/p859-zuckerman.pdf

239