Journal of New Music Research

ISSN: 0929-8215 (Print) 1744-5027 (Online) Journal homepage: http://www.tandfonline.com/loi/nnmr20

Is Vivaldi smooth and takete? Non-verbal sensory

scales for describing music qualities

Maddalena Murari, Antonio Rodà, Sergio Canazza, Giovanni De Poli &

Osvaldo Da Pos

To cite this article: Maddalena Murari, Antonio Rodà, Sergio Canazza, Giovanni De
Poli & Osvaldo Da Pos (2015) Is Vivaldi smooth and takete? Non-verbal sensory scales
for describing music qualities, Journal of New Music Research, 44:4, 359-372, DOI:
10.1080/09298215.2015.1101475

To link to this article: http://dx.doi.org/10.1080/09298215.2015.1101475

Published online: 28 Oct 2015.

Submit your article to this journal

Article views: 90

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=nnmr20

Download by: [Universita di Padova] Date: 23 June 2016, At: 23:55

 Journal of New Music Research, 2015
Vol. 44, No. 4, 359–372, http://dx.doi.org/10.1080/09298215.2015.1101475

Is Vivaldi smooth and takete? Non-verbal sensory scales for

describing music qualities

Maddalena Murari, Antonio Rodà, Sergio Canazza, Giovanni De Poli and Osvaldo Da Pos

University of Padova, Italy

(Received 31 December 2014; accepted 11 September 2015)
Downloaded by [Universita di Padova] at 23:55 23 June 2016

Abstract life, such as entertainment (Zanolla, Canazza, Rodà, Camurri,

& Volpe, 2013), rehabilitation (Rosati, Rodà, Avanzini, &
Studies on the perception of music qualities (such as induced
Masiero, 2013) and sport training. Digital technologies and
or perceived emotions, performance styles, or timbre nuances)
new playback devices (e.g. mobile devices) have rapidly
make a large use of verbal descriptors. Although many authors
changed the way we listen to music: on one hand, one can
noted that particular music qualities can hardly be described
access a huge amount of musical content in any moment
by means of verbal labels, few studies have tried alternatives.
of the day and during almost any activity, including work,
This paper aims at exploring the use of non-verbal sensory
travel, eating, running, swimming, sleeping; on the other hand,
scales, in order to represent different perceived qualities in
mobile devices are equipped with or linked to increasingly
Western classical music. Musically trained and untrained lis-
sophisticated sensors, that can track human actions (such as
teners were required to listen to six musical excerpts in major
arm movements and pace rate) and bodily responses to those
key and to evaluate them from a sensorial and semantic point
actions (e.g. heart rate, skin conductance) and all this informa-
of view (Experiment 1). The same design (Experiment 2) was
tion can be instantaneously shared through social networks.
conducted using musically trained and untrained listeners who
Traditionally, studies on music qualities such as induced
were required to listen to six musical excerpts in minor key.
or perceived emotions, performance styles or timbre nuances
The overall findings indicate that subjects’ ratings on non-
make a large use of verbal labels. The most popular quantita-
verbal sensory scales are consistent throughout and the results
tive experimental method in this field is the semantic differen-
support the hypothesis that sensory scales can convey some
tial (Osgood, Suci, & Tannenbaum, 1957), an instrument for
specific sensations that cannot be described verbally, offering
measuring the connotative meaning of music through bipolar
interesting insights to deepen our knowledge on the relation-
rating scales.
ship between music and other sensorial experiences. Such
Given the recent scenario, where music playback devices
research can foster interesting applications in the field of music
allow us to interact with music by means of touch/actions/
information retrieval and timbre spaces explorations together
gestures, are verbal descriptions alone suitable to represent
with experiments applied to different musical cultures and
the multiform qualities of music? Can we find non-verbally-
contexts.
mediated forms of representation based on an action paradigm?
Can we discover forms of association that can help the under-
Keywords: expression, perception, interfaces, sensory scales,
standing of musical hidden qualities? How takete is Vivaldi
non-verbal description, crossmodal relationships
and how blue is Mozart?
Starting from the well-known semantic differential method,
1. Introduction this paper aims to explore the use of non-verbal sensory scales
to study different perceived qualities in Western classical
Music is a multifaceted reality covering an important role
music. The broad hypothesis is that both musically trained and
in many aspects of our life. The power of music to arouse
untrained subjects are able to consistently differentiate music
in the listener a rich set of sensations, such as images, feel-
excerpts along sensory qualities such as a piece of rough paper
ings or emotions, can have many applications in everyday

Correspondence: Antonio Rodà, University of Padova, Department of Information Engineering, via Gradenigo 6b, Padova, 35131 Italy. E-mail:
roda@dei.unipd.it

© 2015 Taylor & Francis

 360 M. Murari et al.

or a hot object. Moreover, we aim to compare non-verbal we reflect on the way in which scores are thought, analysed and
sensory scales with verbal scales, under the hypothesis that discussed, we can realize how often musicologists and music
non-verbal sensory scales can provide a different representa- theorists treat the composer as if s/he were the principal agent
tion of musical qualities to the traditional semantic differential of meaning. ‘Schubert lingers affectively on this glimpse of
method. bucolic happiness, though without slackening the pace of the
After an overview of previous studies investigating music song’ (Kramer, 2003). ‘Beethoven hurls forth the P0 (primary
from a non-verbal point of view (Sections 2 and 3), we present theme) module, Allegro vivace, in a set of three impetuous,
two experiments (Sections 4 and 5) that use two different sets downward-cast precipitations’ (Hepokoski, 2010). In these
of musical stimuli and the same set of verbal and non-verbal two examples the composer is treated as if he were also the
sensory scales. The comparison on the main results of the two performer. The writer tells us how the piece should sound—
experiments will be finally discussed in Section 6. ‘should’ because the conception is attributed to the composer
as performer. According to Leech-Wilkinson (2012), as soon
as we start thinking about the relationships between notes,
2. Background we have to imagine those notes sounding and our linguistic
translation does not refer to an abstract musical performance,
2.1 Music and language but to a particular performance style, the performance style
The idea of music as a language has been widely debated current around us (Leech-Wilkinson & Prior, 2013).
Downloaded by [Universita di Padova] at 23:55 23 June 2016

in past years (see e.g. Clark, 1982). Though music lacks the Linguistic symbols communicated from sender to receiver
specificity of natural language in terms of semantic meaning do not represent involvement with music in a direct way,
(Patel, 2007), neurological studies have recently shown that since they cannot render that sense of timelessness and loss
the brain seems to ‘see’ music as a language, in the sense that of feelings of self-consciousness that a music flow experience
the mechanisms underlying syntactic processing are shared can bring. Music does not allow the discursive, reciprocal
between music and language (Patel, 2003). Such syntactic communication of meaning, but rather an immediate and inef-
processes are instantiated in overlapping frontal brain areas fable communication (Jankélévitch & Abbate, 2003). Music,
that apply over different domain-specific syntactic represen- said Debussy, is made for that which cannot be expressed
tations in posterior brain regions. Nevertheless, lesion and (Siohan, 1959). ‘The truth we cannot speak about, we must
neuroimaging studies point out that some musical activities sing, we must say it in music’ (Berio, 2006).
like singing, memorisation of music and sight-reading are
functionally and neuroanatomically dissociable from analo-
2.2 Music and action
gous activities that involve speech (Peretz & Morais, 1989).
According to Peretz and Hyde (2003), some processing com- According to Leman (2008), music perception is rooted in
ponents appear to be genuinely specialized for music, while action. People try to be involved with music, because this
other components could be involved in the conjoint processing involvement permits an experience of behavioural resonance
of music and speech. with physical energy. Being aware of the state of the environ-
In any case, there is much evidence that music expresses ment in which music occurs can engage the mind in thinking
something, in the sense that groups of more or less culturally and reasoning processes that may break the magic spell of
homogeneous subjects have similar experiences when they being entrained. The measurement of musical involvement by
listen to the same music piece. Music has even been des- introspection is very difficult, since, as soon as direct involve-
cribed as the ‘language of emotions’ by some authors (Cooke, ment is evaluated and mentally represented in a conscious
1959) and many studies, investigating the relationship bet- way, it seems to disappear.
ween music and emotions, have shown that it is possible Some experimental designs have been proposed with the
to correlate the listeners’ main appraisal categories and the aim of studying the perceived qualities of music without fol-
acoustic parameters, even though there is still much debate lowing the semantic differential approach. De Poli, Mion &
on the distinction between perceived and induced emotions Rodà (2009) carried out experiments using two sets of musical
(Gabrielsson, 2002; Schubert, 2013). According to the func- excerpts, with simple and complex musical structure respec-
tionalist approach, emotional expression in music is partly tively, and three haptic reaction forces, provided by means of a
based on a code for vocal expression of basic emotion cat- force feedback device, which simulate the mechanic concepts
egories; music performance uses largely the same emotion- of friction, elasticity and inertia (FEI). Participants were asked
specific patterns of acoustic cues as does vocal expression to associate each musical piece to one of the three reaction
during verbal language production (Juslin & Laukka, 2003). forces. The results showed that subjects are able to consistently
Music performers communicate emotions to listeners by exp- relate the musical excerpts with the reaction forces, even if
loiting an acoustic code that derives from innate brain pro- the FEI metaphor seems to be more suitable for describing
grams for vocal expression of emotions. expressive cues in simple musical excerpts, where the exp-
But what music expresses is not limited to emotions. In the ressive content is mainly related to performance cues, than
field of music expressiveness, we note that much of what is in complex musical stimuli, where musical structure is more
said about music is actually about performances of pieces. If relevant.
 Is Vivaldi smooth and takete? 361

Amelynck, Grachten, van Noorden and Leman (2012) asked Ward, Huckstep and Tsakanikos (2006) studied a group of
the subjects to express the affective character of a set of mus- synaesthetes reporting colour sensations in response to music
ical fragments by arm movements, while listening to the and other sounds. Their experiment showed that synaesthetes
fragments. The subjects’ arm motion was captured using a choose more precise colours and are more internally consistent
wireless hand-held device equipped with 3D inertial sensors. in their choice of colours given a set of sounds of varying pitch,
The results show that human subjects are generally capable of timbre and composition (single notes or dyads) relative to a
expressing and recognizing affective qualities in music from control group. In spite of this difference, both controls and
arm gestures. synaesthetes appear to use the same heuristics for matching
Barten (1998) identified the effectiveness of metaphorical between auditory and visual domains (e.g. pitch to lightness).
verbal instructions for teaching the physical skills of music Overall, the results support the conclusion that this form of
performance. Often, instructors use imagery of movement, synaesthesia recruits some of the same mechanisms used in
such as an object travelling through the air and falling to the normal crossmodal perception rather than using direct, priv-
ground, to bring attention to the motional aspects of music, ileged pathways between unimodal auditory and unimodal
perhaps invoking musical concepts of ‘phrase’ and ‘line’. This visual areas that are absent in most other adults.
metaphoric language is particularly effective in improving Crisinel and Spence (2010) performed an experiment inv-
the student’s performance, since it is conveyed through a estigating crossmodal correspondences in synaesthetes and
particular vocal inflection, facial expression, gesture, posture, nonsynaesthetes through explicit matching of sounds of vary-
Downloaded by [Universita di Padova] at 23:55 23 June 2016

rendering the linguistic expression more vivid and emotion- ing pitch to a range of tastes/flavours. In addition, participants
ally loaded. also chose the type of musical instrument most appropriate for
each taste/flavour. The association of sweet and sour tastes to
high-pitched notes was confirmed. By contrast, umami and
2.3 Crossmodal correspondences bitter tastes were preferentially matched to low-pitched notes.
The characteristic of music of eliciting a rich number of The choice of musical instrument was primarily driven by a
often shared sensations and images may be based on cross- matching of the hedonic value and familiarity of the two types
modal correspondences which can be interpreted as intrin- of stimuli.
sic attributes of our perceptual system’s organization (Marks, Belkin, Martin, Kemp and Gilbert (1997) investigated the
1978). Many studies investigated the relationship between relationship of certain odours (in particular, those that are
music and colours, tastes, vision and odours, proving that commonly used in perfumery) to the pitch of a tone. The stim-
people exhibit consistent crossmodal responses in different uli were drawn from several fragrance categories (e.g. floral,
sensory modalities. In particular, Martino and Marks (2001) citrus, woody, earthy) and included essential oils extracted
introduce the two concepts of strong and weak synaesthesia, from botanical sources as well as synthetic aroma compounds.
arguing that these two phenomena may share similar neural Discussing the results of the experiment, Belkin hypothesized
mechanisms, probably involving sensory and semantic areas that stimulus intensity and pleasantness cannot account for the
of the brain. Strong synaesthesia refers to a vivid percept matching of odours and sounds, since the relevant perceptual
in one sensory modality induced by stimulation in another, feature appears to be some aspect (or aspects) of odour-quality
while weak synaesthesia refers to a correspondence expressed coding. According to Belkin, subjects could have based their
through language, perceptual similarity and perceptual int- pitch matches along such putative olfactory dimensions as
eractions between different sensory experiences. According dull–aromatic, heavy–light, bright–dark, or hard–soft.
to Spence (2011), crossmodal correspondences between dif- Ramachandran and Hubbard (2001), replicating Köhler’s
ferent combinations of modal and amodal features need to 1929 experiments with maluma and takete using the words
be considered alongside semantic, spatiotemporal, develop- ‘kiki’ and ‘bouba’, suggested that the human brain somehow
mental and affective dimensions, since such correspondences attaches abstract meanings to the shapes and sounds in a
are mediated not only by innate factors, but also by experi- consistent way. They hypothesize that the kiki/bouba effect
ence and statistical factors. In particular, Spence distinguishes has implications for the evolution of language, because it
between structural, statistical and semantic correspondences suggests that the naming of objects is not completely arbitrary.
and hypothesizes that structural correspondences may bear De Götzen (2014) experimented the association of maluma
fruitful comparison with synaesthesia, while other kinds of and takete with two sound movements in space, proving that,
correspondence may need a different theoretical interpreta- also in this particular case, the quality of the movement is
tion. As regards the neural substrates underlying crossmodal easily associated with the quality of words.
correspondences, polysensory areas in the temporal cortex All these studies show that crossmodal correspondences
have been shown to respond more vigorously to crossmodally exist between a variety of auditory, visual, gustatory, haptic
congruent pairings of simultaneously presented auditory and and tactile stimuli. In this regard, we believe that sensory
visual action stimuli (Beauchamp, Argall, Bodurka, Duyn, & scales can represent a suitable tool to measure, describe and
Martin, 2004). Moreover, damage to the angular gyrus can directly access such crossmodal similarities in response to a
interfere with a person’s ability to match stimuli crossmodally piece of music.
as in the bouba/kiki test (Ramachandran & Hubbard, 2001).
 362 M. Murari et al.

3. Methodological issues cluster structure can be also obtained with minor-mode-only

excerpts.
3.1 Cluster analysis of music qualities
In order to study how the affective qualities of music are per-
ceived in the absence of verbal descriptions, Bigand, Vieillard,
Madurell, Marozeau and Dacquet (2005) proposed an exper- 3.2 Non-verbal sensory scales
imental setup that does not use verbal labels to describe dif- The use of linguistic labels is one of the most complex prob-
ferences between the stimuli: musically trained and untrained lems of the studies investigating the emotional aspects of
listeners were asked to listen to 27 musical excerpts taken from music. Musical emotions are so undetermined that it is difficult
the Western classical repertoire and to freely group those con- to render them through words, since the determinedness of
veying similar subjective emotions. By means of cluster analy- language causes an inevitable loss in richness of meanings.
sis and multidimensional scaling (MDS), the musical excerpts The use of verbal labels can encourage participants to simplify
resulted in being grouped into four main clusters, interpreted what they actually experience (Scherer, 1994) and it is still
as high arousal and high valence (HAHV); high arousal and uncertain if research based on the recording of electrophys-
low valence (HALV); low arousal and high valence (LAHV); iological responses to musical stimuli (Khalfa et al., 2002;
low arousal and low valence (LALV). Following this approach, Krumhansl, 1997) can faithfully account for the subtlety of
subjects could choose to divide the stimuli into as many groups musical emotions.
Downloaded by [Universita di Padova] at 23:55 23 June 2016

as they wanted. Some perceptual experiences cannot be expressed by means

A variant of this approach consists in asking subjects to of an accurate, reliable language (lexicon) and this fact is
associate each music piece to one of n representative music proved by the high interindividual variability in the attempts to
stimuli, called attractors (Mion, De Poli, & Rapanà, 2010). describe them. Particular touch experiences, such as variations
By means of attractors, the subjects are induced to focus on of springiness and slipperiness have no correspondent verbal
particular aspects of music qualities, without using a verbal descriptors, even if their perceptual understanding is very
label. For example, Mion et al. (2010) used three attractors, simple (Guest et al., 2011).
obtained asking a musician to play a major scale with three dif- In order to establish a correspondence between concepts
ferent expressive intentions, i.e. three different performance from disparate domains of knowledge, we decided to exploit
styles. Asking the subject to associate a set of music excerpts cross-domain mappings by means of metaphoric association.
with one on the three attractors, they were induced to focus According to psycholinguistic theories (Bowdle & Gentner,
on the performance aspects of the music excerpts. 2005), highly conventional metaphors are processed without
Cluster analysis can also be carried out starting from the any active mapping from the source to the target domain (e.g.
acoustic analysis of the music excerpts. Mion and De Poli from space to pitch) since they are interpreted as directly as
(2008) used machine learning techniques to select the audio literal language.
features that are most relevant for the recognition of differ- Sensory scales were first introduced by Da Pos and Pietto
ent music qualities both in the affective (valence-arousal) (2010) with the aim of substituting Osgood’s verbal scales
and sensorial (kinematics-energy) spaces. The analysis of the with sensory ones. For instance, instead of asking the observer
feature space revealed a structure into three clusters, inter- where he/she would have placed his/her light impression in
preted as hard/heavy/angry, light/happy, and sad/calm/soft the continuum between ‘warm’and ‘cold’expressed by words,
respectively. they made the observer feel his/her hands cool or warm by
Music composers know very well that the music mode and plunging them in cold and warm water. The results obtained
the performance tempo are related to the affective properties confirmed the hypothesis that sensory scales greatly improve
of music. Traditionally, the minor mode has been attributed the efficiency of the semantic differential in enlightening the
to feelings of grief and melancholy whereas the major mode impact that perceived objects and events determine in the
to feelings of joy and happiness (Gerardi & Gerken, 1995; human mind.
Hevner, 1936; Kastner & Crowder, 1990). In order to reduce In our experiments we asked participants to evaluate their
the relevant influence of modality and tempo, Rodà, Canazza musical experience from a subjective sensorial point of view.
& Poli (2014) planned two experiments imposing some con- Seven non-verbal sensory scales (visual, tactile, haptic and
straints on the musical features, namely modality and tempo, gustative) were presented in random sequence in order to stim-
of the stimuli. In particular, the first experiment used music ulate a quasi-synaesthesic response. Participants were con-
excerpts in major-mode only, whereas the second one excer- fronted with the following scales: maluma/takete, blue/orange,
pts in major-mode and performed at 104 bpm. Following an hard/soft, smooth/rough, bitter/sweet, heavy/light, cold/warm
experimental method based on free grouping (Bigand et al., previously tested in an experiment on colour perception
2005), the authors showed that participants grouped the stim- (Da Pos et al., 2013). In this case the aim of the researchers was
uli in a consistent way, structuring them into three main clus- the study of different reactions to normal or iridescent colours
ters, interpreted as low arousal (LA), high arousal and low and it resulted that evaluations in all scales, sensory and sem-
potency (HA-LP), high arousal and high potency (HA-HP) antic, significantly discriminated normal from changeable
respectively. Preliminary experiments showed that this three- colours.
 Is Vivaldi smooth and takete? 363

Another study (Da Pos & Pietto, 2010) on unique hues experiments were conducted to study the clustering of the
led to the discrimination of four characteristic factors: one affective qualities of music (see Section 3.1). For the present
including the three verbal scales typical of Osgood’s findings, study, six music pieces, representing the three main clusters of
one including the warm/cold scale, one characterized by the Rodà et al. (2014), two pieces for each cluster, were chosen.
sensory opposition of dissonant, aloud, and orange against All the selected stimuli were in a major tonality. Each excerpt
consonant, faint, and turquoise and the last including the glass- had a duration of about 30 s. A list of the stimuli is reported in
sandpaper scale, and the light-heavy scale dealing with the Appendix A.
sensation of smoothness given by the light sources. These
two studies confirmed the hypothesis that the sensory scales 4.3 Materials
used in the evaluation of light sources can show up relevant
We prepared the following material:
qualitative aspects otherwise hidden.
As regards the semantic analysis, we decided to rely upon
• maluma–takete (Köhler, 1929), computer visualiza-
the semantic differential, a type of rating scale designed to
tion of the two visual forms (cm 4.3 × 4.3)
measure the connotative meaning of objects, events, and con-
(Figure 1);
cepts ideated by Osgood et al. (1957). Ratings on bipolar adj-
• blue–orange, the computer display of the two colours
ective scales tend to be correlated, and three basic dimensions
(NCS notation: S 2055-B10G, S 1080-Y70R, cm 4.3
of response account for most of the co-variation in ratings.
Downloaded by [Universita di Padova] at 23:55 23 June 2016

× 4.3) (Figure 2);

The three dimensions, which have been labelled Evaluation,
• hard–soft, a piece of wood of cylindrical shape and a
Potency and Activity, have been verified and replicated in an
cylinder of polystyrene foam (16 × 3 × 3 cm; 16 ×
impressive variety of studies. Evaluation can be matched to
6 × 6 cm);
the adjective pair ‘desirable-undesirable’. The ‘gentle-violent’
• smooth–rough, N 1200 and N 30 sandpapers (cm 15×
adjective pair defines the potency factor.Adjective pair ‘active-
10.5);
passive’ defines the activity factor. These three dimensions of
• bitter–sweet: a bitter substance (Zefirus Calma Plant,
affective meaning were found to be cross-cultural universals
two drops in a small cup) and water with sugar
in a study of dozens of cultures. The semantic differential
(1 teaspoon of sugar in a small cup);
technique is a multivariate approach to affect measurement.
• heavy–light: a dark plastic bottle full of liquid (500 ml)
It is also a generalized approach, applicable to any concept or
and the same bottle without liquid (5 g). The dull
stimulus, and thus it permits comparisons of affective reac-
colour of the bottle didn’t allow participants to dis-
tions on widely disparate topics.
tinguish the full from the empty bottle;
• cold–warm: one cup of cold water and one cup of
4. Experiment 1 warm water (temperature: 5◦ and 40◦ ).
4.1 Participants
4.4 Procedure
Twenty participants were recruited on a voluntary basis, of
which 12 were musicians (age range 17–25, mean age 21.4; 8 The sound files were represented on the computer screen by
women and 4 men) and 8 were non-musicians (age range 24– small icons with the number of the musical excerpts. Partici-
68, mean age 42; 4 women and 4 men).Aquestionnaire was ad- pants listened to the excerpt once (30 s), but they had the possi-
ministered to determine the participants’ musical background bility to listen to the stimuli as many times as they wanted. To
and experience, including listing the instruments played and avoid order effects, sensory and semantic scales were adm-
number of years trained in each instrument. inistered randomly. The experiment consisted in expressing
a subjective judgment on the characteristics of the excerpt
4.2 Stimuli heard by placing the indicator inside the slider at the point
that was considered representative of the association strength
The musical excerpts were selected starting from two previous either with the sensations on which the listener was focused,
works by Bigand et al. (2005) and Rodà et al. (2014), in which or with the meaning of the verbal terms proposed. The slider
had no visual references (numbers or internal divisions) apart
from the position of minimum and maximum. The computer
program transposed and registered the chosen position.
After listening to each excerpt, participants could express
their evaluations along the 20 scales (7 sensory and 13 verbal)
and register them by clicking on a bar put under the slider.
Once the subjective rate was confirmed, participants could
express the following one.
The average duration of the experiment was 20–25 min, but,
Fig. 1. The two forms named maluma (on the left) and takete (on the interestingly enough, musicians tended to complete the task
right) (Köhler, 1929). in a longer time. Probably they weighed their answers in a
 364 M. Murari et al.

Fig. 2. The two colours that constitute the second sensory scale.
Downloaded by [Universita di Padova] at 23:55 23 June 2016

Fig. 3. The average scores obtained by the six musical excerpts of Experiment 1 on the seven sensory scales. Bars indicate the standard error.

more careful way, since the musical excerpts stirred emotions differential was used (Da Pos, Degan, Pultrone, & Ciraci,
connected to the autobiographical memory, thus activating 2013). We used 13 verbal scales corresponding to Osgood’s
more complex associations. To evaluate the expressive charac- three main factors: activity (1, 5, 10, 12), evaluation (2, 4, 6,
teristics of the excerpts, a new version of the Osgood semantic 7, 8, 9) and power (3, 11, 13).
 Is Vivaldi smooth and takete? 365

Table 1. Significance p values with FDR correction of the differences between pairs of excerpts of Experiment 1. Blank cells mean p > 0.05.

Ma/Ta Bl/Or Ha/So Sm/Ro Bi/Sw He/Li Co/Wa

1–2 <0.001 0.002 0.009

1–3 <0.001 0.005 0.029 0.001 0.018
1–4 0.019
1–5 <0.001 0.010 0.072
1–6 <0.001 < 0.001 0.003 < 0.001 < 0.001
2–3 0.013
2–4 <0.001 0.032 0.001 < 0.001
2–5
2–6 <.001 0.037 0.005
3–4 <0.001 0.020 0.003 0.001 < 0.001
3–5
3–6 0.038 0.020
4–5 <0.001 0.003 0.004
4–6 <0.001 < 0.001 < 0.001 < 0.001 < 0.001
5–6 < 0.001 0.023 0.016
Downloaded by [Universita di Padova] at 23:55 23 June 2016

• Active–Passive statistic computed for the verbal scales only (α = 0.96) and
• Boring–Interesting for all (both verbal and sensory) the scales (α = 0.94).
• Slow–Fast Figure 3 represents the average subjective evaluations on
• Superficial–Deep each sensory scale; musical excerpts are reported along the
• Tense–Relaxed x-axis and are identified with the number listed in Appendix A.
• Masculine–Feminine In addition, error bars are displayed (standard error of means)
• Clear–Confused for each average value. It can be noted that the musical stimuli
• Undesirable–Desirable are judged very differently on some scales (e.g. maluma–
• Brilliant–Dark takete), while the differences are less marked on other ones
• Simple–Complex (e.g. hard–soft).
• Shaken–Calm An ANOVA analysis was carried out in order to emphas-
• Intimate–Open air ize the average values that are significantly different. Table 1
• Gentle–Violent shows the significance levels (p-value) of the differences
between each pairs of musical excerpts. The p-values are
corrected by means of False Discovery Rate (FDR) using the
4.5 Results Benjamini–Hochberg (1995) procedure, which relies on the
Initially, the average scores for the musician and the non- p-values being uniformly distributed under the null hypothe-
musician subjects were separately calculated. The two sets sis.Accordingly, the procedure consists of sorting the p-values
of values presented a high correlation value (r (118) = 0.87, in ascending order, and then dividing each observed p-value
p < 0.001, where r is the Pearson’s correlation coefficient by its percentile rank to get an estimated FDR. Stimuli 1 and 4
and the p-value is computed using algorithm AS 89 (Best and share the quality ‘maluma’, while 2, 3, 5, 6 belong to the cat-
Roberts, 1975), implying a high agreement between musicians egory ‘takete’. Brahms’ violin concerto (1) and Mozart’s flute
and non-musicians. Then, the following results include the concerto (4) were judged significantly different from all the
responses of both groups. other excerpts. It is interesting to notice that the pairs of stimuli
Amultivariate analysis of variance (MANOVA) was carried discriminated by the scale maluma/takete are the same differ-
out considering the musical excerpts as independent variables entiated by the scale light/heavy. We can hypothesize that the
and the 20 scales (sensory and verbal) as dependent vari- listening of the stimuli aroused in participants an association
ables. A significant multivariate main effect was found for the between maluma and lightness. Examining the research litera-
musical excerpts (Wilks’ λ = 0.013, F(100, 375) = 5.39, ture, we didn’t find a similar association of maluma and light-
p < 0.001, η2p = 0.581). ness intended as a haptic sensation. Nevertheless, Janković,
The Cronbach’s α statistic has been computed to test the Vucković and Radaković (2005) analysed the relationship
inter-subjects reliability. Results show that subjects are able of physical characteristics of visual patterns with consonant
to rate music on the seven non-verbal sensory scales in a types, asking participants to produce pseudo-words corre-
highly consistent way (α = 0.90), i.e. subjects are able to sponding to presented monochromatic abstract visual patterns.
recognize the sensorial differences of the non-verbal scales Takete-like figures were characterized by sharp and dark pat-
and to associate them with musical stimuli in a meaningful terns while maluma-like figures were associated with oval,
way, although this value is slightly lower than the Cronbach’s light, luminous figures. The sensory scales blue/orange and
 366 M. Murari et al.

soft/hard don’t convey any significant result; Mozart and Amultivariate analysis of variance (MANOVA) was carried
Brahms are considered the bluest and the softest stimuli, but out considering the musical excerpts as independent variables
the results don’t allow us to discriminate between couples and the 20 scales (sensory and verbal) as dependent variables.
of excerpts. The scale smooth/rough enucleates excerpt 6 as A significant multivariate main effect was found for the mus-
the roughest. Brahms’ horn trio (6) significantly differs from ical excerpts (Wilks’ λ = 0.001, F(100, 497.39) = 15.51,
excerpts 1, 2, 4, 5. Also Bizet (3) is felt rough enough, since p < 0.001, η2p = 0.749). The Cronbach’s α statistic shows
it differs significantly from 1, 2 and 4. The scale sweet/bitter that subjects of Experiment 2 rated the musical excerpts on
is characterized by results very similar to the previous one. the seven non-verbal sensory scales in a highly consistent way
Brahms and Bizet are significantly different from stimuli 1 (α = 0.90), as for Experiment 1, a value that is slightly lower
and 4. The sweetest excerpt is Mozart and it differs signifi- than the Cronbach’s statistic computed for the verbal scales
cantly from stimuli 2, 3 and 6. ‘The sweet taste of maluma’ only (α = 0.95) and for all (both verbal and sensory) the
investigated by Crisinel, Jones and Spence (2012) is confirmed scales (α = 0.93).
also in our experiment, since both Brahms’s violin concerto Figure 4 represents the average subjective evaluations on
and Mozart’s flute concerto are significantly characterized the scale indicated for each musical excerpt examined. In
by these two qualities. Also in the scale warm/cold, Brahms addition, error bars are displayed (standard error of means).
and Mozart are considered the warmest pieces, since they Table 2 shows the significance levels ( p) of the differences
differ significantly from every other excerpt. Other significant between pairs of tracks. Two couples of excerpts are signif-
Downloaded by [Universita di Padova] at 23:55 23 June 2016

results regard stimulus 6, which differs significantly from all icantly differentiated in every sensory scale (3–5; 4–5). This
the other excerpts, since it is considered the coldest. means that Bach’s Badinerie (5) is felt as juxtaposed both to
Chopin’s Prelude (3) and to Liszt’s Tasso (4) and these pairs
of stimuli are seen as opposites. Another interesting juxta-
position regards excerpts 1 and 3, which significantly differ
5. Experiment 2 in almost every sensory scale except the cold/warm scale.
5.1 Participants The sensory scale maluma/takete significantly differentiates
excerpt 3 (Chopin) from 1 (Mozart), 2 (Wagner), 5 (Bach)
Participants, recruited on a voluntary basis and different from
and 6 (Rossini). This implies that Chopin’s Prelude is felt
the ones of the previous experiment, were 25, of which 10
as the sharpest excerpt and this is probably due to the par-
musicians (age range 23–47, mean age 37.2; 3 women and 7
ticularly percussive style chosen by the pianist. Also excerpt
men) and 15 non-musicians (age range 24–77, mean age 56.1;
4 (Liszt) and 5 (Bach) are significantly differentiated by the
10 women and 5 men).
sensory scale maluma/takete. As regards the sensory scale
blue/orange, it is interesting to notice how Mozart (1) and Bach
(5) are considered the bluest excerpts. Mozart significantly
5.2 Stimuli differs from stimuli 2, 3, 4 and 6, while Bach significantly
Six stimuli were chosen from the Western classical repertoire. differs from 3 and 4. The scale hard/soft particularly discrimi-
Unlike the first experiment, all the selected stimuli were in a nates stimulus 5 (Bach). This excerpt is considered the softest
minor tonality. The six stimuli represent the three main clusters one, since it differs significantly from every other stimulus.
of Rodà et al. (2014), two pieces for each cluster. Four of The same happens for 3 (Chopin), considered the hardest,
the six excerpts were also used in Bigand et al. (2005). Each and significantly differentiated from every other stimulus. As
excerpt had a duration of about 30 s; a list of the stimuli is regards the scales smooth/rough, bitter/sweet and heavy/light,
reported in Appendix A. significant differences can be seen between excerpts 1, 5,
6 and 2, 3, 4 respectively. This result is particularly rele-
vant for the fact that the group formed by tracks 1, 5, 6
is characterized by smoothness, sweetness and lightness as
5.3 Materials and procedure opposed to the roughness, bitterness and heaviness of the
Materials and procedure are the same as Experiment 1. other triad. Inside the scale smooth/rough, even excerpt 3
is differentiated from 4, thus emphasizing the roughness of
Chopin’s Prelude as opposed to Liszt’s symphonic poem. As
regards the scale cold/warm, the key excerpt is no. 5, which is
5.4 Results felt as the warmest and which is significantly different from
The average scores separately computed for the musician and stimuli 1, 3 and 4. In this scale we don’t find extreme values,
the non-musician subjects present a high correlation value since it is characterized by great inter-participant variability,
(r (118) = 0.86, p < 0.001, where r is the Pearson’s correla- determining similar average values. Strangely enough, Wag-
tion coefficient and the p-value is computed using algorithm ner is perceived as warmer than Mozart and this is probably
AS 89 (Best & Roberts, 1975), implying a high agreement due to the fact that this excerpt is perceived as more vio-
between musicians and non-musicians. Thus, the following lent and the heat was intended as a quality associated with
results include the responses of both groups. burning.
 Is Vivaldi smooth and takete? 367
Downloaded by [Universita di Padova] at 23:55 23 June 2016

Fig. 4. The average scores obtained by the six musical excerpts of Experiment 2 on the seven sensory scales. Bars indicate the standard error.

Table 2. Significant p values with FDR correction of the differences between pairs of excerpts of Experiment 2. Blank cell means p > 0.05.

Ma/Ta Bl/Or Ha/So Sm/Ro Bi/Sw He/Li Co/Wa

1–2 < 0.001 < 0.001 0.002 0.001

1–3 0.001 < 0.001 0.001 < 0.001 0.015 0.001
1–4 < 0.001 < 0.001 0.029 0.002
1–5 0.016 0.001 0.004 0.013
1–6 0.050
2–3 0.047 0.020
2–4
2–5 0.050 < 0.001 < 0.001 < 0.001 < 0.001
2–6 0.008 < 0.001 < 0.001
3–4 0.020 0.042
3–5 < 0.001 0.050 < 0.001 < 0.001 < 0.001 < 0.001 0.013
3–6 0.012 0.001 < 0.001 < 0.001 0.002 0.024
4–5 0.011 0.050 < 0.001 < 0.001 < 0.001 < 0.001 0.039
4–6 0.042 < 0.001 0.006
5–6 0.007 0.008 0.002 0.006

6. Discussion a slow tempo with a strong dramatization due to many sudden

crescendos. The association of low arousal and negative val-
As can be seen in Figures 3 and 4, participants’ ratings on sen-
ence may have inspired some particular correlation between
sory scales tended to configure according to a similar profile
the colours and the shapes. In both experiments, Mozart (flute
with great consistency. The quality ‘maluma’ always couples
concerto and piano concerto) is evaluated as the bluest and
with ‘blue’, while the quality ‘takete’ always couples with
the most maluma composer, together with Bach (Badinerie)
‘orange’. The only exception was excerpt 2 in Experiment 2,
and Brahms (violin concerto).
R. Wagner, Tristan, Act 3, in which the quality orange and
As regards the sensory scales, we see how maluma/takete,
maluma were associated, even if with very low values. Proba-
smooth/rough, sweet/bitter and light/heavy could differentiate
bly this exception is due to the fact that this excerpt combines
between musical examples. More problematic were the scales
 368 M. Murari et al.

Table 3. The qualities of the six excerpts of Experiment 1, based on the subjects’ evaluation. Blank cells mean that no significant trend has been
found.

1 2 3 4 5 6

maluma takete takete maluma takete takete

soft soft
smooth smooth smooth smooth rough
sweet sweet bitter
light light heavy
warm warm cold
Active Active Passive Active Active
Interesting Interesting Interesting Interesting Interesting
Slow Fast Fast Slow Fast Fast
Superficial Deep Deep Superficial Deep
Relaxed Tense Relaxed Tense
Feminine Masculine Feminine Masculine
Clear Clear Clear Clear
Desirable Desirable Desirable Desirable Desirable
Downloaded by [Universita di Padova] at 23:55 23 June 2016

Brilliant Brilliant Brilliant

Simple Complex Simple Simple Complex
Calm Shaken Shaken Calm Shaken Shaken
Intimate Open air Open air Intimate Open air
Gentle Gentle Gentle Gentle Violent

Table 4. The qualities of the six excerpts of Experiment 2, based on the subjects’ evaluation. Blank cells mean that no significant trend has been
found.

1 2 3 4 5 6

maluma takete maluma

blue blue
hard soft
smooth rough rough smooth
bitter bitter bitter sweet sweet
heavy heavy heavy light
warm warm
Active Active Active Active Active
Interesting Interesting Interesting Interesting Interesting Interesting
Slow Slow Fast Fast Fast Fast
Deep Deep Deep Deep Deep
Relaxed Tense Tense Tense Relaxed Tense
Masculine Masculine Masculine Feminine
Clear Clear Clear
Desirable Desirable Desirable
Dark Dark Brilliant Brilliant
Simple Complex Complex Complex Simple
Calm Shaken Shaken Shaken Shaken
Intimate Open air Open air
Gentle Violent Violent Gentle

blue/orange and soft/hard; these two scales didn’t provide to be teamed up with darker, greyer, bluer colours (Palmer,
significant values, apart from the blue quality of Mozart’s Schloss, Xu, & Prado-León, 2013). Also in our study, music
piano concerto and Bach’s Badinerie and the soft quality of in the minor mode was in general associated with blue colours.
Brahms’ violin concerto and Mozart’s flute concerto. This In order to observe the interaction of the semantic scales
result can be compared with research based on the association with the sensory ones, Tables 3 and 4 show the qualities of
between music and colours. Researchers at the University the six excerpts of Experiment 1 and 2, based on the subjects’
of Berkeley found that people tended to pair faster-paced evaluation. In particular, only rates significantly different from
music in a major key with lighter, more vivid, yellow colours, 50 (the middle point of the evaluation scale) are reported. It
whereas slower-paced music in a minor key was more likely is interesting to notice how Bizet (3 Exp 1), Brahms’ horn
 Is Vivaldi smooth and takete? 369

trio (6 Exp 1), Chopin (3 Exp 2) and Liszt (4 Exp 2), des- investigate crossmodal relationships. An interesting example
cribed verbally with the same characteristics, (active, inter- is represented by the coupling of the two apparently opposite
esting, fast, deep, tense, masculine, complex, shaken, open qualities takete and smooth as characterizing the excerpts
air and violent), receive significantly different sensorial res- by Vivaldi and Boccherini. These two tracks share the same
ults. This makes us suppose that sensory scales can convey verbal labels, but none of them can catch this apparent paradox
some specific sensations that cannot be described verbally. which is an intrinsic feature of music’s multifaceted nature.
Both Brahms’ horn trio and Chopin are in triple meter with Sensory scales seem to be able to depict music timbral aspects
significant changes from ternary to binary subdivision, they and imaginative shades which escape a linguistic definition.
share a similar articulation (many notes played marcato) per- Results confirm that the tool created for our experiments
formed by the violin and by the right hand of the pianist provides a novel way of allowing qualities of music to be
and a common accompaniment based on repeated piano oc- differentiated. The consistency of results suggests the appli-
taves. Participants were however able to recognize a slight cation of sensory scales to other musical cultures, styles and
sensory difference between the two pieces, represented by genres, overcoming the boundaries of Western classical music.
Chopin’s higher hardness, due to the presence of many ac- Since features (acoustic, structural, style-related or cultural)
cents, and by Brahms’ higher coldness, due to the obsessive- of the music used in the experiments are not controlled or
ness of the rhythmic configuration. Another interesting result systematically accounted for, further experiments are needed
deriving from sensory scales is represented by the apparent un- in order to relate sensory scales to these features.
Downloaded by [Universita di Padova] at 23:55 23 June 2016

usual association of the quality takete with the quality smooth From the application point of view, such research can foster
applied to Vivaldi (2 Exp 1) and Boccherini (5 Exp 1). Also the development of innovative interfaces to browse audio
in this case participants’ evaluation seems to recall timbral digital collections. These new devices will allow users to
elements; the staccato and pizzicato generate an idea of sharp- interrelate in a spontaneous and even expressive way with
ness delimiting only the contours of an ideal figure perceived interactive multimedia systems, relying on a set of advanced
as covered by a smooth surface. Comparing the two Mozart musical and gestural content processing tools adapted to the
excerpts (4 Exp 1 and 1 Exp 2), we see that they have 12 profiles of individual users, adopting descriptions of perceived
features in common (maluma, smooth, interesting, slow, deep, qualities, or making expressive movements.
relaxed, clear, desirable, simple, calm, intimate, gentle). The Another interesting application regards the exploration of
flute concerto differs from the piano concerto in softness, the timbre space of a synthesis algorithm, allowing the tran-
smoothness, sweetness, lightness and warmness, while, from sition between one timbre to the other in an intuitive way.
the verbal point of view, we can recognize their diversity only By exploiting the potentialities of metaphoric associations,
from the higher darkness of the piano concerto. In this case, musicians will be able to navigate timbre spaces generated
the qualities underlined by sensory scales offer us some hints by synthesis algorithms with easiness in exploring unknown
about some ineffable aspect of the music related to affect, sounds.
allowing insights into the depressive and sombre atmosphere Further insights can be provided by direct comparison of
of the second excerpt. Also Brahms’ violin concerto (1 Exp 1) evaluations based on sensory scales and their ‘equivalent’
and Bach’s Badinerie (5 Exp 2) have 12 features in common, verbal scales. An experiment is on the way to explore this
but in this case, sensory scales report a similarity which is relationship. Results can identify possible intrinsic aspects
not matched by verbal scales. Verbal scales are crucial in of music, underlining qualities that are general, universal,
discriminating the swiftness and brilliantness of Bach, but valid prima facie for any listener, but, at the same time, the
the sensory scales provide useful information about timbral expression of subjective and personal characterizations.
aspects, since both stimuli are played by two wind instruments Advances in music research need to be based on the col-
(oboe and flute) with a particularly warm, soft and rich sound. laboration of multiple research fields such as musicology,
Further research is needed to confirm our hypothesis on the psychology and engineering, in synergy with artistic explo-
kind of information provided by sensory scales. In particular, rations. Such research should deepen our knowledge on the
it is interesting to verify which kind of metric relations (if relationship between music and other sensorial experiences,
any) can be found between the sensory scales (e.g. see Zaidi overcoming the limits imposed by verbal scales and help-
et al., 2013). This step could further the definition of a low- ing to develop a holistic consciousness of musical expressive
dimensional sensory space, to be used in music information content.
retrieval applications.

7. Conclusions Acknowledgements
Two experiments were carried out in order to test the possibil- The authors wish to acknowledge Prof. Emery Schubert, School
ity of describing music qualities through non-verbal sensory of the Arts and Media, University of New South Wales, and
scales. The overall findings indicate that subjects’ rates on Prof. Alessandro Paccagnella, Dept. of Information Engineer-
sensory scales are consistent and the results support the hyp- ing, University of Padova, for helpful discussions and
othesis that sensory scales can provide a useful tool in order to suggestions.
 370 M. Murari et al.

References 2014 and 11th Sound and Music Computing Conference, SMC
2014 "Music Technology Meets Philosophy: From Digital
Amelynck, D., Grachten, M., van Noorden, L., & Leman, M.
Echos to Virtual Ethos", Athens.
(2012). Toward e-motion-based music retrieval a study of
De Poli, G., Mion, L., & Rodà, A. (2009). Toward an action
affective gesture recognition. IEEE Transactions on Affective
based metaphor for gestural interaction with musical contents.
Computing, 3(2), 250–259.
Journal of New Music Research, 38(3), 295–307.
Barten, S. S. (1998). Speaking of music: The use of motor-
Gabrielsson, A. (2002). Emotion perceived and emotion felt:
affective metaphors in music instruction. Journal of Aesthetic
Same or different? Musicae Scientiae, 5 (1 suppl), 123–147.
Education, 32(2), 89–97.
Gerardi, G. M., & Gerken, L. (1995). The development of
Beauchamp, M. S., Argall, B., Bodurka, J., Duyn, J. H., &
affective responses to modality and melodic contour. Music
Martin, A. (2004). Unraveling multisensory integration: Patchy
Perception: An Interdisciplinary Journal, 12(3), 279–290.
organization within human STS multisensory cortex. Nature
Guest, S., Dessirier, J. M., Mehrabyan, A., McGlone, F., Essick,
Neuroscience, 7(11), 1190–1192.
G., & Gescheider, G. (2011). The development and validation
Belkin, K., Martin, R., Kemp, S. E., & Gilbert, A. N. (1997).
of sensory and emotional scales of touch perception. Attention,
Auditory pitch as a perceptual analogue to odor quality.
Perception, & Psychophysics, 73(2), 531–550.
Psychological Science, 8(4), 340–342.
Hepokoski, J. (2010). Formal process, sonata theory, and the
Benjamini, Y., & Hochberg, Y. (1995). Controlling the false
first movement of Beethoven’s “Tempest” sonata. Music
discovery rate: a practical and powerful approach to multiple
Downloaded by [Universita di Padova] at 23:55 23 June 2016

Theory Online, 16 (2), http://www.mtosmt.org/issues/mto.10.

testing. Journal of the Royal Statistical Society. Series B, 57(1),
16.2/mto.10.16.2.hepokoski.html
289–300.
Hevner, K. (1936). Experimental studies of the elements of
Berio, L. (2006). Remembering the future. Cambridge, MA:
expression in music. The American Journal of Psychology,
Harvard University Press.
48(2), 246–268.
Best, D. J., & Roberts, D. E. (1975). Algorithm AS 89: The
Jankélévitch, V., & Abbate, C. (2003). Music and the ineffable.
upper tail probabilities of Spearman’s rho. Journal of the Royal
Princeton, NJ: Princeton University Press.
Statistical Society. Series C. Applied Statistics, 24(3), 377–379.
Janković, D., Vucković, V., & Radaković, N. (2005). Consonants
Bigand, E., Vieillard, S., Madurell, F., Marozeau, J., & Dacquet,
in the takete-maluma phenomenon: Manner and place of
A. (2005). Multidimensional scaling of emotional responses to
articulation. Perception, 34, ECVP Abstract Supplement
music: The effect of musical expertise and of the duration of
Juslin, P.N., & Laukka, P. (2003). Communication of emotions in
the excerpts. Cognition and Emotion, 19(8), 1113–1139.
vocal expression and music performance: Different channels,
Bowdle, B. F., & Gentner, D. (2005). The career of metaphor.
same code? Psychological Bulletin, 129(5), 770–814.
Psychological Review, 112(1), 193–216.
Kastner, M. P., & Crowder, R. G. (1990). Perception of the
Clark, A. (1982). Is music a language? The Journal of Aesthetics
major/minor distinction: IV. Emotional connotations in young
and Art Criticism, 41(2), 195–204.
children. Music Perception: An Interdisciplinary Journal, 8(2),
Cooke, D. (1959). The language of music. Oxford: Oxford
189–201.
University Press.
Khalfa, S., Peretz, I., Blondin, J. P., & Robert, M. (2002). Event-
Crisinel, A., Jones, S., & Spence, C. (2012). "The sweet taste of
related skin conductance responses to musical emotions in
maluma": Crossmodal associations between tastes and words.
humans. Neuroscience Letters, 328(2), 145–149.
Chemosensory Perception, 5(3–4), 266–273.
Köhler, W. (1929). Gestalt psychology (2nd ed. 1947). New York:
Crisinel, A. S., & Spence, C. (2010). As bitter as a trombone:
Liveright.
Synesthetic correspondences in nonsynesthetes between
Kramer, L. (2003). Franz Schubert: Sexuality, subjectivity, song.
tastes/flavors and musical notes. Attention, Perception, &
Cambridge, UK: Cambridge University Press.
Psychophysics, 72(7), 1994–2002.
Krumhansl, C. L. (1997). An exploratory study of musical
Da Pos, O., Degan, V. D., Pultrone, D., & Ciraci, I. (2013, July).
emotions and psychophysiology. Canadian Journal of
Towards a psychology of normal and iridescent colours. In
Experimental Psychology, 51(4), 336–353.
Paper presented at the 12th Congress of the International Colour
Leech-Wilkinson, D. (2012). Compositions, scores, perfor-
Association AIC Newcastle.
mances, meanings. Music Theory Online, 18(1), 1–17.
Da Pos, O., Fiorentin, P., Scroccaro, A., Filippi, A., Fontana, C.,
Leech-Wilkinson, D., & Prior, H. (2013). Heuristics for
Gardin, E., & Guerra, D. (2013, April). Subjective assessment
expressive performance. In D. Fabian, E. Schubert, &
of unique colours as a tool to evaluate colour differences
R. Timmers (Eds.), Expressiveness in music performance.
in different adaptation conditions. In Paper presented at CIE
Oxford: Oxford University Press.
Centenary Conference, “Towards a New Century of Light”,
Leman, M. (2008). Embodied music cognition and mediation
Paris.
technology. Cambridge, MA: MIT Press.
Da Pos, O., & Pietto, M. (2010, September). Highlighting the
Marks, L. (1978). The unity of the senses: Interrelations among
quality of light sources. Paper presented at the 2nd CIE
the modalities. New York: Academic Press.
Expert Symposium on Appearance - "When Appearance meets
Martino, G., & Marks, L. E. (2001). Synesthesia: Strong and
Lighting", Gent.
weak. Current Directions in Psychological Science, 10(2),
De Götzen, A. (2014, September). Association of Sound
61–65.
movements in space to Takete and Maluma. Paper presented
at the 40th International Computer Music Conference, ICMC
 Is Vivaldi smooth and takete? 371

Mion, L., & De Poli, G. (2008). Score-independent audio features Appendix A. Description of the musical excerpts
for description of music expression. IEEE Transactions on
Speech, Audio, and Language Process, 16(2), 458–466. Experiment 1
Mion, L., De Poli, G., & Rapanà, E. (2010). Perceptual 1 J. Brahms—Violin Concert in D major, op. 77, Ada-
organization of affective and sensorial expressive intentions in gio. Thematic exposition on the oboe of a slow,
music performance. ACM Transactions on Applied Perception, pure melodic line, built on the tonic major chord,
7(2), 14:1–14:21. and standing apart above a timbrally rich, sustained
Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. (1957). The orchestra. The doubling of lines serves to reinforce
measurement of meaning. Urbana, IL: University of Illinois the fullness of sound of the whole.
Press. 2 A. Vivaldi—Trio Sonata in C major, RV82, Allegro.
Palmer, S. E., Schloss, K. B., Xu, Z., & Prado-León, L. R.
Vigorous and cheerful passage, characterized by a
(2013). Music-color associations are mediated by emotion.
thematic development entrusted to the combination
Proceedings of the National Academy of Sciences of the United
of lute and violin. The violin plays rapid trills, thus
States of America, 110(22), 8836–8841.
Patel, A. D. (2003). Language, music, syntax and the brain.
complementing the lute’s quick, athletic ornaments
Nature Neuroscience, 6(7), 674–681. with its own sharp notes. The ascending tone is
Patel, A. D. (2007). Music, language, and the brain. Oxford: emphasized by the intensive use of progressions
enriched by the continuous dialogue between lute
Downloaded by [Universita di Padova] at 23:55 23 June 2016

Oxford University Press.

Peretz, I., & Hyde, K. L. (2003). What is specific to music and violin.
processing? Insights from congenital amusia. Trends in 3 G. Bizet—Symphony no. 1 in C major,Allegro vivo.
Cognitive Sciences, 7(8), 362–367. The work starts with an opening tutti full of strength
Peretz, I., & Morais, J. (1989). Music and modularity. and force, like a brisk announcement. This bold
Contemporary Music Review, 4(1), 279–293. first idea is answered by a timid pp reply by the
Ramachandran, V. S., & Hubbard, E. M. (2001). Synaesthesia - winds which are soon harassed again by the tutti
A window into perception, thought and language. Journal of repeating the same announcement this time leading
Consciousness Studies, 8(12), 3–34. to G major.
Rodà, A., Canazza, S., & Poli, G. D. (2014). Clustering affective 4 W.A. Mozart—Flute Concerto G Major, II. Andante
qualities of classical music: Beyond the valence-arousal plane. non Troppo. Gentle and relaxed theme developed by
IEEE Transactions on Affective Computing, 5(4), 364–376. the flute through an expanse figuration of demisemi-
Rosati, G., Rodà, A., Avanzini, F., & Masiero, S. (2013). On quavers. The orchestra accompanies this quiet
the role of auditory feedback in robot-assisted movement moment with soft pizzicato and tender eight notes,
training after stroke: Review of the literature. Computational while the violin answers the flute with responding
Intelligence and Neuroscience, 2013, 1–15. demisemiquavers figurations.
Scherer, K. R. (1994). Affect bursts. In S. van Goozen, N. E. van 5 L. Boccherini—String Quintet in E major, op. 11,
de Poll, & J. A. Sergeant (Eds.), Emotions: Essays on emotion
no. 5, Minuetto. This popular piece is full of grace
theory (pp. 161–196). Mahwah, NJ: Erlbaum.
and elegance. The dance rhythm is underlined by the
Schubert, E. (2013). Emotion felt by the listener and
upbeat quarter line in the first violin embellished by
expressed by the music: A literature review and theoretical
investigation. Frontiers in Psychology, 4, 837. doi: 10.3389/fp-
a characteristic grace note.
syg.2013.00837 6 J. Brahms—Trio, piano, violin, and horn, mvt 2.
Siohan, R. (1959). Possibilités et limites de l’abstraction Repetition of a thematic rhythmic motif, above
musicale. Journal de Psychologie, 56, 257–274. major key harmony, punctuated by brass effects,
Spence, C. (2011). Crossmodal correspondences: A tutorial at a rapid tempo and with a very rich sound. The
review. Attention, Perception, & Psychophysics, 73(4), 971– sonority of the French horn enriches the timbral
995. quality of the ensemble, and the structure of the
Ward, J., Huckstep, B., & Tsakanikos, E. (2006). Sound- piece is reinforced by the presence of transposed
colour synaesthesia: To what extent does it use cross-modal harmonic progressions.
mechanisms common to us all? Cortex, 42(2), 264–280.
Zaidi, Q., Victor, J., McDermott, J., Geffen, M., Bensmaia,
S., & Cleland, T.A. (2013). Perceptual spaces: Mathematical Experiment 2
structures to neural mechanisms. The Journal of Neuroscience, 1 W.A. Mozart, Piano concertoAdagio, K 488. Theme
33(45), 17597–17602. in a minor key, played at a very slow tempo.
Zanolla, S., Canazza, S., Rodà, A., Camurri, A., & Volpe,
G. (2013). Entertaining listening by means of the Stanza
Logo-Motoria: An interactive multimodal environment.
Entertainment Computing, 4, 213–220.
 372 M. Murari et al.

Melancholic trochaic rhythm characterized by a 4 F. Liszt, Tasso Lamento and Trionfo (from letter
large intervallic distance between sounds grouped A allegro strepitoso). Powerful orchestral line dev-
by the left hand, and the melody in the high register elops tense minor harmonies on a choppy rhythm
of the right hand, creating a void in the middle of and at a rapid tempo, supported by the entry of the
the range which reinforces the desolate aspect of percussion.
the theme. 5 J.S. Bach, Badinerie from Orchestral Suite no. 2
2 R. Wagner, Tristan, Act 3. Declamation in the low BWV 1067. Exposition of the main theme by the
register of the strings of the orchestra. Very strong flute in the typical dance rhythm characterized by
harmonic tension within a minor key with on the a joyous and light feeling. The orchestral accompa-
6th chord against a dissonant second. Slow and niment is very simple and elegant.
dilated tempo. The upper parts ascend in pitch by 6 G. Rossini, La Gazza Ladra (The Thieving Magpie)—
chromatic movement, with unresolved intervallic Allegro con brio. Particularly fast and tense orc-
tensions. The absence of a bass creates a feeling of hestral passage characterized by frequent accents
vertigo and of ascension into infinity. and chromatic contrasts. Triplets figurations in the
3 F. Chopin, Prelude 22. Motif in the low register violins and violas are punctuated by the other strings
of the piano repeated obsessively and character- playing a very pressing and obstinate rhythm.
ized by pounding octaves in the left hand, dissonant
Downloaded by [Universita di Padova] at 23:55 23 June 2016

harmonies, and accompanied in the right hand by

a panting rhythm, accentuating the weak part of
the beat, and breaking up the violent and hopeless
discourse of the left hand.