FRANK MONAGHAN

WHAT DIFFERENCE DOES IT MAKE?

CHILDREN’S VIEWS OF THE DIFFERENCES BETWEEN SOME
QUADRILATERALS

ABSTRACT. This paper describes an attempt to explore aspects of the language used in
children’s written work in a secondary mathematics classroom as a means of assessing
their mathematical understanding. It focuses on children’s conceptualisations of polygons
and their attempts to differentiate between them. It also examines the use of database and
concordancing software in deepening our understanding of a potentially major source of
students’ (mis) conceptions: the published materials they engage with. It describes their
usefulness in providing a systematic and efficient method of analysis of materials to explore
patterns within them. The method is also used to identify possible sources of students’ mis-
conceptions about the properties of polygons based on their likely exposure to potentially
misleading examples in published materials.
. . . concept formation is a creative, not a mechanical passive, process; . . . a
concept emerges and takes shape in the course of a complex operation aimed
at the solution of some problem . . . memorizing words and connecting them with
objects does not in itself lead to concept formation; for the process to begin, a
problem must arise that cannot be solved otherwise than through the formation of
new concepts.
(Vygotsky, 1986: 99–100)

KEY WORDS: language, mathematics, secondary, quadrilaterals, van Hiele

C ONCEPT DIFFERENTIATION

In this paper, the benefits of concept differentiation in mathematics will

be explored as a means of promoting the kind of cognitive conflict that
Vygotsky (1978, 1984), posits as being essential to successful concept
formation. As Adler (1997, p. 238) makes clear, Vygotsky “recognises the
school as a distinct context entailing distinct kinds of activities leading to
qualitatively different kinds of knowledge from those acquired in everyday
life, in play or in work”. In terms of conceptual development, Vygot-
sky makes a distinction between children’s spontaneous concepts, which
develop unsystematically through the interactions of everyday life and sci-
entific concepts, which are part of a system of concepts and commonly
mediated through schooling (Vygotsky, 1986: 193–194):

Educational Studies in Mathematics 42: 179–196, 2000.

© 2001 Kluwer Academic Publishers. Printed in the Netherlands.
180 FRANK MONAGHAN

The inception of a spontaneous concept can usually be traced to a face-to-face

meeting with a concrete situation, while a scientific concept involves from the first
a ‘mediated’ attitude towards its object.

Additionally, it seeks to explore and promote the value of such a ‘mediated’

and explicit focus on the language of mathematics as a means of develop-
ing students’ command and facility with this particular type of discourse.
As Adler (1997, p. 251) points out:

. . . talking within and talking about mathematics within the classroom and its
mathematical practices, while deeply related, do not place the same communic-
ative demands on the speaker . . . The move between talking within and talking
about is not spontaneously or tacitly learnt. It requires mediation.

This paper describes an example of such language-focussed mediation

based around an activity requiring students to engage in concept differen-
tiation, which in essence involved presenting students with pairs of related
concepts and asking them to analyse the differences between them. It is a
relatively simple and economic method of gaining insights into students’
current conceptual understandings. By promoting cognitive conflict, how-
ever, it may have greater potential as a means of developing such under-
standings further and of recognising and correcting possible sources of
misunderstandings.
In this study, the cognitive conflict is brought about by asking students
to describe in their own words the differences between pairs of quad-
rilaterals. The underlying assumption is that in describing such differ-
ences students will, amongst other things, reveal valuable insights into
their conceptual understanding and will thereby assist teachers in their
efforts to scaffold their students’ learning. This assumption is founded on
a constructivist model of learning (Glaserfeld, 1995; Mercer, 1995; Ernest,
1994; Sierpinska, 1994). Adler has investigated this area of mathematics
education and acknowledges the value of such an approach with teach-
ers. In reporting on the results of workshops with teachers she comments
(Adler 1999, p. 54–55):

The teachers agreed that pupils’ saying what they were thinking would, at least,
help the teacher to know what learners were constructing and to respond appropri-
ately. One summed up this view in the workshop discussion: “Hearing what it is
pupils think and articulate can help you [the teacher] see what they understand”.

That view is endorsed by this paper, which seeks to locate such under-
standings within the van Hiele model of geometric development.
 CHILDREN’S VIEWS OF DIFFERENCES BETWEEN SOME QUADRILATERALS 181

T HE VAN H IELE MODEL

The van Hiele model of geometric development, as described in Crow-

ley (1987), proposes five levels of understanding: visualization, analysis,
informal deduction, formal deduction, and rigor. The model asserts that
learners move through this hierarchical sequence from the initial, or ba-
sic, level (visualization), where the properties of figures are not explicitly
recognized, to the highest level (rigor), which is concerned with formal
abstract aspects of deduction. Crowley (1987, p. 1–3) describes these five
stages as follows:

Level 0 (Basic level): Visualization

. . . Geometric figures . . . are recognized by their shape as a whole, that is, by their
physical appearance, not by their parts or properties.

Level 1: Analysis
At level 1, an analysis of geometric concepts begins. For example, through obser-
vation and experimentation students begin to discern the characteristics of figures.
These emerging properties are then used to conceptualize classes of shapes. Thus
figures are recognized as having parts and are recognized by their parts. . . . Rela-
tionships between properties, however, cannot yet be explained by students at this
level, interrelationships between figures are still not seen, and definitions are not
yet understood.

Level 2: Informal deduction

At this level, students can establish the interrelationships of properties both within
figures (e.g., in a quadrilateral, opposite sides being parallel necessitates opposite
angles being equal) and among figures (a square is a rectangle because it has
all the properties of a rectangle). Class inclusion is understood. Definitions are
meaningful. Informal arguments can be followed and given. The students at this
level, however, do not comprehend the significance of deduction as a whole or the
role of axioms. Empirically obtained results are often used in conjunction with
deduction techniques. Formal proofs can be followed, but students do not see
how the logical order could be altered nor do they see how to construct a proof
starting from different or unfamiliar premises.

Level 3: Deduction
At this level, the significance of deduction as a way of establishing geometric
theory within an axiomatic system is understood. The interrelationship and role
of undefined terms, axioms, postulates, definitions, theorems, and proof is seen.
A person at this level can construct, not just memorize, proofs; the possibility of
developing a proof in more than one way is seen; the interaction of necessary
and sufficient conditions is understood; distinctions between a statement and its
converse can be made.

Level 4: Rigor
At this stage the learner can work in a variety of axiomatic systems, that is, non-
182 FRANK MONAGHAN

Euclidean geometries can be studied, and different systems can be compared.

Geometry is seen in the abstract.
The levels are assessed through the students’ ability to talk, draw or write
about their conceptual understanding: at level 0, it proposes observation;
at level 1 recognition; at level 2 definition; at level 3 proof; and at level 4
axiomatic systems. It is clear from this outline that the model entails many
linguistic implications, as Van Hiele (1984, p. 246) points out: “Each level
has its own linguistic symbols and its own systems of relations connecting
these symbols”.
Certainly in the early years of secondary schooling (the focus of this
paper) very few students would be expected to have gone beyond level
2. By the time students reach secondary school they are generally able to
identify basic shapes in isolation, so that an exercise such as the following
(at van Hiele level 0) causes little or no difficulty:

Figure 1. Shape discrimination.

Children are likely to be able to identify these shapes and should also
have little difficulty in identifying particular shapes from among a group
of shapes, as in the following exercise:

Figure 2. Shape identification.

In my experience, when faced with a more complex exercise such as

in figure 3, more advanced students will generally be able to identify the
individual shapes as required:
School mathematics curricula intend that children have a clear concept
of what shapes look like and are able to use this mental map to navigate
 CHILDREN’S VIEWS OF DIFFERENCES BETWEEN SOME QUADRILATERALS 183

Figure 3. More complex shape discrimination.

their way through exercises such as those above. There are significant dif-
ferences, however, between being able to identify shapes, describe shapes
and classify them, with implications for the degree of mathematical un-
derstanding that is assumed. England’s National Curriculum Orders, for
example (DFE, 1995, p. 27) locate the classification process at both level
3 (of 8): “Pupils classify 3-D and 2-D shapes in various ways using math-
ematical properties”, and at level 6, “They know and use the properties of
quadrilaterals in classifying different types of quadrilateral”.
At the lower levels, classifying may, perhaps, be taken to be closer in
meaning to describing (akin to van Hiele levels 0 and 1) and, at higher
levels, to defining (akin to van Hiele level 2). There is, of course, a clear
difference to be made between describing shapes and defining them, but
the term classifying is used at diverse levels with no operational distinction
made between them. Such statements of attainment, of course, are gener-
ally written with summative assessment by examination in mind rather
than as a description of a systematic and coherent progression through
a defined hierarchy of mathematical concepts. This, perhaps, accounts for
abrupt and unprogrammed shifts within such descriptors but it does little to
help teachers faced with the difficulty of planning a curriculum that devel-
ops their students’ ability to demonstrate their knowledge of mathematical
concepts through the medium of language. Nor does it assist teachers in the
formative assessment of such abilities, let alone help the students develop
the skills of self-assessment, which are increasingly recognised as being of
great value, as argued by Gipps (1994, p. 26–27):
If pupils are to become competent assesors of their own work, as developments
in metacognition tell us they should, then they need sustained experience in ways
of questioning and improving their work, and supported experience in assessing
their work, in addition to understanding what counts as the standard expected and
the criteria on which they will be assessed.
Any such planning for future learning must be founded upon an under-
standing and appreciation of students’ current conceptualisations. The fol-
184 FRANK MONAGHAN

lowing study represents an attempt to use concept differentiation as a means

of formatively assessing a group of Year 7 students’ (aged 11–12 years)
understanding of shapes and their use of language to express that under-
standing.

T HE STUDY

The setting
The study was carried out in a large (ca. 2,000 students) comprehens-
ive school in an inner London borough marked by high levels of social
deprivation. Approximately two thirds of the students regularly use a lan-
guage other than English, and about one third of this group are in the early
stages of acquiring the language. All students are taught mathematics in
mixed-attainment classes from the ages of 11–16, with every class con-
taining a wide spread of ability, but with more students falling into the
lower end of the range than the higher.

The study
The conceptual distance that students must cover to move from the stage of
recognising such gross visual features of shapes as straightness or length
to more abstract concepts such as parallelness or perpendicularity is far
greater than the mere difference in vocabulary might suggest. As has been
pointed out, students very early on are able to recognise and distinguish
shapes. What is less clear is the basis on which they make such distinctions.
In order to investigate this, an activity was devised in which a group
of 24 Year 7 students were asked not to identify shapes in isolation but
to differentiate between them. The aim was to analyse the language stu-
dents used to describe differences between shapes in order to explore what
it reveals about their conceptualisation of the shapes. The impetus for
this derives from Vygotsky’s notion of cognitive conflict discussed earlier,
whereby the formation of concepts is regarded as a creative process in
which concepts emerge as a result of solving a problem, rather than through
the mere memorization of words and connection of them to some object.
The exercise took place during the students’ first half term in the school
at a time when they had not done any work on classifying shapes. (They
will, of course, have worked on shape at their primary schools but this
type of definitional exercise was new to them.) By requiring the students to
identify the salient attributes which distinguish between shapes, it was as-
sumed that evidence would emerge about the basis on which they identified
shapes. The students were asked to answer the following five questions:
 CHILDREN’S VIEWS OF DIFFERENCES BETWEEN SOME QUADRILATERALS 185

1. What’s the difference between a square and a rectangle?

2. What’s the difference between a rectangle and a parallelogram?
3. What’s the difference between a square and a rhombus?
4. What’s the difference between a parallelogram and a kite?
5. What’s the difference between a trapezium and a parallelogram?
The following is an analysis of their responses. (For reasons of space, I
will report only on their responses to the first two. The students’ spelling is
preserved and square brackets are used to suggest possible interpretations
where the original meaning was not clear.)

S TUDENT RESPONSES

1. What’s the difference between a square and a rectangle?

Three students made no real distinction between the shapes:
The difference is that a square has 4 right angles and a rectangle all the opposite
sides are equal. (Matthew)
Whilst neither of Matthew’s statements is incorrect neither of them illus-
trates a difference between the two shapes as both have four right angles
and both have opposite sides equal.
there both quaralagle [quadrilaterals] and they have both got 4 siads. (Chris)
Chris attempts to use a mathematical term (quadrilateral) but by going
on to produce an unintentional tautology suggests that he does not really
understand it. Had he just been asked to describe a rectangle, however, and
used the term, his teacher is unlikely to have realised that he is not clear
about its meaning and may have even been impressed by his ‘knowledge’
of mathematical terms.
a square have 4 sides shapes and rectangle is a 4 sides shapes why because the
square is a 4 right langles and a rectangle is a 4 right angles (Munira)
Munira has grasped the point that both shapes are right-angled quadri-
laterals (and she is unusual in commenting on angles) but fails to make
any distinction between the two. She may have grasped the underlying
principle that a square is a special case of rectangle, which would be a
considerable achievement on her part. However, her answers to the other
questions make this conjecture appear unlikely.
Four of the twenty-four pupils gave an answer that focused on the es-
sential difference between a rectangle and a square, i.e. that in a rectangle
opposite sides must be equal and in a square all four sides must be equal:
A rectangle has each side the same. The oppiste side are the same.
186 FRANK MONAGHAN

A square has 4 equal sides, A rectangle has 2 oppisite sides are equal.

A rectangle has two opposite sides equal and a square has four equal sides.

A square is a special rectangle because it is all equal sided.

The majority of students (sixteen of them) perceive the difference between
a square and a rectangle in a similar but significantly different way. The
salient difference for these students is the matter of horizontal length. This
perception, of course, is commonly held but is mathematically inaccurate
as it ignores the square as a special case of rectangle. More importantly,
what this perception reveals is the extent to which students’ conceptualisa-
tions of shape are governed by the standard representations they generally
encounter in mathematics materials. They generalise that rectangles are
inherently longer than squares and that it is differences in the horizontal
sides that distinguishes them, as these examples show:
Longer
A rectangle is a four sided shape but it is longer. (Marcus)

The lenth of square is more short then the rectangle. (Celina)

Horizontal
A square has all four sides the same length and a rectangle alway has two opposite
sides the same length and the other two are the same length but a different than
the top and bottom (Del)

Rectangle has the opposite sides which are equal. The top and botum are the
same, and the square have got same sides and angles. (Ghalib)

Both longer and horizontal

A rectangle has two sides (the bottom and top) wich are wider than the two sides
and a square has all four sides the same. (Andrew)
It is, of course, possible to construct a square and a rectangle with the
sides of the square longer than those of the rectangle and students will
recognise which is which. Equally, they are unlikely to fail to recognise
a rectangle where the vertical width is greater than the horizontal length.
(For children ‘length’ seems normally to refer to horizontal sides and width
to the vertical. The terms ‘length’ and ‘width’ of a rectangle will be used
here in that sense.)
It could be argued that the students are being forced into this ‘error’
by being asked to make a difference where there may be none and so
alight on the more ‘relaxed’ conditions for the rectangle, where all four
sides may be equal as opposed to the square where all four sides must be
equal. The important point that emerges from their descriptions, however,
is the rigidity of their view of the rectangle as an oblong. The salience
 CHILDREN’S VIEWS OF DIFFERENCES BETWEEN SOME QUADRILATERALS 187

of such ‘prototypes’ and the obstacles they can represent has been widely
discussed (Hasegawa, 1997; De Villiers, 1994; Laborde, 1994; Fischbein,
1993; Parzysz, 1991; Sfard, 1991; Hershkovitz, 1990) and informs this pa-
per. Students tend to overgeneralise the properties of one type of rectangle
to the whole class. As Hasegawa (1997, p. 157–158) comments:

The prototype is a result of our visual-perceptual limitations which affect the

identification ability of individuals, and individuals use the prototypical example
as a model in their judgements of other instances.

Conceptually speaking, there is no such thing as a rectangle, there are

two types: the oblong and the square. This is not how students perceive
it, however; their concept of a rectangle has become fixed as being syn-
onymous with an oblong. This rigidity, I would argue, stems from the
way students are presented with a particular orientation of rectangles in
teaching materials in which there is a standard one-to-one object-word
match.
To test this I used a database I had created as part of a broader study
(Monaghan, 1997) of the text of the materials from a mathematics scheme
used widely in London secondary schools and beyond. I then used con-
cordancing software, which allowed me to investigate all occurrences of
particular strings in context, and so was able to explore various patterns
that emerged in the texts of the materials.
An examination of the representation of rectangles in the materials
would suggest that the materials themselves may – albeit unwittingly –
be providing an erroneous deductive basis that might account for the com-
monly found perception among students that a rectangle is always longer
than it is wide (or vice versa!) – as implied by Celina’s statement above
that, “the lenth of square is more short then the rectangle”.
The analysis covered over 1,400 activities, which was some 93% of
the total materials available. 32 of these activities contained illustrations
of some 48 rectangles. Whilst there was a wide range of different sizes of
rectangles, an analysis of the ratio of the longer and shorter sides of the
depicted shapes suggests a ‘standardised’ representation of the rectangle
as having one pair of sides approximately twice the length of the other.
(The modal ratio was indeed found to be 1:2 and the mean ratio 1: 2.19.)
Some two-thirds of the rectangles depicted in the materials showed the
rectangle with the horizontal length greater than the vertical width and in
all but a handful of the activities the rectangles were oriented perpendicular
to the page. Kerslake (1979, p. 34) in her study of 5–11 year olds, in which
children were asked to identify squares presented in various orientations
as squares, found that:
188 FRANK MONAGHAN

It does seem that squares were increasingly more difficult to recognise as they
became more ‘tilted’.
Parzysz (1988) also showed how students unconsciously tended to transfer
the geometrical properties of the graphical representation to the object
represented, and vice versa. Pimm (1987) contains similar examples of
such student behaviour.
That the dominance of such an orientation-driven perception continues
to be underpinned by curriculum materials must be of pedagogical con-
cern. Nor is it limited to squares and rectangles, as will become clear from
the discussion of the other question, which follows.

2. What’s the difference between a rectangle and a parallelogram?

The dominant perception here is that a rectangle is straight whereas a
parallelogram leans:
A parelelagram bends . . . (Menelick)

Parallelogram are leaning sides way but a rectangle is a right angle (Sudaratt)

A parallelogram slightly leans. But a rectangle is straight with two equal sides.
(Ellen)
Ellen’s description of the parallelogram as ‘slightly’ leaning illustrates
the degree to which previous representations of parallelograms appear to
condition a one-to-one matching of word and concept.
The question also illuminates the students’ understanding of ‘straight’,
which they appear to regard as synonymous with horizontal or vertical:
A parrallelogram is sort of leaning to one side but a rectangle is strait. (Andrew)

The difference is that a rectangle is straight and a parallelogram leans to one side.
(Matthew)

The Parallelogram side are not all stirght and equl but the rectangle side are all
stight and equl. (Celina)

A parallelogram has it’s four side but not straight but a rectangle has it’s four
side straight but not all of them are the same size and not all of them are equal.
(Yarun)
Students will sometimes use the mathematical terms ‘horizontal’ and ‘ver-
tical’ to make essentially the same point:
A rectangles edges are either horizontal or vertical and a parallelogram only has
two horizontal sides. (Jon)

A rectangle has two horizontal sides and 2 parellel side’s but a parallelogram
ha’s 2 horizontal and two side’s tillting towards the right. (Alomgir)
 CHILDREN’S VIEWS OF DIFFERENCES BETWEEN SOME QUADRILATERALS 189

As with Ellen’s use of ‘slightly’, Alomgir’s description of the rectangle

“tilting towards the right” is again evidence of the dominance of a particu-
lar representation.
Clearly, the dominant model is one in which a parallelogram is not a
parallelogram unless it looks like this:

Figure 4. Parallelogram in standard orientation.

rather than like this:

Figure 5. Parallelogram in non-standard orientation.

Similarly, students will identify this ‘leaning’ feature with diagonality:

A parallelogram has 2 short sides and 2 long sides as well but the short sides go
a bit like diagolone. (Joynub)

A rectangle is straight, a parallelogram is also straight but the sides on the right
and left are diagnal. (Del)

A rectangles sides are horizontal and vertical but a parallelograms side lines are
diagnal. (Amir)

The parallegram’s sides go vartkil and the rectangles sides go strage. (Ghalib)
Ghalib is probably using vertical incorrectly to mean diagonal and the
distinction he makes between verticality and straightness is likely to cause
him difficulties later on. Again, these descriptions illustrate the dominant
representation of diagonal lines in polygons that are themselves presented
along a horizontal axis. This is reflected in the apparent ease with which
students are able to draw in diagonal lines on a shape presented in ‘stand-
ard’ orientation, ie. with the base horizontal to the edge of the page, but
have greater difficulty when the same shape is presented in a different
orientation.
The concept of diagonality has been rendered essentially inoperative
due to the dominance of one system of representation. It has become fixed
as a single visual representation rather than as an abstract property that can
occur in a variety of modes. A separate analysis of the pattern of use of the
190 FRANK MONAGHAN

lexeme ‘diagonal’ in a mathematics scheme (Monaghan, 1997) demon-

strated clearly the predominance of the use of the term as a synonym for
oblique over its mathematical meaning so it is not surprising, perhaps, that
here again such confusion should surface in children’s writing about shape.
Arguably it even accounts for it to a considerable degree. As Triadafillidis
(1995, p. 226) argues:
The ‘problem’ in the relationship between diagram and definition arises from the
persuasiveness of visual models in the learning process. Although no generalised
diagrams exist for a given shape, there is a tendency to rely on certain prototypes
of a geometrical concept. . . . Reliance on a prototype may also lead to the inad-
vertent expansion of the mathematical definition to include additional non-critical
attributes (e.g. orientation effects).
The evidence from this discussion suggests strongly that students use terms
such as diagonal, vertical, and horizontal, as fixed and defining attributes
of specific shapes rather than as descriptors of particular representations of
individual cases. Del’s use of right and left to describe the shorter rather
than the longer sides is also noteworthy as it indicates a not uncommon
preference for ‘everyday’ English over ‘mathematical’ English and high-
lights a recurring difficulty students have in disentangling the mathemat-
ical description of sides in terms of length, width, height, perpendicularity,
diagonality, horizontality, etc.
The view that language used ambiguously in materials can exert a neg-
ative influence on children’s learning is central to Hanley’s (1978, p. 28)
perception:
Over the years, many words have been abused in mathematics, a subject where
numerous everyday words are expected to have a precise mathematical meaning
which may differ from its commonplace connotation. Sometimes these words are
casually misused in a context where the correct meaning becomes apparent at
a later stage. This may be acceptable as part of normal language development
but other words are regularly misused both orally and in print and this should
not be condoned. The printed misused word or any poor use of language has a
permanence and, in consequence, potentially such a wide audience that it becomes
a definite threat to good teaching.
This may be overly moralistic in tone – and overly simplistic in its view of
language development – but it certainly raises the issue of what teachers in
direct contact with their students are to do in terms of intervening in this
development. Hanley concludes on a positive note (p. 30):
The best school mathematics learning situations exist where the language can be
used freely as the interactive medium and the best resource for this . . . is still the
classroom teacher.
Indeed, the role of the teacher in helping students bridge both their current
conceptual understanding and their linguistic expression of it is obviously
 CHILDREN’S VIEWS OF DIFFERENCES BETWEEN SOME QUADRILATERALS 191

crucial. The authors of published materials are always going to be remote

from the child that (mis-)uses them. The teacher is in the best position to
chart the individual child’s path through this complex network and help
them re-negotiate the trickier passages.
This form of analysis should not, of course, be confined to identifying
only negative features of students’ writing. There are more positive be-
nefits to it. For example, some students reveal a richly dynamic, plastic
conception of shape in which they describe the one as a transformation of
the other:
The differance between a rectangle and a parallelogram is that the parallelogram
goes side ways and turns into a different shape and if it didn’t go side ways it
would have been a rectangle. (Razia)

Although Razia’s use of ‘side ways’ maintains the orientation stereotype,

it is nonetheless a very powerful way of visualising shapes as it allows
the students to focus on the defining points of a polygon. It is tempting to
speculate whether this sort of response would have been produced if the
question had been about a square and a parallelogram.
Other students reveal a more tactile aspect:
A parrelogram is like a rectangle except that it is pulled at two corners and the
lines are parall. (Ania)

. . . a parallelogram is like a rectangle but it is pulled at two corners . . . (Jacklyn)

It is interesting to speculate where this ‘pulling’ metaphor might come

from. One possibility is that students have had experience of constructing
polygons with interlocking straws or geostrips. A rectangle produced in
this way is not rigid and can be bent to form a parallelogram. Whatever
the source, students who are able to visualise shapes in this way have
a very useful skill for engaging in higher levels of transformations and
are less likely to have difficulty in seeing shapes not as immutable, fixed
visual entities but as fluid. They should experience less difficulty with the
conceptual integration that is required to see one shape as a special case of
another. The use of computer packages that allow students to manipulate
geometric figures on screen might also offer useful ways in to perceiv-
ing such transformations for students who currently operate more rigid
models.
Teachers need to take a positive attitude to this sort of metaphorical
language as it can reveal much about how students conceive of shapes.
Take this example from Munira:
a rectangle is a 4 sides It’s shot. a Parallelogram is sthreching and it look like a
door. (Munira)
192 FRANK MONAGHAN

She reveals an imaginative (if not entirely accurate) use of perspective

in this description of a door ajar set against the rectangular frame that
houses it. She has transferred the two shapes into a familiar setting and
transformed the rectangular door by moving it in her imagination through
space until it matches her mental image of a parallelogram (though it
would actually look more like a trapezium, of course). It is interesting to
speculate on her use of ‘shot’. If it is used to mean ‘short’ then this would
indicate the common perception of rectangles only as oblongs. However,
if it is read as ‘shut’ it can be seen as further clarification of the door
metaphor, with the rectangle viewed as a closed door and a parallelogram
as an open door. Either way, it indicates a dominance of visual perception
over mathematical abstraction.
A similar mental moulding operation is suggested by Shupi’s descrip-
tion of a parallelogram:
A parallelogram is a lettel same as a square. But parallelogram looks like if it is
came fould [folded]. (Shupi)
This sort of topological transformation is often assumed to be a higher
level skill but it is clear that students operating at lower levels are able to
make use of such skills when needing to resolve a conceptual conflict.

C ONCLUSION

Following Kress’ (1994, p. 132) view that “the child’s language may be
regarded as a window to the child’s conceptual world”, this exercise was
devised as a relatively straightforward method of capturing an insight into
students’ conceptual understanding and their ability to express it through
language.
What emerges is that students (over-)rely on standard representations
of shapes as a means of identifying and discriminating between them. It
was shown that curriculum materials tend to underpin such perceptions.
Similar effects were found by Geddes et al. (1982) in an examination of
geometry textbooks used in a high school in the United States.
This is potentially detrimental to the development of the students’ use
of the mathematics register (the mathematical use of natural language –
vocabulary and structures – to create mathematical meanings). Unless they
are forced into a Vygotskyan conflict with their current conceptual under-
standing – as evidenced by their single representation of shapes that have
many possible representations – they are unlikely to develop the higher
order conceptualisations that will enable them to recognise that trapezium
is inclusive of ‘square’, ‘parallelogram’, ‘rectangle’, and ‘rhombus’. They
 CHILDREN’S VIEWS OF DIFFERENCES BETWEEN SOME QUADRILATERALS 193

confuse a mental image with a concept and will need the mediation of
the teacher to unscramble the two as differentiated by Fischbein (1993,
p. 140):
What . . . characterizes a concept is the fact that it expresses an idea, a general,
ideal representation of a class of objects, based on their common features.
In contrast, an image (we refer here to mental images) is a sensorial representation
of an object or phenomenon.
From the teacher’s perspective, this sort of exercise gives valuable insights
into the multiplicity of ways their students conceive of shape and help them
move beyond the van Hiele level 0 and 1 conceptions demonstarted above.
It can enable them to identify which concepts are not known at this stage
(enabling them to plan at a whole-class level), or which concepts particular
students do not have, (enabling them to plan at the individual level). It can
throw light on mistaken conceptions that students may have but that do not
appear so evidently in a conflict-free exercise of simple identification. In
the context of the van Hiele model, Shaughnessy and Burger (1985, p. 425)
have pointed out that:
. . . it is very likely that the teacher and students are reasoning about the same
concepts but at different levels. While the teacher is writing a careful definition
on the chalkboard (level 2), the students may be thinking about all the properties
that the teacher has left out (level 1).
The evidence of the students’ comments discussed above clearly illustrates
the potential gap between the assumed level inherent in the materials and
that of the students. In the Geddes study cited above (Geddes, 1992) school
textbooks were analaysed in terms of the van Hiele levels they reflected.
It was found that most of the texts presented material at van Hiele level
3, but had problem sets that frequently jumped from level 0 to level 3. By
focusing on the language the teacher is able discover which aspects of the
register are in need of development.
The use of database and concordancing software enables the teacher or
researcher to make such analyses much more easily than would otherwise
be possible. For example, the information underpinning the analysis of the
ratio of sides in rectangles was generated in a matter of seconds by identi-
fying activities from containing the words ‘rectangle’ and/or ‘square’. A
manual search through all of the materials would clearly take very much
longer.
The database can also be used to identify activities which specifically
address particular shapes. A simple coding system can be devised, e.g
a new field containing details of diagrammatic features of the materials,
identifying which shapes are represented and whether in standard or non-
standard orientation. This would allow teachers, researchers, and materials
194 FRANK MONAGHAN

devisers to explore patterns of presentation and to identify opportunities for

either revising existing materials or devising new ones to address particular
concerns.
From a research perspective, the evidence from the students’ writing
suggests that a substantial concordance of student writing in the mathem-
atics classroom would provide invaluable insights into how students con-
ceptualize shape and how they might be helped linguistically to develop
their perceptions further (see Stubbs, 1996; Hoey, 1991; Sinclair, 1991 for
a broader discussion of the role of such concordances in genre analysis and
Morgan, 1998 for a discussion of how teachers’ assessments of students are
affected by their use of language). Deeper explorations of their use of such
words as pointing, diagonal, leaning, right angle, etc. would reveal import-
ant evidence about their understanding and communication of orientation,
which appears to be more frequently viewed in terms of horizontality or
verticality than as a mathematically defined line. Shaughnessy and Burger
(1985) commented on the fact that most geometry courses do not contain
materials or problems designed to help move students from level 0 to level
1, or from level 1 to level 2 in van Hiele’s terms. Dina van Hiele developed
an entire course (van Hiele-Geldof, 1957) based on the perceived need for
visual and exploratory lessons as a preparation for deduction. As Sfard
(1991, p. 3) comments:

Unlike material objects, however, advanced mathematical constructs are totally

inaccessible to our senses – they can only be seen with our mind’s eyes . . . The
sign on the paper is but one among many possible representations of some abstract
entity, which by itself can be neither seen nor touched.

The software can be used to identify activities which specifically address

quadrilaterals and comparisons made between the presentation and the
representation of such shapes found in a mathematics scheme and the
language students use to describe them. The coding system in the database
would need to be refined to take account of, for example, diagrams of
shapes and their orientation.
Curriculum statements and syllabi provide a basis for comparing the
degree of fit between the students’ current use of language and the inten-
ded curriculum requirements. Similarly, comparison of a specific corpus
of students’ usage of the register and of published mathematics texts with
a large corpus of ‘ordinary’ English (such as COBUILD) would reveal
the extent of discrepancies between the registers. Such an analysis would
enable teachers and writers of materials to build in a more explicit explora-
tion and negotiation of the varied meanings of words used by mathematics
educators and their students.
 CHILDREN’S VIEWS OF DIFFERENCES BETWEEN SOME QUADRILATERALS 195

Curriculum statements and syllabi for mathematics can also be used to

identify the need for new materials. Such resources provide an opportunity
to focus on both the language and the content by presenting the language
as content and vice versa. If an aim of mathematics is to reduce commu-
nication to the maximum then this sort of exercise may be a way towards
fulfilling that aim.

ACKNOWLEDGEMENTS

I would like to thank my colleagues and students at North Westminster

Community School, London who took part in this study. I would also like
to thank the anonymous reviewers of earlier drafts of the paper for their
valuable criticisms and suggestions for improvements.

R EFERENCES

Adler, J.: 1999, ‘The dilemma of transparency: Seeing and seeing through talk in the
mathematics classroom’, Journal for Research in Mathematics Education 30(1), 47–64.
Adler, J.: 1997, ‘A participatory-inquiry approach and the mediation of mathemat-
ical knowledge in a multilingual classroom’, Educational Studies in Mathematics 33,
235–258.
Crowley, M.: 1987, ‘The van Hiele model of the development of geometric thought’, in
M.M. Lindquist and A.P. Shulte (eds), Learning and Teaching Geometry, K-12, National
Council of Teachers of Mathematics, 1987 Yearbook, Virginia, pp. 1–16.
DES: 1989, Mathematics in the National Curriculum, DES, London.
De Villiers: 1994, ‘The role and function of a hierarchical classification of quadrilaterals’,
For the Learning of Mathematics 14(1) (Feb. 1994), 11–18.
DFE: 1995, The National Curriculum, HMSO, London.
Ernest, P. (ed.): 1994, Mathematics, Education and Philosophy. An International Perspect-
ive, The Falmer Press, London.
Fischbein, E.: 1993, ‘The theory of figural concept’, Educational Studies in Mathematics
24, 139–162.
Geddes, D., Fuys, D., Lovett, J. and Tischler, R.: 1982, ‘An investigation of the van
Hiele model of thinking in geometry among adolescents’, Paper presented at the Annual
Meeting of the American Educational Research Association, March 1982, New York.
Gipps, C.: 1994, Beyond Testing: Towards a Theory of Educational Assessment, The
Falmer Press, London.
Hanley, A.: 1978, ‘Verbal Mathematics’, Mathematics in School 7(4), 27–30.
Hasegawa, J.: 1997, ‘Concept formation of triangles and quadrilaterals in the second
grade’, Educational Studies in Mathematics 32, 157–179.
Hershkovitz, R.: 1990, ‘Psychological aspects of learning geometry’, in P. Nesher and J.
Kilpatrick (eds), Mathematics and Cognition, Cambridge University Press, Cambridge.
Hoey, M.: 1991, Patterns of Lexis in Text, OUP, London.
Kerslake, D.A.: 1979, ‘Visual mathematics’, Mathematics in School 8(2), 34–35.
196 FRANK MONAGHAN

Kress, G.: 1994, Learning to Write (2nd edition), Routledge, London.

Laborde, C.: 1994, ‘Enseigner la géometrie: permanences et révolutions’, Proceedings of
ICME 7 Quebec, 1, 47–75.
Mercer, N.: 1995, The Guided Construction of Knowledge: Talk among Teachers and
Learners, Multilingual Matters, Clevedon.
Monaghan, F.: 1997, Language and Mathematics: The Analysis of Written Text in the
Secondary Classroom, Unpublished PhD thesis, University of Reading.
Morgan, C.: 1998, Writing Mathematically: The Discourse of Investigation, The Falmer
Press, London.
Parzysz, B.: 1991, ‘Representation of space and students’ conceptions at high school level’,
Educational Studies in Mathematics 22, 575–593.
Parzysz, B.: 1988, ‘Knowing vs. seeing. Problems of the plane representation of space
geometry figures’, Educational Studies in Mathematics 19(1): 79–92.
Pimm, D.: 1987, Speaking Mathematically, Routledge, London.
Sfard, A.: 1991, ‘On the dual nature of mathematical conceptions: reflections on processes
and objects as different sides of the same coin’, Educational Studies in Mathematics 22,
1–36.
Shaughnessy, J.M. and Burger, W.F.: 1985, ‘Spadework prior to deduction in geometry’,
Mathematics Teacher 78(6), 419–427.
Sierpinska, A.: 1994, Understanding Mathematics, The Falmer Press, London.
Sinclair, J.: 1991, Corpus, Concordance, and Collocation, C.U.P., Cambridge.
Stubbs, M.: 1996, Text and Corpus Analysis, Blackwell, Oxford.
Triadafillidis, T.A.: 1995, ‘Visual limitations in geometry’, Educational Studies in Math-
ematics 29, 225–233.
van Hiele, P.M.: 1984, ‘A child’s thought and geometry’, in D. Geddes, D. Fuys and R.
Tischler (eds), English Translation of Selected Writings of Dina van Hiele-Geldof and
Pierre M. Van Hiele, part of the research project, ‘An investigation of the van Hiele
Model of Thinking in Geometry among Adolescents’, Research in Science Education
(RISE) Program of the National Science Foundation, Grant No. SED 7920640.
van Hiele-Geldof, D.: De Didaktiak van de Meetkunde in de eerste Klas van het V.H.M.O.,
Utrecht, Ph.D. Dissertation, University of Utrecht.
von Glaserfeld, E.: 1995, Radical Constructivism: A Way of Knowing and Learning, The
Falmer Press, London.
Vygotsky, L.: 1986, Thought and Language, MIT Press, Cambridge, Mass.
Vygotsky, L.: 1978, Mind in Society: The Development of Higher Psychological Processes,
Harvard University Press, London.

F RANK M ONAGHAN
Literacy & Numeracy Coordinator/Professional Tutor
North Westminster Community School
Penfold Street
London NW 16 RX
United Kingdom
E-mail: kca06@dial.pipex.com