Towards A Systemic Functional Analysis of Multisemiotic Mathematics Texts

Towards a systemic functional analysis of
multisemiotic mathematics texts
KAY O'HALLORAN
Introduction
Mathematics may be viewed as a semiotic construction as, for example, in

Marcus's (1997) recent examination of the metaphorical nature of Claude
Levi-Strauss's symbolically expressed canonic formula of myth, and others
such as Rotman (1987, 1988, 1993), Coleman (1988), Mortensen and
Roberts (1997), and Thorn (1983) who argue for a semiotically orientated
philosophy of mathematics. Discourse analysis of mathematical texts must
therefore necessarily take into account the multisemiotic nature of its
makeup. That is, mathematical discourse involves the codeployment of the
resources of natural language, mathematical symbolism, and visual display
in the form of diagrams and graphs. M. A. K. Halliday's systemic func-
tional model of language offers a comprehensive theoretical framework for
investigating such semiotic systems as demonstrated by O'Toole's (1990,
1992, 1994, 1995) systemic model for the visual forms of painting, archi-
tecture, and sculpture. Other extensions of the functional approach include
visual images (Kress and van Leeuwen 1990, 1996), music (van Leeuwen
1988), and preliminary observations on the multisemiotic nature of scien-
tific discourse and mathematics (Lemke 1998; O'Halloran 1996).
Following Lemke (1998), mathematical texts are necessarily multi-
semiotic as the systems specific to each semiotic differentially construct
reality. That is, although semantic commonalities may occur, the meanings
realized by one semiotic cannot be exactly replicated by another. In what
follows, I discuss the functions of mathematical symbolism, visual display,
and language from the perspective of the evolution of modern mathemat-
ical symbolism from linguistic and visual forms. The need for such
investigation of the changing roles and functions of each code has also been
expressed by those seeking a semiotically based account of the nature and
foundations of mathematics (Coleman 1988; Mortensen and Roberts 1997,
for instance). Following this, I present preliminary systemic frame-
works which may be viewed as a first step in the construction of systemic
Semiotica 124-1/2 (1999), 1-29 0037-1998/99/012-0001

Brought to you by | Stockholms Universitet
Walter de Gruyter
Authenticated
Download Date | 12/4/16 1:25 PM
2 K. O'Halloran
grammars for mathematical symbolism and visual display, with the latter
based on O'Toole's (1994) systemic model for paintings. Finally, I discuss
the joint construction of meaning in the mathematical texts through an
examination of the process of translation between codes, which gives rise to
the phenomenon of semiotic metaphor.
Mathematics as a bridge between linguistic descriptions and

perceptual reality
Various forms of mathematical symbolism evolved from natural language

and, in some instances, visual representations, to fulfill particular functions
and, as Joseph (1991) makes clear, historically these developments were not
confined to the Western world. However, in the efforts to solve practical
problems arising from the political and economic interests of seventeenth-
century Europe, modern mathematical symbolism evolved to bridge the gap
between perceptual reality and linguistic descriptions. That is, mathemati-
cians such as Descartes (1596-1650) and Fermat (1601-1665) became con-
cerned with investigating curves like ellipses, parabolas, and hyperbolas
which described phenomena of the physical world such as the paths of
planets, comets, and projectiles. These curves were important for solving
immediate practical problems such as those associated with warfare, navi-
gation, and trade. In investigating these curves, the idea was developed that
'to each curve there belongs an equation that uniquely describes the points of
that curve and no other points' (Kline 1972: 198). Before this time, it is
reported that algebraic symbolic notation was in some state of disarray,
fulfilling no obvious purposeful activity. For example, Kline (1972) reports
Descartes as explicitly criticizing algebra 'because it was so completely
subject to rules and formulas "that there results an art full of confusion and
obscurity calculated to embarrass, instead of science fitted to cultivate the
mind"' (1972: 193). From Descartes's links of the equation to curve, the
study of motion and change was independently developed by Newton and
Leibniz. This represented a major extension in mathematical activity since
'previous mathematics had been largely restricted to the static issues of
counting, measuring and describing shape' (Devlin 1994:2). That is, the link
from text to visual was achieved with the development of Cartesian geom-
etry and calculus where the 'grammatical metaphor' in the form of sym-
bolism was linked to the 'visual metaphor' of the abstract diagrams and
graphs.
Galileo's (1564-1642) plan for studying nature through quantitative
mathematical description (Kline 1972) had directed Descartes's explora-
tions in mathematics and science. A scientific revolution (Khn 1970)

Authenticated
Multisemiotic mathematics texts 3
followed in which quantitative mathematical descriptions of the material

world replaced physical explanations of phenomena (Kline 1972, 1980;
Wilder 1981). Science was no longer to be based on metaphysical,
theological, and mechanical explanations of the causes and reasons for
events in the material world. The new goal of science was to seek mathe-
matical formulas to describe phenomena independently of explanations.
However, the path to the 'unified' discipline of modern mathematics reveals
the discontinuous nature of mathematical knowledge (Foucault 1970,
1972) with shifts in theoretical paradigms (Azzouni 1994; Grabiner 1986;
Kline 1980; Tiles 1991; Wilder 1981) and intense rivalry over forms of
mathematical notation as documented by Cajori (1927,1952, 1974, 1991).
From a contemporary viewpoint, following Lemke (1998), natural
language primarily realizes typographical modalities or categorical
descriptions, while mathematics realizes topological modalities or
descriptions of continuous variation. Thus the descriptive power of
mathematics outstrips the potential of language in the field of continuous
covariation and descriptions of relations of parts to a whole. However,
although the symbolism allows for complete descriptions of these relations,
trends and patterns which are present in these formulations are often dif-
ficult to discern. The visual display of symbolic notation in the form of
graphs and diagrams allows these trends and patterns to be revealed per-
ceptually (Lemke 1998). However, these visual patterns are only partial
descriptions which are further limited in terms of manipulative and cal-
culatory power. As Lemke (1998) explains, the symbolism is thus more
powerful but less intuitive than the visual displays.
Modern mathematics evolved as a written semiotic and so may be con-
textualized with respect to the semantic space occupied by written and
spoken language. Halliday makes the point that speech and writing dif-
ferentially represent reality. 'Written language represents phenomena as
products. Spoken language represents phenomena as processes' (Halliday
1985: 81). Mathematical symbolic descriptions may be related to the costs
involved in which written texts construct a synoptic world of things and
their relations while oral texts construct a dynamic world of happenings
and processes. Halliday formalizes the cost of written language as 'some
simplifying of the relationship among its parts, and a lesser interest in how it
got the way it is, or in where it may be going next' (Halliday 1985: 97). On
the other hand, the cost of the dynamic view is 'less awareness of how things
actually are, at a real or imaginary point of time; and a lessened sense of how
they stay that way' (Halliday 1985: 97). Mathematical symbolic descrip-
tions are concerned with dimensions of meaning which occur in the dis-
junction between these forms of language. That is, mathematics is
concerned with capturing continuous patterns of variation and relations of

Authenticated
4 K. O'Halloran
parts to the whole which reveal the status quo at all points of time.
Mathematics captures exact dynamic descriptions of relations as things
frozen in time through the lexicogrammar of mathematical symbolism.
Mathematical symbolism
Any account of the contemporary semantics and grammar of mathematical

symbolism as distinct from natural language should be considered from an
historical perspective. Initially, mathematical texts were written in the
prose form of verbal 'rhetorical algebra'. This was later followed by
abbreviations for recurring participants and operations in what has been
called 'syncopated algebra'. In the last five hundred years, the use of vari-
ables and signs of operation resulted in 'symbolic algebra' and hence the
development of the lexicogrammatical systems specific to modern mathe-
matics. That is, the contemporary form of mathematical symbolism grew
directly out of the lexicogrammar of natural language, which explains the
integration of symbolic and verbal forms in mathematics texts. In its evo-
lution, however, it appears that mathematical symbolism adopted only
particular selections from the meaning potential of language. This con-
traction of options meant that eventually maximal structural condensation
could occur unambiguously within the symbolism while allowing semantic
extensions that outstripped the meaning potential of language. However, as
a consequence of the nature of this contraction and expansion, as a semiotic
system mathematical symbolism never gained an overall functionality and,
as a result, requires codeployment with language and visual display, with
the former also acting as the metalanguage of mathematics. The limitations
of the semantic dimensions of mathematical symbolism may be captured
through discussion of the comparisons of the restricted lexicogrammatical
selections from the English language which are symbolized in mathematics.
Expansion of experiential meaning
Invoking Halliday's (1994) system of TRANSITIVITY, semantic exten-

sions in mathematical symbolism perhaps first took place experientially
through the introduction of new participants and processes initially
afforded by the numerical systems. That is, numerical systems resulted in
new combinations of symbols for previously unconceived, quantities as in
the case of extremely small or extremely large numbers, and intermediate
integer and fractional quantities. The systems also allowed for a new pro-
cess type which I have called Operative'. With the introduction of these
processes, the semantics of Material processes of combining and increasing,

Authenticated
decreasing and sharing physical objects were replaced with arithmetical

notions of adding, multiplying, subtracting, and dividing, respectively.
Operative processes are actions performed by human Agents on or between
the mathematical objects such as numbers and later, variable and 'abstract'
quantities. Operative processes therefore include the four arithmetic
operations of addition, subtraction, multiplication, and division and later
algebraic mathematical operations such as those found in advanced
and higher mathematics. The human Agent is most commonly deleted
in contemporary mathematical symbolism. The notion of agency in
mathematical symbolism is further discussed later on.1
It is possible to conceive Operative processes as metaphorical realiza-
tions of existing processes found in natural language. Indeed, my analyses
of oral mathematical classroom discourse (O'Halloran 1996) demonstrate
that linguistic selections suggest metaphorical shifts between Material
processes (a process of doing) and Mental processes (a process of thinking).
While these differing interpretations are not problematical in the context
of classroom discourse, neither of these categories exactly describes the
nature of the mathematical processes which appear in the mathematical
symbolic text. That is, Operative symbolic processes appear be semantically
different from other processes described in Halliday's (1994) system of
TRANSITIVITY. For instance, there appear to be multiple Mediums
connected with Operative processes as opposed to the notion of a single
Medium found in the process types of natural language. Halliday (1994)
describes a Medium as being central to a process: 'Every process has
associated with it one participant that is the key figure in that process: this
is the one through which the process is actualised, and without which
there would be no process at all. Let us call this element the MEDIUM,
since it is the entity through the medium of which the process comes into
existence' (1994: 163). For illustrative purposes, we may consider this
mathematical statement.
-b Vb2 - 4ac
X=
2a
This statement realizes a Relational Identifying process through the

symbol = with Token (Agent) and Value (Medium) (b Vb2 4ac)/2a.
However, if the right-hand side of the equation is considered as a case of
clausal rankshift involving nuclear configurations of the processes of
addition, subtraction, multiplication, square root, and division with par-
ticipants a, ft, c, 2, and 4, then it no longer becomes possible to consider this
solely as one lexical item. When these processes are considered, it is difficult,
however, to identify one Medium through which there would be no process

Authenticated
6 K. O'Halloran
at all. For example, if one considers 4ac as 'four multiplied by a multiplied

by c* (or alternatively, consider 'a times b times c') then the notion of a
central participant becomes increasing difficult to apply. This brings into
focus the question of the differences in the nature of reality construed by
language and that construed by mathematical symbolic notation and
highlights the need for further investigation. If considered from the point of
view of the ergative interpretation of language (Halliday 1994: 163-172),
we are perhaps dealing here with semantic differences whereby experien-
tially the descriptive forms of symbolic reality are such that notions of
causation and agency, central to our linguistic constructions of reality, do
not apply in mathematical symbolic descriptions. With the development of
a comprehensive lexicogrammar of mathematical symbolism, the seman-
tics of Operative processes and their accompanying participant functions
may be fully explored.
Contraction of experiential and interpersonal meaning
The contraction of meaning potential of language in mathematical sym-

bolism occurred in the realms of experiential and interpersonal meaning. In
the first case, options for selections of experiential meaning as described by
Halliday's system of TRANSITIVITY are dramatically reduced in mathe-
matical symbolic notation. Mathematics is concerned with descriptive
statements involving a restricted variety of processes (Relational, Exis-
tential, and Operative) and, as a direct result, fewer participant functions
which include Token and Value, Carrier and Attribute, Existent and
Operative participants. There is no longer a requirement to indicate the
functional descriptive category in mathematical symbolism through the
use of prepositions, as, for example, 'for' to indicate Beneficiary. This
functional consistency of the participants and processes extends to the
narrowing of options of circumstantial elements which are expressed
symbolically. As mathematics is concerned with descriptive as opposed to
explanatory statements, Halliday's circumstantial elements of Manner
(realized through prepositions such as 'with', 'by', 'like'), Role (realized
through the preposition 'as'), and Cause (realized through prepositions
such as 'through' and 'for') are not usually relevant. On a cautionary note,
however, exceptions do occur and prepositions are found in hybrid forms
of mathematical symbolic statements.
Other condensatory lexicogrammatical strategies in mathematical
symbolism include dispensing with rankshifted prepositional phrases
through the inclusion of circumstantial elements of Extent, Location, and
Accompaniment in participant and process structures. For instance, the
circumstantial element of Extent may be realized in the participant

Authenticated
structure as, for example, in the case of the representation of 'after t sec-
onds' in the participant structure s(t). Other strategies include a restricted
range of experiential components in the nominal group with an absence of
Deictic selections and experiential (and interpersonal) epithets. In addition,
the Thing is often represented by conventionalized symbols where the
participant function is apparent at first glance (for example, for the
independent variable). The latter suggests that another major con-
densatory strategy includes the use of pronouns to replace lexical items. As
I discuss below, in combination with restricted interpersonal meaning,
the rigorous ordering of fewer functional components in mathematical
symbolism allowed accessible and intelligible conventions to be developed.
Significantly, as mathematical symbolism evolved as absolute descriptive
statements, whole areas of interpersonal meaning found in language dis-
appeared. That is, the English language realizes different speech functions
(statement, interrogative, question, command) with a range of modalities
through a changeable order and selection of functional units in the Mood
structure. This range of options does not exist in mathematics since the
symbolism is concerned with descriptive statements and, in a restricted
sense, commands which consistently select for maximal modality and
modulation. This means that the discourse is imbued with a sense of
absolute certainty and obligation. There is no equivalent of the Finite to
realize plays with tense or modality and Polarity is realized by the same
symbol that represents the verbal group with negative polarity usually
indicated by a slash. Together with the lack of Mood Adjuncts to realize
probability, usuality, or inclination, interpersonal meaning in terms of
attitudinal lexis is absent. Mathematical symbolism consists only of pro-
nouns and core lexical items (Carter 1987). That is, there are no shades of
meaning derived from selection of lexical items which culturally occupy a
non-central position.
At the level of discourse semantics, the mathematical symbolism con-
tains a series of statements. Following Martin's (1992: 529) indicators of
status, together with the linguistic selections described above, the nature of
Exchange structure (Martin 1992: 31-91) as a series of moves by a 'primary
knower' positions the text as dominating. This interpersonal meaning is
also communicated through the style of production and contrasts in scripts
and font in stylized computer-generated texts.
Restricted logical meaning
Mathematical discourse is concerned with only selected logic which post-

structuralist thinkers (Azzouni 1994; Tiles 1991, for example) have argued

Authenticated
8 K. O'Halloran
is not entirely deductive in nature. By definition, deductive logic is con-

cerned with two premises and a conclusion, all of which are categorical
propositions. It is argued, however, that mathematics is a collection of
algorithmic systems where the system is generated from a postulate base
and includes sets of derivable sentences which are not based on syllogistic
reasoning alone. However, the explicit statement of the algorithmic steps
involved in solving a problem or the derivation of a result gradually became
implicit as the grammar of the symbolism developed and the results became
generalized. The algorithms became implicit procedures using established
results derived from a recursive set of axioms and a recursive set of inference
rules. For instance, Tiles (1991) gives the following example, which cannot
be represented by syllogistic logic alone.
Consider the function/(jc) = 4x 4.x2

Since 4x 4x2 = 4x(\ x), we can easily see that/(;c) will take the value 0 when
jc - 0 and when = 1 (1991: 34)
If I represent this example using mathematical notation, the resulting

form would be something like this.
Problem : Algebraically find the zeroes off(x) =4x- 4x2

Solution : J[x) = 4x-4x2
Q = x(\ -x)
.'. = 0 or = 1
i.e. the zeroes off(x) = 4x 4x2 are = 0 and = 1
Apart from the initial formalization of the system when the postulate
basis is first established, the symbolism realizes a mathematical description
of the result after each algorithmic step, but no explicit statement of the
algorithm occurs in the mathematical symbolic text. The implicitness of the
deductive and operative relations in generic steps in the mathematical
symbolic texts cause problems in an educational context as often there are
long implication chains that give little or no indication of the results,
definitions, axioms, operational properties, or laws that have been used.
The analysis of the logico-semantic relations for clause complex rela-
tions (Halliday 1994: 215-273; Martin 1992: 179) in classroom discourse
(O'Halloran 1996) reveal that the most common types of logical relations
are consequential relations (for example, 'so', 'if, 'because', 'so as'), tem-
poral relations (for example, 'then'), and additive relations (for example,
'and'). Given the field-structured nature of the pedagogical discourse, the
occurrence of the latter relations is not surprising. However, the nature of
the logical relations is predominantly orientated towards consequence,

Authenticated
either as occurring 'naturally' or as a consequence of some predetermined

condition as described earlier.
Specification of textual meaning
The textual organization of mathematical texts is highly conventionalized

with strategies which include the spatial arrangement of sequential mathe-
matical statements and the ordering of participant functions. From the
system of THEME (Halliday 1994: 37-67), the notion of a Theme as the
point of departure of a message and Rheme as the part in which the Theme
is developed is highly organized to the point of being generic. Referential
cohesion for the tracking of participants, however, depends on the implicit
deductive and operational relations with participant configurations being
transformed through axiomatic definitions, derived results, operational
properties, and repetition of reference. For example, if we consider the
previous problem involving f(x) = 4x 4x2, the Theme,/(x), is the left-
hand side of the equation and the Rheme, describing the function is the
right-hand side. In the second line, the Theme is 0 after/(jc) is substituted
according to definition of the zero of a function. That is, a zero of a function
is 'a value of the argument of a function at which the value of the function is
zero' (Borowski and Borwein 1989: 641). In the second line, the Rheme is
transformed into x(\ x) according to established results for extracting a
common factor. Experientially, the positionings help track the altered form
of the participants. In line three, there are two mathematical statements on
the one line. In each case the Theme is x, the zero of the function with values
0 and 1 given as Rheme. As often occurs in mathematical texts, the fourth
line is a hybrid of mathematical symbolism and language where lexical
items are cases of rankshifted mathematical symbolic statements.
The graphology of mathematical symbolism
The contraction, condensation, and extension of ideational, interpersonal,

and textual meaning of mathematical symbolism is encoded in the gra-
phology at the level of the language plane. The forms of symbols used in
mathematics include letters of the Roman alphabet with upper- and lower-
case letters of varying sizes written with varying fonts and scripts, the Greek
alphabet, and a limited range of other alphabets and abbreviations. Other
sources include punctuation symbols, brackets, iconic representations, and
the invention of new symbols. Mathematical symbolic text is usually
written in italic font when embedded in language co-text.

Authenticated
10 K. O'Halloran
Spatial positioning in combination with serial positioning plays an

important role in the semantics of mathematical symbolism. This includes
textual and ideational meaning in the spatial layout of the entire mathe-
matical text (genre), in the sequential ordering of statements within that text
(discourse semantics), in the serial ordering of functional elements within a
statement (lexicogrammatical) and, possibly its most significant departure
from language, in experiential meaning as dependent upon spatial position
(for example, x?). Also, serial notation in the structure of the nominal group
follows conventions which differ from language as, for example, in the case
of the decimal place value system. In combination, the different types of
positional and serial notation allow an economy of expression that is
impossible in language and thus make available an entirely new meaning
potential. This is further evidence that mathematical expression has
incorporated and built on resources from both language (sequential and
serial positioning) and visual display (spatial positioning).
A framework for a systemic functional grammar

of mathematical symbolism
Table 1 gives a descriptive framework for a systemic functional grammar of

mathematical symbolism.
As the preceding discussion relates to these systems, I limit my discussion
to a brief reconsideration of the experiential meaning for the symbolic
statement, = (-b Vb2 - 4ac)/2a. The nuclear relations at the level of
the discourse semantics (Martin 1992: 319) for this statement are displayed
in Table 2. What is immediately apparent is the narrow range of processes,
the multiple levels of rankshift and the absence of a Periphery realizing
Circumstance. The process type is either Relational Intensive Identifying
( = ) or Operative (, , ^, with the fraction line realizing division and
implicit encoding of multiplication).
The condensatory strategy of the multiple rankshift of Operative pro-
cesses is in part possible because of the conventions regarding the order of
operations (the sequential order of brackets, powers, and multiplication/
division and addition/subtraction respectively), ellipsis of the Operative
process of multiplication, and the use of spatial graphology. In addition,
condensation is achieved through the nature of the components in that
there is maximum lexical density as each unit is either a lexical item or a
pronoun. The grammar of mathematical symbolism thus functions to
condense and compact experiential meaning while dispensing with
peripheral information where possible. Together with the dimensions of
interpersonal and experiential meaning, this lack of peripheral information

Authenticated
SU
^
<
t
5
H
Q

^
I
as
0}
I
s
unctional grammai
S
&
'S,
^
%
's
^
1
1
j
J

%
<
XPERIENTIAL OGICAL :NTERPER
fl
04
00
W
<
P
H-J O
MTER-STATEMENTAL
11
Discourse Semantics)
(U
.
c

Collocation (lexical relati Conjunctions an Speech ftu Ltial cohesion
So

Ic
~
8
8
and strings through taxo: conjunctions (im and limite on definition
,
'
<1J
CO

43 -g 'U
>-i ~ ..=.
< .55 O

S3 o 5
So 'g SI
definitions, axioms, theoi generalization, d command^ onal properti
111

4)
CO
CO
0

axiom, operatior repetition
8.
I
*-> J-H

2 e &
-5
property, operati ence)
11
. 2 0 ^
.2
deduction, and ii
^
g
1
c
0
0

s|| s s I

lal notation
>S
Nuclear relations (consist Implicit and exp!
C
>
52
Ofl O
'ii
13
S
8
luential dowr
^
of Participant and Procei conjunction
^
a S
-t-
?nt of statem
1 5- B
sitional place
rial componei
G ^ >
CO
<4-

CO
(
"
u
Q

*>
G ^%

8 ' fi S

^
DH ^ 113
Sequential placei ition (operati

'!
Activity sequences
(progressive steps of statements representa
simplification and solutic
"a"
|
O
^j
'
G
00

Positional notation Extension of tax lent clauses
if
^ <u
WH
t**
a
^
^>S- c
S
^ c
'S *
8
(indicating continuations long implication tic or spatiall
>.
il
?
^
CO

8
strings and activity seque
'&
CO
CO
I) 52
c^

o<
Lexis Apposition
t c .
1
to fl
Si
.^
G
CO
c3
(4-1
^
2

Authenticated

s
Parenthesis
^
1
22 8
a
Multisemiotic mathematics texts
1
1
^2 ""*
g o g
3 < II iS

111
s-i32! || I
11
12 K. O'Halloran
S
'*
_
Cd - <
S^ - *5" td '*3
** (L) CX ^
o 3 -S g ^ ^ 8-
J " o 8* S g |
S .H ^ g S -2 |
t t litiifg!
EXTUAL
<
1 1. t . .1-tlllSI
H
^ S &
c S- to co
1 l3 i 8
s i !i1
Z 0 P
^
Production pr
INTERPERSO
participants/p
Abstractness
(nature of
r; >^ ,1 *-*< --H .Jii
" c d > ^ 2 t ^ ^ c o o j
1 li *1P
2 . t cd!2 o ^ c d N
"8
w CU > N. ' c/3 Ci' Vj cd
So
cd
p
So
d

c
cd
CO 9

J
1!*
U
HH 12
8
H-l
U

*

vE

____^ cd d
(30 rt C
d
3 JL-S ?
Attributive/Intensive
(Relational-Identifyi
TABLE 1 (Continued)
Circumstantial and
^ S^ ^ 5
STATEMENTS //
^^^3^0
EXPERIENTIAL
'S "32
^S'Sg2"
.M l? S -S f
Processes
8s .2Itl 1 s $
(Clause)
t: ts P .S 3 'g
O ^ c d W - S cdS
, , ^ u sE-


Authenticated
~-J

Operative processes y Conventional order of
|
<L>
co
2
c s
"3-5
2

8-1
Degree of embeddedness
II
and explicitness operations
5
J8
'S P (4-!i
(indicated by maximal realized dding) (which may be changed
|f l

number of brackets conventi through use of differing
s|

<u
CO
*8
0
CX
from the outside to ration types of brackets)
1
the innermost
embedded process)
o>

Participants (often the of abstractness
<u
<D l*
result of embedded of participants and
Operative processes)
a
^c
li
I
.&
Circumstantial elements
11 cd
"lo
^ .8

(fused with participant)
COMPONENTS
(Atoms)
.
CO

d
G
CO
c
Q
c
Restricted range of units i Convent
&
1
1
1
"cd

I cLer

the nominal group combina (spatial, serial position,
(the absence of Deictics, and brackets)
attitudinal, and
experiential Epithets)

_c
c
Q
c

Qualifiers Contrast
1
Ul
<*
cd
_"cd
cd
0
0
I

538

Authenticated
group without the need
for embedding as phrases]
CO
CO
I
Classifiers
0>
11
Conventionalized use of
specific symbols
14 K. O'Bailoran
CN ^! =
-H
I
a>
HH
7 a2
M
2

Authenticated
contributes to the view of mathematics as universal truth when in fact

its 'truth' is contextually dependent on the system within which it is
constructed.
Visual display
The emphasis of visual forms is directed towards the representation of

speech in western culture (Kress and van Leeuwen 1990). This may be
compared to other cultures; for example, Aboriginal culture, where
drawings, sand-paintings, and carvings hold an equally important place
alongside verbal language as meaning-making practices. This bias is
reflected in the relative status of the different semiotic systems in mathe-
matics where there is a traditional prejudice against the use of visual display
amongst mathematicians and philosophers (Shin 1994). Diagrams, for
example, are considered as a heuristic tool only and not as a means of
establishing a valid proof. Shin (1994) believes there are two reasons for
this; first, the limitations in presenting knowledge and, second, the possible
misuse of diagrams as can happen with making unwarranted assumptions
in geometry. While agreeing that diagrams cannot represent all that is
desired, Shin (1994) nevertheless argues that the negative prejudice against
diagrams is not warranted on the basis of these two arguments. Despite this
prejudice, as Lemke (1998) explains, visual representation plays an
important role in displaying trends and patterns which are otherwise
indiscernible in mathematical symbolically expressed relations. The visual
display is thus important for intuitive understanding of the reality
constructed through mathematical symbolism.
A framework for a systemic functional grammar of

mathematical visual display
Using O'Toole's (1994) systemic functional model in The Language of

Displayed Art, the systems for mathematical visual display in Table 3 are
metafunctionally organized according to rank with the Modal, Repre-
sentational, and Compositional functions corresponding to interpersonal,
ideational, and textual functions of language. Following O'Toole's (1994)
formulations, Diagram/Graph is the whole representation resulting from
interactions or Episodes of Figures which are composed of Parts.
As inferred from O'Toole's (1994) comments on abstract art, the
cultural styles of mathematical diagrams and graphs could be traced

Authenticated
00
Cd 00
T3 00 C
u 3 cd
-C c3 d,
M
8 a ^S
</T C * G "g
ri n
5 O Q U S G^ ^
C cd m T3 g .52 o

SS Q
O
* -2 1 c _G
O as'S
^
28 .2-2
| 12
OO "rj C '~
P 00 <
C2
8
Position!]
G
COMPOSI'
00 ^
2 S /-
sS S*s -s
c cd
I "o V g>Jj
>
g S S 2 n
D -5 S U 11!

1 i
00
^
.
1

'S * c
* cd ^
'S
(

c J? ^ S
O
SX s" s" G rt
C/3
43
2
h;
S c c 5
"^
00 O .G O >
hi
S^>
G '+2 0 U
1
"^
c 0,2 *a
^ cd s? O (U
G
|
| 00
" G
uT 00 J3
MODAL
. 11
0
2
J | cd ^"c
<, Ci i-J X5 Cd u11 UH
S I
1 -3 aG ^T 2l ^

t!0 .. ^co s s
rt c^
nd systems in spatial represem
Mathematical Symbolic 1
Topological modalities
continuous co-variations
(for example, geometrica
(for example, Venn diagr
S
>
-
cd
2Sg
. 2 0, T3 ^
a* 1 litt
REPRESENTATION
Perceptual Reality
Representation of:

g^ i S
iiiiifiii
data graphs)
t .. G cd <Z ^-H G
^^2i3Sg&^--g
&8Hrt^9c S|-c
^ 0 - CN JS U >
<3
03
s:
S

S
^
v<
CO
~ in
3 'S
f2 5
Authenticated
i <U .u
i-l rxs
3
co
CQ
C/3
(
cd
c
,c C EO
00
3 1 ss
>> 'S
o -^ T3 hfl cd
S S - l *
!'isl
B ?

M 0
D >< '^3 .2
ill -si
8^
S S
fOH
, , ^
jfj*
1 I| p^ P (
.S
J 12 2
^
1 ^
ss ^5
2
-B M
II 00
-S
00
00 C
.S 5 131 ?1 1
"cl 00 * 'S ^
>,c ^
* ^ 8 ^ 2 ?n
C/ C
5 3l O cd -1~ t4_
OH PH ^ 0
c O O -G <U S
-6 U U c* fti o

2
t~o 00
IS (
8
VH
C*
O fll
>% ^ 8
o
2 O , >
4> O \3 , 3
lo.-gl l 8g o |
-5 ^ cd
o .& C! 5 T3
.
1 11
"cd
O CX S c' '-3
c o ig c8 8

o-2|S o c
<s a s a ia . 1 ll s
3
^ u
1 S's S2 1 1
cd
U g
Qd
.52 -S
I S
1.
icd 3
- 2
o
J_l
G " 00
^ 3
,> 8S -Scd Si 00 LL
l^
S "^ 5 "C U 12
"cd ? : S. s 's
c
.2
"o ^
<S ^ Q^ <yj
Intersection '.
Lines, Curve
> Circumstanti

Slope of Par
> Portrayal of
with Process
oi
o1
111
1 1 2
<>

<
PISODE
IGURE
pq [^T I <S
i

Authenticated
18 K. O'Halloran
historically to reveal metafunctional tendencies which have resulted in the

conventionalized forms of contemporary visual representation. In this
case, the process seems to have paralleled that found in mathematical
symbolism whereby simultaneously there is a contraction and expansion of
the meaning potential of visual display. That is, as mathematical symbolism
selected for contracted interpersonal meaning, expanded circumscribed
forms of experiential meaning and refined textual meaning, similar trends
occurred in visual mathematical representations with the final result being a
monofunctional tendency oriented towards Representational meaning.
In the realm of the interpersonal, Modal meaning developed to a con-
tracted and direct form. That is, strategies for engaging the viewer of the
mathematical visual display do not operate through nuance as found in
forms of art, but rather select for a direct unmarked command, 'look here'.
With exceptions, such as clines of definiteness associated with lines (from
bold to solid, to dashes to dot-dashes, to dots), the use of Labels, Titles, and
explicitly marked features direct the viewer's gaze to certain features of the
diagram with little or no discretion. Another feature of the contracted
interpersonal meaning is the maximal value of the Modality. As Kress and
van Leeuwen (1990: 52) explain:
visual modality rests on culturally and historically determined standards of what is

real and what is not, and not on the objective correspondence of the visual image to a
reality defined independently of it.
Hence, although mathematical representations clearly do not corre-

spond exactly to material reality, they nevertheless function as 'truth' since
the Galilean reality of science underlies visual Modality; that is, reality
based on 'size, shape, quantity and motion' (Mumford 1934, quoted in
Kress and van Leeuwen 1990: 53). The maximal modality is encoded
through the Degree of Idealization, Abstraction, Quantification, and the
conventionalism of the display. In combination with the verbal discourse
and symbolic statements, the visual displays function to realize the 'truth'
of mathematics.
The viewer's gaze is thus unequivocally directed to certain dimensions of
mathematical visual display and, as opposed to individualization through
Stylization, there is uniformity through Conventionalization. This feature
is functional as it enables experienced viewers to apprehend Representa-
tional meaning at a glance and lowers the likelihood of misleading inter-
pretation. The wider context is realized through the surrounding verbal text
while overt markers on the visual display indicate critical circumstantial
features such as, for example, congruent lines and angles in the display.

Authenticated
Following OToole (1994), if more than one Figure is involved in the visual
display, then the Metaphorical Narrative may involve, for example, a 'plot'
such as the intersection of a line with a plane or intersections between
relations.
At this stage, it is useful to differentiate between abstract graphs, sta-
tistical graphs, and diagrams. Generally, abstract graphs refer to graphs
showing a functional relationship between two or more participants by
means of a curve or surface. The points are plotted on a set of co-ordinate
axes and include only those points which satisfy the given relation. Statis-
tical graphs show the relationship between certain sets of quantities or
numbers by means of a series of bars, lines, or points plotted with respect to
a set of co-ordinate axes. The term diagram is used here in the broadest
sense to include pictorial or graphical representation of entities and their
relations such as Venn diagrams, geometrical figures, and other graphs
such as those used in graph theory and topology. However, it is not possible
to extend this discussion to include the analysis of statistical graphs and
diagrams. In what follows I use the systemic framework to consider the
graph of/(jc) = 4x - 4x2 displayed in Figure 1.
f(x) = 4x - 4x2
Figure 1. Abstract graph off(x) = 4x - 4x*

Authenticated
20 K. O'Halloran
Analysis of an abstract graph
We may note that mathematical graphs and diagrams are themselves

multisemiotic texts. Language and mathematical symbolism are used, for
example, for Titles and Labels in the visual display. The functions of each of
these components are discussed below.
The modal function
The viewer's attention is drawn immediately to the curve through multiple

strategies of Framing, opposition between the Lines of the Axes and the
Curvature of the Figure, and Labeling. (If the graph were embedded in
verbal text, the visual display would be given a Caption to which the viewer
is directed through Size, Underlining, Font, and positioning of the text.)
The viewer's attention is thus initially drawn by the set of Axes which is an
extreme case of explicit Framing. The Axes function modally to direct the
viewer's gaze to the central part of the display, the origin of the Axes, where
the Figure, the parabola illustrating the pattern of covariation, is located. If
the Figure is located askew from this central location, the viewer's attention
is immediately directed to the Figure, usually through the opposition
between the straight Lines of the Axes and the Curvature of the Figure. The
relation is usually Labeled according to the mathematical symbolic
description of the curve as, for example,/(x) = 4x 4x2. The Axes are also
Labeled according to the appropriate variable and function names with
each Axis being calibrated with a numerical Scale. The Arrows on the end of
the Axis give the sense of continuity and limitless expansiveness.
Following the positioning of the Label, the viewer's gaze may rest at the
peak of the curve, which is marked by its departure from the remaining
shape of the curve. On other occasions, the viewer's attention may be
directed to salient parts of the curve by additional Labeling of points as
either ordered pairs of the co-ordinates or as an upper-case letter. In some
cases, especially with graphs with two or more Figures, both strategies are
used. Although these devices are not used in Figure 1, attention is drawn
from the peak of the curve to the intersection of the curve with the axis by
two dots on zero and one. In relation to the previous discussion of the
algebraic problem involving/(x) =4x 4x2, these are the zeroes of the
function. Other devices may also be used to engage the viewer, such as
Shading to indicate areas and Line Solidarity to indicate dominance of one
line over another.
The shape of the curve, or the Curvature also acts modally in engaging
the viewer. In Figure 1, the viewer's gaze is directed downwards through the

Authenticated
shape and the Arrows of the curve. In combination with the perfect sym-
metry of the curve, the viewer intuitively senses the unfailing continuity of
the established pattern of covariance. These graphs contribute to the
ideology of perfection associated with mathematical descriptions although,
needless to say, all relations do not exhibit such properties. In these cases,
other selections function to realize maximum modality.
Representational function
Once the sharp short Modal impact subsides, Representational Meaning

dominates and the viewer is engaged with the experiential meaning of the
graph or diagram. The conventionality of display often allows experienced
viewers to digest the experiential meaning of graphs and diagrams at a
glance.
The pattern of variation of f(x) 4x 4x2 displayed in Figure 1 is
dependent on the domain. That is, the picture of the pattern of covariation
changes according to the selection Scales of the axes. For example, the
pattern of covariation realized by Figure 1 would be radically different if the
values were only plotted for the negative values of x. Although this is not a
problem for experienced viewers, it is potentially misleading to the novice.
This is a major difference compared with symbolic representations which
are complete descriptions of the relations. Visual descriptions are therefore
necessarily incomplete as they give partial descriptions. This problem is
discussed further in relation to Compositional Meaning.
The Representational meaning of abstract graphs such as Figure 1 relies
on implicit knowledge of the grammar of mathematical visual display,
without which it would be impossible to unpack the experiential meaning of
a graph such as/(jc) = 4x 4x2. This feature of the grammar of abstract
graphs deserves special mention.
In Figure 1, the graph represents the relation or the covariation between
the two sets obtained by the mapping/: 4x 4x2. Formally, it is the
set of ordered pairs (x,f(x)). The curve encodes two dimensions of
experiential meaning; first, the corresponding values of and f(x) and
second, a view of the rate of change off(x) with respect to x. In the first case,
each value of x displayed in the graph corresponds to a value off(x) which is
limited to those values displayed on the graph. In the second case, the Slope
of the curve at a point, or rather the slope of the tangent at that point, realizes
the rate of change of f(x) with respect to x. The positive rate of change
gradually decreases as x approaches the value 0.5. When x has the value 0.5,
the rate of change is zero. After this point, the rate of change decreases at
an increasing rate. The critical value is therefore at x = 0.5 where the

Authenticated
22 K. O'Halloran
derivative, or rate of change, is zero. Interpretation of the experiential

meaning of abstract graphs relies on knowledge of both the grammar of
mathematical visual display and symbolic notation. Further to this, if the
problem had involved finding graphically the zeroes of the function,
f(x) = 4x 4x2, the viewer must be able to interpret this as meaning the
intersection of the curve and the x-axis. It appears that, although visual
display in mathematics is more intuitive than the symbolic descriptions,
interconnections must be made within and between the two codes if the
Representational meaning of the visual display is to be appreciated.
The compositional function
As with language, the three metafunctions work together and are only
separated for the convenience of analysis. In his analysis of visual art,
O'Toole (1994: 22) explains:
decisions about the arrangement of forms within the pictorial space, about line and
rhythm and colour relationships, have been made by the artist in order to convey
more effectively and more memorably the represented subject and to make for a
more dynamic modal relation with the viewer.
Thus, the composition of mathematics visual display only makes sense

with respect to 'what is represented' and 'how it is conveyed to the viewer'
(O'Toole 1994: 23). In this regard, mathematical graphs and diagrams are
concerned with directly engaging the viewer with unambiguous Repre-
sentational meaning. In particular, there is a direct link between the
Composition function and the Representational function in relation to the
Figure in mathematical abstract graphs since the Positioning of the Figures
is controlled by the Scale selection on the axes.
In mathematical visual representations, horizontal and vertical align-
ment is explicitly realized through the set of Axes. In contrast, diagonal
alignment is associated with the curves realizing the patterns of covariation
in the case of abstract diagrams and perspective in geometrical repre-
sentations. Following O'Toole (1994), the horizontal and vertical Axes
'contribute to stability and harmony, while their relation to the diagonal
axes tends to create energy and dynamism' (1994: 23). Nowhere could this
be more true than in mathematical visual representations where perhaps
this notion originated. For example, in abstract and statistical graphs, the
participants are represented as stable elements (the and/(x) coordinates)
while tension between them is created through the curve as it visually traces
their dynamic relationship.

Authenticated
In connection with the Slope of the diagonalism itself in abstract graphs,

the representation has been conventionalized to correspond to the semi-
otics four physical/psychological perception. In the case of perfect covaria-
tion with changes in resulting in an exact corresponding change in/(x),
the relation is depicted as a straight line at forty-five degrees to the axes. The
steeper the line, the greater the rate of change or gradient. On the other
hand, the flatter the diagonal line, the less the tension, since a large variation
in creates only a small change in/(x) until the rate of change is zero. In this
case, the tension is resolved and stability attained.
Joint construction: Shifts between codes and semiotic metaphor
In addition to the choice of options in the systems constituting each

semiotic, meaning is realized through joint constructions across codes in
mathematical discourse. As an aid to this joint construction, mathematical
texts contain explicit discursive links across codes. For example, the verbal
text may contain lexical items and statements which are symbolically
expressed. These symbolic configurations may also appear in the visual
displays. However, as mathematical symbolic statements form the basis for
the description of relations, they are usually spatially separated from the
main body of the verbal text. Prominence of the symbolic notation is also
achieved through the use of italics. In mathematical texts, as for scientific
texts (Lemke 1998), the reading path is non-sequential. That is, mathe-
matical symbolic statements, worked examples, and visual displays usually
are the initial and primary focus. If clarification is required, then the main
body of the verbal text may be consulted.
One facet of the joint construction of meaning in a mathematical text is
the critical phase of shifts between codes. This results in a phenomenon
which I have called semiotic metaphor (O'Halloran 1996, in press). This
process is similar to grammatical metaphor in that a shift in the functions
of elements occurs and new entities are introduced, but in this case,
these shifts take place as a result of movements between semiotic codes.
Semiotic metaphor is thus concerned with shifts in functions and the
introduction of new elements occurring as the result of choices across dif-
fering lexicogrammatical systems. For example, semiotic metaphor occurs
in the verbal translation 'the square root of b squared minus four ac' for the
mathematical symbolic expression Vb2 - 4ac since the Operative process
realized by y in the symbolism shifts to a Thing or entity in the verbal
translation as evidenced by the nominal group in the linguistic version. An
alternative congruent linguistic version would be 'b squared minus four ac
raised to the power of one half. Thus it may be argued that this is a case of

Authenticated
24 K. O'Halloran
grammatical metaphor in language, but as will be seen later on, this

interpretation is not always possible.
The analysis of the discourse in a secondary-school mathematics trigo-
nometry lesson (O'Halloran 1996) provides further examples of semiotic
metaphor. In this lesson, the diagram in Figure 2 is the vehicle through
which the following verbal problem is translated into mathematical
symbolism for its ultimate solution.
Teacher: a man is actually at this point here

he is climbing a cliff
and /ahh doesn't know
how high up he is
and he looks down of course
and looks at the river
and doesn't know
how wide the river is
so with this information, he has a ten metre rope and a device that
measures angles,
we are asking the question
how can the man determine, firstly, the height of the cliff at point A
and, secondly,
the width of the river
Cliff
face
river R
Figure 2. Diagrammatic representation of trigonometric problem, HOW HIGH and HOW

WIDE?

Authenticated
There are several explanations for the pivotal role of the diagram in this
example. First, diagrammatic representation corresponds to our physical
and physiological perceptions and so enlists knowledge of the material
world which is not construed through language alone. Second, the diagram
expands experiential meaning through semiotic metaphor and the intro-
duction of new participants which previously did not exist in either the
perceptual or linguistic constructions. This allows the mathematical sym-
bolic relations realized through the diagrammatic representation to be
formulated. These facets of the visual display are respectively examined
below.
The accompanying labels and the direct correspondence between the
diagrammatic visual display and physical perceptual reality mean that the
experiential meaning of the diagram corresponds to our intuitive under-
standing of the world which is more accessible than the symbolic text and
certain segments of the oral discourse. That is, the vertical line segment
together with its shading and the horizontal line correspond to perceptions
of a cliff and a river respectively. These participants are clearly labelled as
'Cliff face' and 'river' and the height as 'h' and the width as 'r' with arrows
explicitly marking the respective distances.
The diagrammatic representation of the verbal problem is the first
instance of semiotic metaphor. In the oral discourse, experiential meaning
is congruently realized. However, in the semiotic shift to visual display, the
verbal Material process of looking down is transformed into an entity in the
diagram as it is represented by the diagonal line segment. This new entity is
later introduced in the verbal discourse as 'the line of sight'. It is dynam-
ically marked in the diagram with downward arrows indicating the process
of looking down. The Material processes of taking an angle measurement
become the participants and with arrows marking the direction through
which the angles are measured and the dashed line representing the hor-
izontal line of sight. In a similar manner, the circumstance of'how high' and
'how wide' realizing Extent (spatial distance) is transformed into partici-
pants h and later A-10 which are marked with arrows indicating the relative
distances in the diagram. Significantly, the introduction of the line of sight
together with the width of the river and the height of the cliff mean the
Figure of the right triangle appears for the first time. From this point,
congruency of the angles of depression and the angles of elevation is
established in the visual display and is realized through these same labels a
and with the arrows pointing upwards.
This introduction of new participants, the triangles and the congruent
angles, forms a crucial step in the solution of the problem as from this
vantage point the problem is translated into mathematical symbolic
representation as displayed below. Therefore, as a result of semiotic

Authenticated
26 K. O'Halloran
metaphor and the consequent introduction of new participants, the dia-

grammatic display becomes the point of departure for the expression of the
mathematical relationship which exists between the sides and angles of a
right triangle in the form of the tangent ratio. This relationship forms the
basis for the symbolic solution of the problem.
In right :
~^ (1)
In right AABR:
h
(2)
The shift to the semiotic modality of mathematical symbolism also

involves semiotic metaphor as new entities that previously did not exist in
the verbal or diagrammatic forms are created. Although the nuclear con-
figurations realizing the tangent ratios are derived from the diagram by
identifying the opposite and adjacent side with the respective angles, the
new participants, tan and tan 0, are not realized diagrammatically. Fur-
ther to this, the rankshifted symbolic participants which result from the
process of dividing the opposite side by the adjacent side similarly are not
realized in the visual display. The diagrammatic representation has limits in
terms of its meaning potential. Through the symbolism, the newly con-
structed Figure of the triangle is decomposed and the proportional rela-
tionship of the relative size of its Parts is exactly stated. Visually one gains
an idea of the relative sizes of these Parts, but not the exact ratio which is
realized through the symbolism.
These observations reveal the circumscribed functionality of each of the
semiotic codes involved in the solution of this problem. First, a 'real life'
problem is expressed linguistically. Second, this problem is translated into a
visual representation whereby shifts that I have described as semiotic
metaphor result in the creation of entities from verbal processes and
Circumstantial elements. These entities combine to create new participants
which have no counterpart in the verbal discourse. In particular, the
Figures of the triangles emerge as configurations of the participants h, r and
10 and a and . This means that these participants may be now viewed as a
connected whole rather than as isolated entities. Mathematical symbolism
allows the description of the proportional relationship that exists between
these parts of the connected whole. From this relationship, the pattern of

Authenticated
variation that exists between the angles and the sides of a right triangle, the
solution to the problem is symbolically derived. That is, verbal discourse
functions to describe commonsense reality, visual display connects our
physiological perceptions to this reality and in combination with meta-
phorical shifts, creates new entities which are intuitively accessible.
Mathematical symbolism allows exact descriptions of the relations existing
between the parts of the new entities. These descriptions create new entities
which may be configured through Operative processes.
Given that the verbal discourse acts as metadiscourse for each stage in the
derivation of the solution, linguistic choices must function to describe not
only commonsense reality, but also scientific and mathematical reality.
That is, language has evolved to account for the extensions of meaning
realized by visual display and mathematical symbolic description. Apart
from the creation of new register-specific items, new participants are thus
introduced in language through the process of grammatical metaphor.
That is, grammatical metaphor has evolved in part as a response to the
extensions of meaning which have occurred in other semiotic codes
including those resulting from the process of semiotic metaphor. In reac-
tion to this expanded meaning potential, new participants like 'the angle
of elevation' and 'angles of depression' have thus come to function as
paradigmatic choices in languages.
In conclusion, I believe that identification of the phenomenon of semiotic
metaphor is significant for several reasons. First, this notion may provide
the means through which the contributions and limitations of different
semiotic codes in multisemiotic texts may be more fully appreciated
because these shifts in meaning allow for semantic expansions that would
not have otherwise been possible. Second, the evolution of grammatical
metaphor in language is perhaps a response to the semantic expansions
which occur through semiotic metaphor and the use of mathematical
symbolism and visual display for the construction of a scientific view of the
world. Indeed, as dynamic multisemiotic representations are becoming an
increasing part of our lives, investigation of semiotics outside the confines
of linguistic representations becomes imperative. So, too, does investiga-
tion of mathematical and scientific discourse which forms the basis for what
Foucault (1980, 1984) describes as our 'regime of truth'.
Note
For an alternative interpretation of agency and the 'semiotic subjects' involved in mathe-
matical activity, see Rotman's (1988, 1993) formulations of 'Mathematician/Subject',
'Agent', and 'Person', and subsequent discussions by Bell (1995) and Mortensen and
Roberts (1997), for example.

Authenticated
28 K. O'Bailor an
References
Azzouni, Jody (1994). Metaphysical Myths, Mathematical Practice: The Ontology and Epis-
temology of the Exact Sciences. New York: Press Syndicate of the University of Cambridge.
Bell, John L. (1995). Review of Rotman's Adlnfinitum. Philosophica Mathematica 3,218-221.
Borowski, Ephraim J. and Borwein, Jonathan M. (eds.) (1989). Collins Dictionary of Mathe-
matics. London: HarperCollins.
Cajori, Florian A. (1927). A History of Elementary Mathematics, revised edition. New York:
Macmillan.
(1952). A History of Mathematical Notations, vol. 2, third edition. Chicago: Open Court
Publishing Company.
(1974). A History of Mathematical Notations, vol. 1, second edition. Chicago: Open Court
Publishing Company.
(1991). A History of Mathematics, revised edition. New York: Chelsea.
Carter, Ronald (1987). Vocabulary: Applied Linguistic Perspectives. London: Allen & Unwin.
Coleman, Edwin (1988). The role of notation in mathematics. Unpublished Ph.D thesis,
University of Adelaide.
Devlin, Keith (1994). Mathematics: The Science of Patterns. New York: Scientific American
Library.
Foucault, Michel (1970). The Order of Things. London: Routledge.
(1972). The Archaeology of Knowledge. London: Tavistock Publications.
(1980). Power/knowledge: Selected Interviews and Other Writings 1972-1977. London:
Harvester Press.
(1984). Truth and power. In The Foucault Reader, Paul Rabinow (ed.), 51-75. London:
Penguin.
Grabiner, Judith V. (1986). Is mathematical truth time-dependent? In New Directions in the
Philosophy of Mathematics, Thomas Tymoczko (ed.), 201-213. Boston: Birkhauser.
[Reprinted from American Mathematical Monthly 81 (4) (April 1974), 354-365.]
Halliday, Michael A. K. (1985). Spoken and Written Language. Geelong, Victoria: Deakin
University Press. [Republished by Oxford University Press, 1989.]
(1994). An Introduction to Functional Grammar. London: Edward Arnold. [First
ed. 1985.]
Joseph, George G. (1991). The Crest of the Peacock: Non-European Roots of Mathematics.
London: I. B. Tauris Company.
Kline, Morris (1972). Mathematics in Western Culture. London: Pelican.
(1980). Mathematics: The Loss of Certainty. New York: Oxford University Press.
Kress, Gnther and van Leeuwen, Theo (1990). Reading Images. Geelong, Victoria: Deakin
University Press.
(1996). Reading Images: The Grammar of Visual Design. London: Routledge.
Khn, Thomas S. (1970). The Structure of Scientific Revolutions, second edition. Chicago:
University of Chicago Press.
Lemke, Jay L. (1998). Multiplying meaning: Visual and verbal semiotics in scientific text. In
Reading Science, James R. Martin and Robert Veel (eds.), 87-113. London: Routledge.
Marcus, Solomon (1997). The logical and semiotic status of the canonic formula of myth.
SemioticaU6(2/4), 115-188.
Martin, James R. (1992). English Text: System and Structure. Amsterdam: John Benjamins.
Mortensen, Chris and Roberts, Lesley (1997). Semiotics and the foundations of mathematics.
Semiotica 115 (1/2), 1-25.
Mumford, Lewis (1934). Technics and Civilisation. New York: Harcourt, Brace.

Authenticated
O'Halloran, Kay L. (1996). The discourses of secondary school mathematics. Unpublished

Ph.D. Thesis, Murdoch University, Western Australia.
(in press). Interdependence, interaction, and metaphor in multisemiotic texts. Social
Semiotics.
O'Toole, Michael (1990). A systemic-functional semiotics of art. Semiotica 82 (3/4), 185-209.
(1992). A functional semiotics for the visual arts. In Poetics of the Text: Essays to Celebrate
Twenty Years of the Neo-Formalist Circle, Joe Andrew (ed.), 57-78. Amsterdam: Rodopi.
(1994). The Language of Displayed Art. London: Leicester University Press.
(1995). A systemic-functional semiotics of art. In Discourse in Society: Systemic Functional
Perspectives: Meaning and Choice in Language: Studiesfor Michael Halliday, Peter H. Fries
and Michael Gregory (eds.), 159-179. Norwood, NJ: Ablex.
Rotman, Brian (1987). Signifying Nothing: The Semiotics of Zero. London: Macmillan Press.
(1988). Towards a semiotics of mathematics. Semiotica 72 (1/2), 1-35.
(1993). Adinfinitum. Stanford, CA: Stanford University Press.
Shin, Sun-Joo (1994). The Logical Status of Diagrams. New York: Cambridge University
Press.
Thorn, Rene (1983). Mathematical Models of Morphogenesis, trans, by W. M. Brookes and
D. Rand. Chichester: Ellis Horwood.
Tiles, Mary (1991). The Image of Reason. London: Routledge.
Van Leeuwen, Theo (1988). Music and ideology: Notes towards a sociosemiotics of mass
media music. In Sydney Associationfor Studies in Society and Culture Working Papers, Terry
Threadgold (ed.), 19-45. Sydney: Sydney Association for Studies in Society and Culture.
Wilder, Raymond L. (1981). Mathematics as a Cultural System. Oxford: Pergamon Press.
Kay O'Halloran (b. 1958) is a Lecturer at the National University of Singapore

<ellkoh@leonis.nus.edu.sg>. Her principal research interests include systemic functional
linguistics and its application to semiotic systems other than language, mathematical dis-
course, educational linguistics, and analysis of multisemiotic texts. Her major publications
include The discourses of secondary mathematics' (1996) and 'Interdependence, interaction,
and metaphor in multisemiotic texts' (in press).

Authenticated
Authenticated

Towards A Systemic Functional Analysis of Multisemiotic Mathematics Texts

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Towards A Systemic Functional Analysis of Multisemiotic Mathematics Texts

Загружено:

Авторское право:

Доступные форматы

Towards a systemic functional analysis of

multisemiotic mathematics texts

Mathematics may be viewed as a semiotic construction as, for example, in

Semiotica 124-1/2 (1999), 1-29 0037-1998/99/012-0001

Mathematics as a bridge between linguistic descriptions and

Various forms of mathematical symbolism evolved from natural language

Brought to you by | Stockholms Universitet

followed in which quantitative mathematical descriptions of the material

Brought to you by | Stockholms Universitet

Any account of the contemporary semantics and grammar of mathematical

Expansion of experiential meaning

Invoking Halliday's (1994) system of TRANSITIVITY, semantic exten-

Brought to you by | Stockholms Universitet

decreasing and sharing physical objects were replaced with arithmetical

This statement realizes a Relational Identifying process through the

Brought to you by | Stockholms Universitet

at all. For example, if one considers 4ac as 'four multiplied by a multiplied

Contraction of experiential and interpersonal meaning

The contraction of meaning potential of language in mathematical sym-

Brought to you by | Stockholms Universitet

Restricted logical meaning

Mathematical discourse is concerned with only selected logic which post-

Brought to you by | Stockholms Universitet

is not entirely deductive in nature. By definition, deductive logic is con-

Consider the function/(jc) = 4x 4.x2

If I represent this example using mathematical notation, the resulting

Problem : Algebraically find the zeroes off(x) =4x- 4x2

Brought to you by | Stockholms Universitet

either as occurring 'naturally' or as a consequence of some predetermined

Specification of textual meaning

The textual organization of mathematical texts is highly conventionalized

The graphology of mathematical symbolism

The contraction, condensation, and extension of ideational, interpersonal,

Brought to you by | Stockholms Universitet

Spatial positioning in combination with serial positioning plays an

A framework for a systemic functional grammar

Table 1 gives a descriptive framework for a systemic functional grammar of

Brought to you by | Stockholms Universitet

Sequential placei ition (operati

Download Date | 12/4/16 1:25 PM

Brought to you by | Stockholms Universitet

Restricted range of units i Convent

Download Date | 12/4/16 1:25 PM

Brought to you by | Stockholms Universitet

contributes to the view of mathematics as universal truth when in fact

The emphasis of visual forms is directed towards the representation of

A framework for a systemic functional grammar of

Using O'Toole's (1994) systemic functional model in The Language of

Brought to you by | Stockholms Universitet

(for example, Venn diagr

Brought to you by | Stockholms Universitet

historically to reveal metafunctional tendencies which have resulted in the

visual modality rests on culturally and historically determined standards of what is

Hence, although mathematical representations clearly do not corre-

Brought to you by | Stockholms Universitet

Figure 1. Abstract graph off(x) = 4x - 4x*

Brought to you by | Stockholms Universitet

Analysis of an abstract graph

We may note that mathematical graphs and diagrams are themselves

The modal function

The viewer's attention is drawn immediately to the curve through multiple

Brought to you by | Stockholms Universitet

Once the sharp short Modal impact subsides, Representational Meaning