Studying mathematics for teaching inside teacher education

Full text

Education as Change, submission December 2009
1
Mathematics for Teaching Matters
1
Jill Adler
In this paper, I illuminate the notion of mathematics for teaching (its matter) and argue that
it matters (it is important), particularly for mathematics teacher education. Two examples
from studies of mathematics classrooms in South Africa are described, and used to illustrate
what mathematics teachers use, or need to use, and how they use it in their practice: in
other words, the substance of their mathematical work. Similarities and differences across
these examples, in turn, illuminate the notion of mathematics for teaching, enabling a return
to, and critical reflection on, mathematics teacher education.
Introduction
This paper explores the notion of mathematics for teaching, and why it matters for the
teaching and learning of mathematics in general, and mathematics teacher education in
particular. This exploration builds on the seminal work of Lee Shulman. In the mid-
1980s Shulman argued cogently for a shift in understanding, in research in particular, of
the professional knowledge base of teaching. He highlighted the importance of content
knowledge in and for teaching, criticising research that examined teaching activity
without any concern for the content of that teaching. He described the various
components of the knowledge base for teaching, arguing that content knowledge for
teaching included subject matter knowledge (SMK), pedagogical content knowledge
(PCK) and curriculum knowledge (Shulman, 1986; 1987). Shulman’s work set off a
research agenda, with a great deal focused on mathematics. This paper draws from and
builds on the mathematical elaboration of Shulman’s work.
The profound insight of Shulman’s work was that being able to reason
mathematically, for example, was necessary but not sufficient for being able to teach
others to reason mathematically. Being able to teach mathematical reasoning involves
recognising mathematical reasoning in others’ discourse, and at various curriculum
levels, being able to design and adapt tasks towards purposes that support mathematical
reasoning, and critically working with or mediating the development of such in others.
We could say the same for being able to solve algebraic or numeric problems. Most
mathematics teachers and mathematics teacher educators would agree with this
assertion. Yet, in the particular case of mathematical reasoning, its actuality in
curricular texts, classroom practices and learner performances remains a challenge in
many, if not most, classrooms (Stacey & Vincent, 2009). We could say the same for
learner performance in many areas of mathematics, as well as algebra. Despite the
longevity and consistency of elementary algebra in school mathematics curricula
worldwide, large numbers of learners experience difficulty with this powerful symbolic
system (Hodgen, Kuchemar, Brown & Coe, 2008).
In this paper I argue that strengthening our understanding of the mathematical work
of teaching, what some refer to as mathematics for teaching, is a critical dimension of
enhancing its teaching and learning. Mathematics for teaching matters, for all our
1
An earlier version of this paper was presented at the Australian Association of Mathematics Teachers Conference in
Perth, July, 2009. I am grateful to the AAMT for permission to publish this revised paper in Education as Change.

ADLER
Education as Change submission Dec 2009
2
learners, as do its implications for mathematics teacher education. I will develop this
argument through examples from school mathematics classrooms that invoke
mathematical reasoning, together with comment on developments in mathematics
teacher education in South Africa. Ultimately, the argument in this paper poses
considerable challenges for mathematics teacher education.
Teaching and learning mathematics in South Africa
Post-apartheid South Africa has witnessed rapid and intense policy and curriculum
change. New mathematics curricula are being implemented in schools across Grades 1–
12, where there is greater emphasis than before on sense-making, problem-solving and
mathematical processes, including mathematical reasoning, as well as on new topics
related to data handling and financial mathematics. New education policy and curricula
have strong equity goals, a function of the deep and racialised inequality produced
under apartheid that affected teachers and learners alike. New policies and
qualifications have been introduced into teacher education, with goals for improving the
quality of teachers and teaching. In the case of mathematics, there is also a quantitative
goal – of need to address enduring critical shortages of qualified secondary mathematics
teachers. Tertiary institutions have responded, offering new degree and diploma
programs for upgrading teachers in service, retraining into teaching, and preparing new
teachers.
It is in moments of change that taken-for-granted practices are unsettled, in both
inspiring and disconcerting ways. Moments of change thus provide education
researchers and practitioners with challenging opportunities for learning and reflection.
Of pertinence to this paper is that the challenge of new curricula in schools and thus
new demands for learning and teaching, on top of redress, bring issues like the selection
of knowledges for teacher education development and support to the fore. Mathematics
teacher educators in all tertiary institutions have had the opportunity and challenge to
make decisions on what knowledge(s) to include and exclude in their programs, and
how these are to be taught/learnt. This has meant deliberate attention to what
mathematics, mathematics education and teaching knowledge teachers need to know
and be able to use to teach well. This is no simple task: in South Africa, teaching well
encompasses the dual goals of equity and excellence. At the same time as strengthening
the pool of mathematics school leavers entering the mathematical sciences and related
professions, high quality teaching also entails catering for diverse learner populations,
and inspiring school learners in a subject that all too often has been alienating.
Hence the question: what selections from mathematics, mathematics education and
teaching
2
are needed to provide the greatest benefit to prospective and in-service
teachers?
Shulman’s categories provide a starting point to answering this question. Others,
particularly Ball and her colleagues working on mathematical knowledge for teaching in
Michigan USA, have argued that these categories need elaboration; and that elaboration
requires a deeper understanding of mathematics teaching, and hence, of teachers’
mathematical work. Ball, Thames and Phelps (2008) have elaborated Shulman’s
2
Mathematics education here refers to the field of research and other texts related to mathematics curricula; teaching
refers to the professional practice.

ADLER
Education as Change submission Dec 2009
3
categories, distinguishing within subject matter knowledge, between Common and
Specialised Content Knowledge where the latter is what teachers in particular need to
know and be able to use. Within Pedagogical Content Knowledge, they distinguish
knowledge of mathematics and students, and knowledge of mathematics and teaching.
These latter are knowledge of mathematics embedded in (and so integrated with) tasks
of teaching, that is, a set of practices teachers routinely engage in or need to engage in.
In their more recent work where they examine case studies of teaching, Hill, Blunk,
Charalambos, Lewis, Phelps, Sleep and Ball (2008) note that while their elaboration is
robust, compelling and helpful, they underestimated the significance of what Shulman
identified as Curriculum Knowledge. What this reflects is that all teaching always
occurs in a context and set of practices, of which curricular discourses are critical
elements. Ball et al.’s elaboration of Shulman’s categories is useful, particularly as it
has been derived from studies of mathematics classroom practice. They provide a
framework with which to think about and make selections for teacher education. At
immediate face value, they suggest that mathematical content in teacher education and
for teaching requires considerable extension beyond knowing mathematics for oneself.
I go further to say we need to understand what and how such selections take shape in
mathematics teacher education practice. As in school, teacher education occurs in a
context and set of practices, and is shaped by these. In addition, as intimated above, in
mathematics teacher education, mathematics as an “object” or “focus” of learning and
becoming, is integrated with learning to teach. The research we have been doing in the
QUANTUM
3
project in South Africa (that now has a small arm in the UK) has done
most of its work in teacher education as an empirical site, complemented by studies of
school mathematics classroom practice. The goal is to understand the substance of
opportunities to learn mathematics for teaching in teacher education, and how this
relates to the mathematical work teachers do in their school classrooms.
In this paper, I select two examples from studies of mathematics classrooms in South
Africa. I use these to illustrate what mathematics teachers use, or need to use, and how
they use it in their practice: in other words, the substance of their mathematical work.
Similarities and differences across these examples, in turn, illuminate the notion of
mathematics for teaching, enabling a return to, and critical reflection on, mathematics
teacher education.
Designing and mediating productive mathematics tasks
Example 1: Angle properties of a triangle
The episode discussed below is described in detail in Adler (2001)
4
, and takes place in a
Grade 8 classroom. This teacher was particularly motivated by a participatory
pedagogy, and developing her learners’ broad mathematical proficiency (Kilpatrick,
Swaffold & Flindell, 2001). She paid deliberate attention to supporting her learners’
participation in mathematical discourse (Sfard, 2008), which in practice involved
3
For details on QUANTUM, a project focused on Qualifications for Teachers Underqualified in Mathematics, see
Adler & Davis (2006), Davis, Parker & Adler (2007); Adler & Huillet (2008), Adler (2009)
4
The focus of the study reported in Adler (2001) was on teaching and learning mathematics in multilingual
classrooms. There I discuss in detail the learners’ languages, and how and why talking to learn worked in this class.
I have since revisited this data, reflecting on the teachers’ mathematical work (see Adler, 2006).

ADLER
Education as Change submission Dec 2009
4
having them learn to reason mathematically, and verbalise this. It is interesting to note
that the empirical data here date back to the early 1990s and long before curriculum
reform as it appears today in South Africa was underway.
As part of a sequence of tasks related to properties of triangles, the teacher gave the
activity in Figure 1 to her Grade 8 class. The questions I will address in relation to this
task are: What mathematical work is entailed in designing this kind of task, and then
mediating it in a class of diverse learners?
If any of these is impossible, explain why; otherwise, draw it.
 Draw a triangle with 3 acute angles.
 Draw a triangle with 1 obtuse angle.
 Draw a triangle with 2 obtuse angles.
 Draw a triangle with 1 reflex angle.
 Draw a triangle with 1 right angle.
Figure 1. A triangle task.
The task itself evidences different elements of important mathematical work entailed
in teaching learners to reason mathematically
. Firstly,
this is not a “typical” task on the
properties of triangles. A more usual task to be found in text books, particularly at the
time of the research, would be to have learners recognise (identify, categorise, name)
different types of triangles, defined by various sized angles in the triangle. What the
teacher has done here is recast a “recognition” task based on angle properties of
triangles into a “reasoning” task (reasoning about properties and so relationships). She
has constructed the task so that learners are required to reason in order to proceed. In so
doing, she sets up conditions for producing and supporting mathematical reasoning in
the lesson and related proficiencies in her learners. Secondly, in constructing the task so
that learners need to respond whether or not particular angle combinations are
“impossible” in forming a triangle, the task demands proof-like justification—an
argument or explanation that, for impossibility, will hold in all cases. In this task,
content (properties of triangles) and processes (reasoning, justification, proof) are
integrated. The question, of course, is what and how learners attend to these
components of the task, and how the teacher then mediates their thinking.
In preparation for this lesson and task, the teacher would have had to think about the
mathematical resources available to this classroom community with which they could
construct a general answer (one that holds in all cases). For example, if as was the case,
learners had worked with angle sum in a triangle, what else might come into play as
learners go about this task? What is it about the triangle as a mathematical object that
the teacher needs to have considered and that she needs to be alert to as her learners
engage in reasoning about its properties?
Before engaging further with the details of the teachers’ mathematical work, let us
move to the actual classroom, where students worked on their responses in pairs. The
teacher moved between groups, probing with questions like: “Explain to me what you
have drawn/written here?”, “Are you sure?”, “Will this always be the case?” She thus
pushed learners to verbalise their thinking, as well as justify their solutions or proofs. I

ADLER
Education as Change submission Dec 2009
5
foreground here learners’ responses to the second item: Draw a triangle with two obtuse
angles. Interestingly, three different responses were evident.
• Some said, “It is impossible to draw a triangle with two obtuse angles, because
you will get a quadrilateral.” They drew the shape shown in Figure 2.
Figure 2. Student drawing of a triangle with two obtuse angles
• Others reasoned as follows: “An obtuse angle is more than 90 degrees and so two
obtuse angles give you more than 180 degrees, and so you won’t have a triangle
because the angles must add up to 180 degrees.”
• One learner (Joe) and his partner reasoned in this way: “If you start with an angle
say of 89 degrees, and you stretch it [to make it larger than 90 degrees], the other
angles will shrink and so you won’t be able to get another obtuse angle.” Their
drawing is shown in Figure 3.
Figure 3. Joe and his partner’s response.
The range of learner responses to this task is indicative of a further task-based
teaching skill. The task is designed with sufficient openness, and so diverse learner
responses are possible and indeed elicited. In addition, the third, unexpected, response
produced much interest in the class, for the teacher, and for myself as researcher. The
first two responses were common across learners and more easily predicted by the
teacher.
Having elicited these responses, it is the teacher’s task to mediate within and across
these responses, and enable her learners to reason whether each of these responses is a
general one, one that holds in all cases
5
. In the many contexts where I have presented
the study and this particular episode, much discussion is generated both in relation to
the mathematical status of the responses, and their levels of generality, as well as
simultaneous arguments as to what can be expected of learners at a grade 8 level. What
constitutes a generalised answer at this level? Are all three responses equally general? Is
5
The interesting interactions that followed in the class are described and problematised in Adler (2001) and will not
be focused on here.

ADLER
Education as Change submission Dec 2009
6
Joe’s response a generalised one? How does the teacher value these three different
responses, supporting and encouraging learners in their thinking, and at the same time
judging/evaluating their mathematical worth?
These are mathematical questions, and the kind of work this teacher did on the spot
as she worked to evaluate and value what the learners produced was also mathematical
work. The point here is that this kind of mathematical work i.e. working to provoke,
recognise and then mediate notions of proof and different kinds of justification, is
critical to effective teaching of “big ideas” (like proof) in mathematics. In Ball et al.’s
terms, this work entails specialised content knowledge (judging the mathematical
generality of the responses), knowledge of mathematics and teaching (designing
productive tasks) and mathematics and students (and mediating between these and
learners’ mathematics).
We need to ask further questions about subject matter knowledge, or content in this
example, and specifically questions about the angle properties of triangles. The insertion
of a triangle with a reflex angle brought this to the fore in very interesting ways. Some
learners drew the following, as justification for why a triangle with a reflex angle was
possible; and so provoked a discussion of concavity, and interior and exterior angles.
Figure 4. Learner drawings to justify triangles with reflex angles.
The tasks of teaching illuminated in this example are: task design where content (angle
properties of triangles) and process (reasoning, justifying) are integrated; mediation of
both mathematical content and processes; and valuing and evaluating diverse learner
productions. The mathematical entailments of this work are extensive, and are
illustrative of both subject matter knowledge and pedagogical content knowledge. The
teacher here reflects a deep understanding of mathematical proof, and in relation to a
specific mathematical object and its properties. To effectively mediate Joe’s response
and the two above, she would also need to ask suitable questions or suggest productive
ways forward for these learners, so that their notions of proof and of the mathematical
triangle are strengthened. Indeed, as learners in the class engaged with the second
triangle drawn above, their focus was that the answer was incorrect because there were
three reflex angles not one, and the teacher had a difficult time shifting them from this
focus and onto the interior angles.
In Adler (2001), I show that as the teacher mediated the three different responses to
the triangle with two obtuse angles, she worked explicitly to value each contribution

ADLER
Education as Change submission Dec 2009
7
and probe learner thinking. However, her judgment of their relative mathematical worth
was implicit. She accepted the first two responses above, but probed Joe’s, with
questions to Joe that implied she was not convinced of the generality of his argument. I
argued there that if teacher judgment of the varying mathematical worth of learner
responses offered is implicit, it is possible that only those learners who can themselves
make such judgements, or who are able to read the implicit messages in the teacher’s
actions, will appreciate and so have access to what counts mathematically. Sociological
theory and empirical research inform us that these kinds of practices favour students
with school cultural capital, and so can reproduce inequality. In Bernstein’s (1996)
terms, implicit practices will connect with learners who already understand the criteria
for what are most legitimate responses; and alienate or pass by those who are not “in”
the criteria. Typically these will be already disadvantaged learners (Parker, 2009).
The example here is compelling in a number of ways, and provokes the question:
Where, when and how does a mathematics teacher learn to do this kind of work, and in
ways that are of benefit to all learners? Before attempting to answer this and so shift
back into teacher education, we need to look at additional and different examples of the
mathematical work of teaching.
Example 2: Polygons and diagonals — or a version of the “mystic rose”
The second example is taken from a Grade 10 class (see Naidoo, 2008), where the
teacher posed the following task for learners to work on in groups: How many diagonals
are there in a 700-sided polygon?
Here too, the teacher has designed or adapted a task and presented learners with an
extended problem. They have to find the number of diagonals in a 700-sided polygon, a
sufficiently large number to require generalising activity, and so mathematical
reasoning. I pose the same questions here as for Example 1: What mathematical work is
entailed in designing this kind of task, and mediating it in a class of diverse learners?
Many teachers will recognise the “mystic rose” investigation in this problem. The
mathematical object here is a polygon and its properties related to diagonals. Yet the
problem has been adapted from a well known (perhaps not to the teacher) mathematical
investigation of points on a circle and connecting lines — a different, though related
object. Here learners are not asked to investigate the relationship between the number of
points on a circle and connecting lines, but instead to find an actual numerical solution
to a particular polygon, albeit with a large number of sides and so approaching a circle.
I have discussed this case in detail in Adler (2009), where I point out that unlike
triangles and their properties, the general polygon and its properties is not an explicit
element of the secondary school curriculum. However, the processes and mathematical
reasoning required for learners to solve the problem are desired mathematical processes
in the new curriculum.
My concern in this paper is not with the merits of the problem and its adaptation in
an isolated way. Rather, I wish to reflect on the mathematical work of the teacher in
presenting the problem, mediating learner progress, valuing and evaluating their
responses, and managing the integration of mathematical content and mathematical
processes as foci in the lesson. I present selections from the transcript of the dialogue in
the classroom to illuminate these four components of the teachers’ mathematical work.

ADLER
Education as Change submission Dec 2009
8
The teacher (Tr), standing in the front of the class, explained what the class had to
do.
Tr: I want you to take out a single page quickly. Single page and for the next five
minutes no discussion. I want you to think about how would you possibly solve
this problem? (pointing to the projected problem: How many diagonals are there in
a 700-sided polygon?
After seven minutes, the Teacher calls the class’ attention. (Learners are referred to as Lr A, B,
etc.)
Tr: Ok! Guys, time’s up. Five minutes is over. Who of you thinks they solved the
problem? One, two, three, four, five, six.
Lr A: I just divided 700 by 2.
Tr: You just divided 700 by 2. (Coughs).
Lr A: Sir, one of the side’s have, like a corner. Yes… [inaudible], because of the
diagonals. Therefore two of the sides makes like a corner. So I just divided by
two… [Inaudible].
Tr: So you just divide the 700 by 2. And what do you base that on? …
[ ]
Tr: Let’s hear somebody else’s opinion.
Lr B: Sir what I’ve done sir is … First 700 is too many sides to draw. So if there is four
sides how will I do that sir? Then I figure that the four sides must be divided by
two. Four divided by two equals two diagonals. So take 700, divide by two will
give you the answer. So that’s the answer I got.
Tr: So you say that, there’s too many sides to draw. If I can just hear you clearly; …
that 700 sides are too many sides, too big a polygon to draw. Let me get it clear. So
you took a smaller polygon of four sides and drew the diagonals in there. So how
many diagonals you get?
Lr B: In a four-sided shape sir, I got two.
Tr: Two. So you deduced from that one example that you should divide the 700 by two
as well? So you only went as far as a 4 sided shape? You didn’t test anything else.
Lr B: Yes, I don’t want to confuse myself.
Tr: So you don’t want to confuse yourself. So you’re happy with that solution, having
tested only one polygon?
Lr B: [Inaudible response.]
Tr: Ok! You say that you have another solution. [Points to learner D] Let’s hear.
[ ]
Lr A: I just think it’s right… It makes sense.
Tr: What about you Lr D? You said you agree.
Lr D: He makes sense… He proved it… He used a square.
Tr: He used a square? Are you convinced by using a square that he is right?
Lr E: But sir, here on my page I also did the same thing. I made a 6-sided shape and saw
the same thing. Because a six thing has six corners and has three diagonals.
Lr A: So what about a 5-sided shape, then sir?
Tr: What about a 5-sided shape? You think it would have 5 corners? How many
diagonals?
I have underlined the various contributions by learners, and italicised the teachers’
mediating comments and questions. These highlight the learners’ reasoning and the
teacher’s probing for further mathematical justification.
At this point in the lesson, the teacher realises that some of the learners are confusing
terms related to polygons, as well as some of the properties of a general polygon and so
deflects from the problem for a while to examine with learners, various definitions (of a
polygon, pentagon, a diagonal, etc.). In other words, at this point, the mathematical

ADLER
Education as Change submission Dec 2009
9
object in which the problem is embedded comes into focus. It is interesting to note here
that at no point was there reflection on the polygons in use in developing responses to
the problem. All were regular and concave. A little later in the lesson, another learner
offers a third solution strategy. The three different solution representations are
summarised in Figure 5, illustrating the varying orientations students adopted as they
attempted to work towards the solution for a 700-sided polygon.
Fi
gure 5. Three different representations and reasoning.
As with Example 1, we see four tasks of teaching demanded of the teacher: task
design or adaptation; mediation of learners’ productions; valuing and evaluating their
different responses; and managing mathematics content and processes opened up by the
task.
The representations offered by learners give rise to interesting and challenging
mathematical work for the teacher. All responses are mathematically flawed, though the
approaches of Learners B and C show attempts at specialising and then generalising
(Mason, 2002). While this is an appropriate mathematical practice, the move from the
special case to the general case in both responses is problematic, though in different
ways. Does the teacher move into discussion about specialising and generalising in
mathematics (and if so, how)? Open-ended investigations and problem-solving as
described above open up possibilities for this kind of mathematical work in class. Such
opportunities were not taken up here. Each was negated empirically, and not elaborated
more generally. Should they have been taken up by the teacher, and if so, how?
Tasks of teaching and their mathematical entailments
In selecting and presenting two different examples from different secondary school
classrooms in South Africa, I have highlighted four inter-related tasks of teaching, each
of which entails considerable mathematical skill and understanding over and above (or
underpinning) the teaching moves that will ensue. The four tasks (two of which are
discussed in each of the bulleted sections below) further illustrate categories of

ADLER
Education as Change submission Dec 2009
10
professional knowledge developed by Shulman and elaborated by Ball et al. in
mathematics.
Designing, adapting or selecting tasks,
and managing processes and objects
In the first example, the process of mathematical reasoning was in focus, as was the
triangle and its angle properties. I will call this an object-and-process-focused task.
Angle properties of triangles are the focus of reasoning activity. Learners engage with
and consolidate knowledge of these properties through reasoning activity, and vice
versa. Here the integration of learning content and process appears to keep them both in
focus, and thus provides opportunities for learning both. Example 2 is also focused on
mathematical reasoning. It is a process-focused task, having been adapted (what I would
refer to as recontexualised) from an investigation and re-framed as a problem with a
solution. The mathematical object of the activity, the polygon, is backgrounded. At a
few points in the lessons, it comes into focus, when understanding polygons and their
properties is required for learners to make progress with the problem: some learners
make assumptions about what counts as a diagonal, perhaps a function of assuming
regularity (and so finding three diagonals in a hexagon); some generalise from one
specific case (a four-sided figure); while others over-generalise multiplicative processes
from number, to polygon properties.
The intricate relationship between mathematical objects and processes has been an
area of extensive empirical research in the field of mathematics education. It appears
from studying two examples of teaching that selecting, adapting or designing tasks to
optimise teaching and learning entails an understanding of mathematical objects and
processes and how these interact within different kinds of tasks. The teaching of
mathematical content and mathematical processes is very much in focus today. Reform
curricula in many countries promote the appreciation of various mathematical objects,
their properties and structure, conventions (how these are used and operated on in
mathematical practice), as well what counts as a mathematical argument, and the
mathematical processes that support such. In Example 1, we see opportunity for
developing reasoning skills, and understanding of proof at the same time as
consolidating knowledge about triangles. In Example 2, it is not apparent whether and
how either proof or reasoning will flourish through this example and its mediation. The
relevance of the mathematical object in use is unclear. Thus the question: Do we need a
mathematics for teaching curriculum that includes task interpretation, analysis and
design with specific attention to intended mathematical objects and processes and their
interaction?
In other words, should a mathematics for teaching curriculum include attention to the
mediation of mathematical content and processes as these unfold in and through
engagement with varying tasks? If so, is this to be part of the mathematics curriculum,
or part of the teaching curriculum? And hidden in this last question is a question of who
teaches these components of the curriculum in teacher education? What competences
and expertise would best support this teaching?
Valuing and evaluating diverse learner productions
Diverse learner productions are particularly evident in Examples 1 and 2, given their
more open or extended nature. Thus, in each example, the teacher dealt with responses

ADLER
Education as Change submission Dec 2009
11
from learners that they predicted, and then those that were unexpected. In Example 1,
the teacher needed to consider the mathematical validity of Joe’s argument for the
impossibility of a triangle with two obtuse angles, and then how to encourage him to
think about this himself, and convince others in the class. Similarly, we can ask in
Example 2: what might be the most productive question to ask Learner C and so
challenge the reasoning that, since 700 can be factored into 7 × 100, finding the
diagonals in a 7-sided figure is the route to the solution to a 700-sided figure? Such
questioning in teaching needs to be mathematically informed.
Together these examples illuminate how teachers need to exercise mathematical
judgement as they engage with what learners do (or do not do). This is particularly so if
teachers are building a pedagogical approach and classroom environment that
encourages mathematical practices where error, and partial meanings are understood as
fundamental to learning mathematics. In earlier work I referred to this as a teaching
dilemma, where managing both the valuing of learner participation and evaluation of
the mathematical worth of their responses was important (Adler, 2001); and illuminated
the equity concerns if and when evaluation of diverse responses—i.e., judgements as to
which are mathematically more robust or worthwhile—are left implicit.
So, a further question needs to be asked of the curriculum in mathematics teacher
education, and the notion of mathematics for teaching. Learner errors and
misconceptions in mathematics are probably the most developed research areas in
mathematics education. We know a great deal about persistent errors and
misconceptions that are apparent in learners’ mathematical productions across contexts.
These provide crucial insight into the diverse responses that can be anticipated from
learners. Yet, as Stacey (2004) argues, the development of this research into contents
for teacher education has been slow. We have shown elsewhere that the importance of
learner mathematical thinking in mathematics teacher education is evident in varying
programs in South Africa (see Davis, Adler & Parker, 2007; Adler, forthcoming;
Parker, 2009). Yet there are significant differences in the ways this is included in such
programs, and so with potential effects on who is offered what in their teacher
education. How should a mathematics for teaching curriculum then include such
content?
Mathematics for teaching matters
I have argued that mathematics for teaching matters for teaching and also for
opportunities to learn mathematics. I have suggested that what matters are task design
and mediation, as well as attention to mathematical content, objects and processes
within these. I have played on the word “matters” by suggesting firstly that these are the
“matter” or the content of mathematics for teaching; and at the same time that they
matter (have significance) in and for teacher education. Secondly, I have suggested that
there are equity issues at stake.
I now return to the context of teacher education in South Africa where various
innovative teacher education programs are grappling with a curriculum for mathematics
teachers that appreciates the complexity of professional knowledge for teaching and its
critical content or subject basis. I will focus here on what we have observed as objects
of attention (and so meanings) shift from classrooms to teacher education and back

ADLER
Education as Change submission Dec 2009
12
again, observations that support the argument in this paper, that we need to embrace our
deeper understanding of the complexities of teaching and so our task in teacher
education.
In more activity-based, participative or discursively rich classroom mathematics
practice, there is increased attention to mathematical processes as critical to developing
mathematical proficiency and inducting learners into a breadth of mathematical
practices. The examples in this paper illustrate how mathematical processes are always
related to or based on some mathematical object. If the latter is not well understood, in
the first instance by the teacher, in ways that enable her to notice when it goes out of
focus or is completely missed by students, then their reasoning is likely to be flawed or
mathematically empty. This phenomenon is apparent in classrooms in South Africa, and
more so in historically disadvantaged settings, thus perpetuating rather then attacking
inequality. Mathematical objects and processes and their interaction are the central
“matter” of mathematics for teaching. The shift in new curricula to mathematical
processes creates conditions for diminished attention to mathematical objects. Attention
to objects and processes need to be embraced in the context of teaching if access to
mathematics is to be possible for all learners.
Herein lies considerable challenge. In each of the two examples in this paper, a
mathematical object was embedded in a task that worked varyingly to support
mathematical reasoning processes. What the teacher in each case faced was different
learner productions as responses to the task. These become the focus of the teachers’
work, requiring integrated and professional based knowledge of mathematics, teaching
tasks and learner thinking. So what then, is or comes into focus in teacher education,
and not only into teacher education, but into school curricula? What we have observed
(and I have seen elements of this in elementary mathematics teacher education in the
UK), is that learner thinking and the diversity of their responses become the focus, with
the mathematical objects and tasks that give rise to these, out of focus. What one might
see in the case of the triangle properties is a task that requires learners to produce three
different arguments for why a triangle cannot have two obtuse angles. And there is a
subtle but impacting shift of attention: from how to mediate diverse responses, to
multiple answers or solutions being the required competence in learners; from teachers’
learning to appreciate diverse learner productions and their relative mathematical worth,
and more generally, multiple representations, and how to enable learners to move
flexibly between these, to these being the actual content of teaching. Simply, there are
curricular texts that now require learners to produce multiple solutions to a problem. I
leave this somewhat provocative assertion for discussion and further debate.
In conclusion, there is an assumption at work throughout this paper that teacher
education is crucial to quality teaching. In South Africa, all pre-service and formal in-
service teacher education has become the responsibility of universities. Tensions
between theory and practice abound. I hope in this paper to have provided examples that
illuminate the mathematical work of teaching, and through these opened up challenges
for mathematics teacher education. Mathematics for teaching, and its place in
mathematics teacher education, particularly in less resourced contexts, matters
profoundly.

ADLER
Education as Change submission Dec 2009
13
Acknowledgements
This paper forms part of the QUANTUM research project on Mathematics for
Teaching, directed by Jill Adler, at the University of the Witwatersrand. Dr Zain Davis
from the University of Cape Town is a co-investigator and central to the theoretical and
methodological work in QUANTUM. The elaboration into classroom teaching was
enabled by the work of Masters students at the University of the Witwatersrand. This
material is based upon work supported by the National Research Foundation under
Grant number FA2006031800003. Any opinion, findings and conclusions or
recommendations expressed in this material are those of the author and do not
necessarily reflect the views of the National Research Foundation.
Jill Adler holds the FRF Mathematics Education Chair at the University of the Witwatersrand
and the Chair of Mathematics Education at King’s College, London. Contact:
jill.adler@wits.ac.za
References
Adler, J. (2001). Teaching mathematics in multilingual classrooms. Dordrecht: Kluwer.
Adler, J (2006). Mathematics teacher education in post-apartheid South Africa: A focus on the
mathematical work of teaching across contexts. In M. Borba (Ed.) Trends in mathematics education,
Brazil (in Portuguese, pp. 45–64). São Paulo: Autêntí.
Adler, J. (2009). A methodology for studying mathematics for teaching. Recherches en Didactique des
Mathématiques.
Adler, J. (forthcoming) Look at me, look at yourself, look at the practice: Modelling teaching in
mathematics teacher education, and the constitution of mathematics for teaching. In K. Ruthven & T.
Rowland (Eds), Mathematical knowledge in teaching. Routledge.
Adler, J. & Davis, Z. (2006). Opening another black box: Researching mathematics for teaching in
mathematics teacher education. Journal for Research in Mathematics Education. 37(4), 270–296.
Adler, J. & Huillet, D. (2008). The social production of mathematics for teaching. In T. Wood & P.
Sullivan (Eds), International handbook of mathematics teacher education, Vol. 1: Knowledge and
beliefs in mathematics teaching and teaching development (pp. 195–222). Rotterdam, the
Netherlands: Sense Publishers.
Ball, D. L., Thames, M. H. & Phelps, G. (2008). Content knowledge for teaching: What makes it special?
Journal of Teacher Education, 59, 389–407.
Bernstein, B. (1996), Pedagogy, symbolic control and identity: Theory, research and critique. London:
Taylor & Francis.
Davis, Z., Adler, J. & Parker, D. (2007), Identification with images of the teacher and teaching in
formalized in-service mathematics teacher education and the constitution of mathematics for teaching.
Journal of Education. 42, 33–60.
Hill, H., Blunk, M., Charalambos, Y., Lewis, J., Phelps, G., Sleep, L. & Ball, D. (2008). Mathematical
knowledge for teaching and the mathematical quality of instruction: An exploratory study. Cognition
and Instruction. 26, 430–511.
Hodgen, J., Kuchemar, D., Brown, M. & Coe, R. (2008). Children’s understandings of algebra 30 years
on. In Joubert, M. (Ed.) Proceedings of the British Society for Research into Learning Mathematics
28(3) November 2008. Available at www.bsrlm.org.uk.
Kilpatrick, J., Swafford, J. & Findell, B. (Eds) (2001). Adding it up: Helping children learn mathematics.
Washington: National Academy Press.
Mason, J. (2002). Generalisation and algebra: Exploiting children’s powers. In L. Haggerty (Ed.), Aspects
of teaching secondary mathematics: Perspectives on practice (pp. 105–120). London: Routledge
Falmer.

ADLER
Education as Change submission Dec 2009
14
Naidoo, S. (2008). Mathematical knowledge for teaching geometry to Grade 10 learners. Johannesburg,
South Africa: School of Education, The University of the Witwatersrand.
Parker (2009). The specialisation of pedagogic identities in initial mathematics teacher education in post-
apartheid South Africa. Unpublished PhD Thesis. Johannesburg: University of the Witwatersrand.
Sfard, A. (2008). Thinking as communicating: Human development, the growth of discourses, and
mathematizing. Cambridge, UK: Cambridge University Press.
Shulman, L. (1986). Those who understand: knowledge growth in teaching. Educational Researcher,
15(2), 4–14.
Shulman, L, (1987). Knowledge and teaching: Foundation of the new reform. Harvard Educational
Review, 57(1), 1–22.
Stacey, K (2004). International perspectives on the nature of mathematical knowledge for secondary
teaching: Progress and dilemmas. In M. J. Hoines & A. B. Fuglestad (Eds), Proceedings of the 28th
Conference of the International Group for the Psychology of Mathematics Education – PME 28 (pp.
167–196). Bergen, Norway.
Stacey, K. & Vincent, J. (2009). Modes of reasoning in explanations in Australian eighth-grade
mathematics textbooks. Educational Studies in Mathematics. Published online 25 March 2009.
Springer.