TEN-YEAR-OLD STUDENTS SOLVING

LINEAR EQUATIONS
BRBARA BRIZUELA, ANALCIA SCHLIEMANN [1]
We seek to re-conceptualize the perspective regarding students difficulties with algebra. While acknowledging that
students do have difficulties when learning algebra, we also
argue that the generally espoused criteria for algebra as the
ability to work with the syntactical rules for solving equations is rather narrow. Certainly, learning and understanding
the syntactic rules for solving equations should be part of
algebra instruction. However, algebra also involves work
with other, multiple notation-systems (such as function
tables, Cartesian coordinate graphs) and with variables and
functions. Therefore, in our teaching experiments, it is only
after the students have demonstrated familiarity with the concepts and representations of variables and functions, we
introduce equations and work with algebraic syntax.
The crux of the argument of this article is that if we can
present evidence of younger, elementary school children
engaging with algebra, and using and understanding the syntactic rules of algebra, we have to ask ourselves why so
many adolescents face difficulties with algebra. Perhaps it is
not that the students are not prepared or ready for learning
algebra, but that the teaching or curriculum to which the students have been exposed has been preventing them from
developing mathematical ideas and representations they
would otherwise be capable of developing.
It is our belief that, as previously stressed by many, the
difficulties middle- and high-school students have with
algebra result from their previous experiences with a mathematics curriculum that focuses exclusively on arithmetic
procedures and computation rules. In this article, we
describe classroom and interview data showing that, if children are given the opportunity to discuss algebraic relations
and to develop algebraic notations, even ten-year-old students can solve algebraic equations with unknown amounts
on both sides of the equals sign.
Background
Numerous researchers have suggested that algebra should
pervade the mathematics curriculum and be introduced at an
early age, thus avoiding an entrenchment of arithmetic (e.g.,
Bodanskii, 1991; Booth, 1988; Brown and Coles, 2001;
Crawford, 2001; Davis, 1985; Dougherty, 2003; Henry,
2001; Kaput and Blanton, 2001; Schoenfeld, 1995; Vergnaud, 1988; Warren, 2001). They agree that young students
can grasp both algebraic concepts and notation, and that the
boundaries between the arithmetic and algebra curricula can
and should be downplayed.
Many researchers have suggested that students difficulties with algebra arise from their previous lack of experience
with arithmetic and algebraic generalizations that should be
integral to the arithmetic curriculum:

the difficulties that students experience in algebra are

not so much difficulties in algebra itself as problems in
arithmetic that remain uncorrected. (Booth, 1988, p. 29)
Given the problems in arithmetic, researchers have turned
to introducing and embedding the elementary school curriculum with algebraic concepts. But, can children do it? We
will illustrate this kind of work by describing how a group of
ten-year-old students come to produce and solve linear equations with unknowns and variables on both sides of the equal
sign, an achievement that has been thought to be out of reach
for many middle- and high-school students.
There is relatively wide agreement among mathematics
educators and among policy makers (see, for instance,
NCTM, 2000; Sutherland, 2002) that algebra should become
part of the elementary school curriculum. How this should
be done, however, is still a matter of discussion and for systematic research. Successful implementation of algebraic
activities and discussions in elementary school have been
described by, for example, Bodanskii (1991); Carpenter and
Franke (2001); Davis (1985); Dougherty (2003); Fujii and
Stephens (2001); Kaput and Blanton (2001); and Schifter
(1999). These researchers have aimed at algebra in the early
grades, but their approaches have varied and focused on different aspects of what is considered to be algebra.
A first group of studies focused on the use of written equations and algebraic syntax, on childrens understandings of
equalities and inequalities, and on the use of letters to represent variables and unknowns:
Bodanskii (1991), in Russia, focused mainly on the
development and solution of written equations to solve
verbal problems. He found that ten-year-old students
who were introduced to algebra problems and notation
for equations from first grade (six- or seven-year-olds)
performed significantly better than twelve- and thirteen-year-old students who only had access to algebra
from age eleven.
Brito Lima (Brito Lima and da Rocha Falco, 1997)
and Lins Lessa (unpublished doctoral dissertation) in
Brazil, highlight issues related to both algebraic concepts and notation for equations. In their studies they
have shown that elementary school children can
develop written representations for algebraic problems
and that their success on algebraic problems is based on
the development of written equations and on the use
of algebra syntactic rules for solving equations.
Carpenter and Franke (2001) focus on algebraic thinking
through childrens discussions on equations and inequalities. They show that young children, who participated in

For the Learning of Mathematics 24, 2 (July, 2004)

FLM Publishing Association, Kingston, Ontario, Canada

33

classroom activities that explore mathematical relations, can understand and explain that an equation such
as a + b b = a is true for any numbers a and b.
More recent experiences of introducing algebraic concepts from the early school years have been developed by
the Measure up project in Hawaii (Dougherty, 2003). The
basic premise behind the project is the need to think in a
non-traditional way about what constitutes appropriate
mathematics for young children. What Dougherty and her
colleagues propose, following Davydovs (1991) work, is
for children to begin their mathematics instruction with no
numbers. Instead, they focus on comparing physical attributes of objects. These comparisons and relationships
between attributes are described with relational statements,
using letters. Only gradually are numbers introduced into the
curriculum, without ever setting aside the main focus on
comparisons and relationships. In this project, children as
young as five and six years old deal with equalities, inequalities, differences, commutativity, associativity, and inverses.
Daviss (1985) work in the Madison Project takes a different approach and focuses on eleven-year-old students
working with concepts and notations for variables, Cartesian
coordinates, and functions. Our own work has shown that
nine-year-old students can learn to think of arithmetical
operations as functions rather than merely as computations
on particular numbers, that they can operate on unknowns
(Carraher et al., 2001) and work with mapping notation [2],
such as n 2n 1 (Carraher et al., 2003). We have also
found that function tables, graphs of linear functions, and
establishing connections between tables and graphs are
within reach of ten-year-old students (Brizuela, 2004;
Schliemann and Carraher, 2002).
These demonstrations, however, may not have convinced
some mathematics educators that young children can learn
algebra. Previous research has highlighted students difficulty
in solving equations when unknown quantities appear on both
sides of the equality (e.g. Filloy and Rojano, 1989; Herscovics and Linchevski, 1994). These researchers have attributed
such findings to developmental constraints and the inherent
abstractness of algebra, concluding that even adolescents may
not be ready to learn algebra (Collis, 1975; Filloy and Rojano,
1989; Herscovics and Linchevski, 1994; Linchevski, 2001;
MacGregor, 2001; Sfard and Linchevski, 1994).
Some researchers have claimed that students are engaging
in algebra only if they can understand and use the syntax of
algebra and solve equations with unknown amounts on both
sides of the equals sign (see Filloy and Rojano, 1989). Filloy
and Rojano (1989), for example, have proposed that there is a
cut-point separating arithmetic from algebraic thought. Similarly, Herscovics and Linchevski (1994) point to students
inability to operate spontaneously with or on the unknown (p.
59, emphasis in the original) and refer to their failure to represent and manipulate unknowns as a cognitive gap.
Our approach to algebra notation
Algebra-symbolic notation is one of several basic representational systems of mathematics. In a narrow sense,
algebraic reasoning is concerned only with algebra-symbolic
notation. In the broad sense that we adopt in our research
34

and in this paper, algebraic reasoning is associated with and

embedded in many different representational systems.
Although some educators argue against any and all uses of
algebra-symbolic notation in the early grades, we feel it is
better to frame the issue in a broad context. By broad context
we mean to ask more generally how (different) written notations relate to mathematical reasoning and algebraic concepts in particular. Because algebra-symbolic notations are
embedded and associated with algebra reasoning, and are
integral to algebra, in our research we focus both on childrens reasoning and on their use and understanding of
algebra-symbolic notation.
Conventional notations help extend thinking (Cobb, 2000;
Lerner and Sadovsky, 1994; Vygotsky, 1978), but if they
are introduced without understanding, students may display
premature formalization (Piaget, 1964). For these reasons,
students need to be introduced to mathematical notations in
ways that make sense to them.
Our approach relies on introducing new notations as
variations on students spontaneous ways of representing
open-ended verbal problems. Our classroom intervention
data has shown that young students can learn to use algebra-symbolic notation meaningfully to express generalizations they have reached while exploring problems in
open-ended rich contexts. We have found that children can
use mathematical notations not only to register what they
understand, but also to structure and further their thinking, allowing them to make inferences they might
otherwise not have made. We also showed that algebra
notation can constitute a tool for generalizations, for understanding of linear functions, and for solving problems
(Brizuela, 2004).
Our next step was to investigate whether elementary
school children could also deal with written algebra equations and with the syntactic rules of algebra. Previous
studies have shown that even seven-year-old students can
understand a basic principle implicit in the rules for solving equations, namely, that adding or subtracting the same
amount to two equal amounts does not destroy an equality, and that nine-year-old students can develop notations
for algebraic problems and, with help from the interviewer,
solve linear equation problems using different solution
strategies, including the use of the syntactic rules of algebra (Bodanskii, 1991; Brito Lima and da Rocha Falco,
1997).
In the longitudinal study we partially report here, we
introduced children to equations as an extension of their
work on functions, function tables, and graphs of linear
functions. We will briefly describe the general approach we
adopted for introducing children to algebra equations and
will report on the final results of our study concerning tenyear-old students use of algebra notation and strategies for
solving equations. We will describe in detail:

the students discussions and written materials produced during the last lesson in one of the four
classrooms we worked in, and

the same students responses to an individual interview.

The classroom intervention and its results

We worked with 70 students in four classrooms, from grades
2 (children between 7 and 8 years of age) to 4 (children
between 9 and 10 years of age). Students were from a multiethnic community (75% Latino) in Greater Boston, and
83.09% of the students in the school were on the free or
reduced meal program.
Each semester, from the beginning of their second semester in second grade to the end of their fourth grade, we
implemented and documented six to eight Early Algebra
activities in their classrooms, each one lasting about 90 minutes. The activities related to arithmetic operations, fractions, ratio, proportion, and negative numbers. Our goal was
to examine how, as they participated in the activities, the students would work with variables, functions, positive and
negative numbers, algebra notation, function tables, graphs,
and equations.
The last six lessons we taught in fourth grade focused on
equations and algebra notation. Each lesson focused on a
problem that had unknown amounts in it and that could be
represented with equations, as in the two examples shown in
Figure 1.
Problem 1:
Mike and Robin each have some money. Mike has $8 in
his hand and the rest of his money is in his wallet. Robin
has altogether exactly three times as much money as
Mike has in his wallet. How much money could there
be in Mikes wallet? Who has more money?
Problem 2:
Which phone plan is better? Plan #1: You pay $0.10 per
minute for all calls. Plan #2: You pay $0.60 per month
plus $0.05 per minute for calls.
Figure 1: The two examples.
When presented with the problems, children were not asked
to find a right answer, but to consider all possibilities, to
draw the graphs of two functions, and to consider an answer
only after they had gone through these steps. During the
weeks leading up to the lesson we will focus on, the children
felt fairly comfortable dealing with unknown amounts and
some of the children were able to gradually use N to represent the unknown amounts, although some of them still used
iconic notations. During the last lesson in fourth grade, the
problem in Figure 2 was presented to the class.

We chose this problem as it reflects part of the controversy

regarding what exactly counts as algebra. This problem can
be represented through the following equation:
y + 2x + 7 = y + x + 20.
The problem involves variables and unknowns on both sides
of the equal sign, following Filloy and Rojanos (1989) constraints for when students are doing algebra. In this way we
also adjust to Herscovics and Linchevskis (1994) constraint
of learning algebra as involving manipulations of the
unknown, which our students would have to do if they
worked on the above problem.
We started off the lesson acting out the problem. At the
front of the class, a box, two tubes, and seven candies in
a transparent bag were put on a table that we labeled as
Brianas table. Next to this table, a bag, a tube, and twenty
candies in a transparent bag were put on a table that we
labeled as Susans table. Figure 3 shows the distribution of
elements on the tables.

Figure 3: The distribution of elements on the table for Susan

and Brianas candy problem.
In our work with the children, we had encouraged them to
use two different kinds of strategies when dealing with algebraic problems. The first had to do with matching up
amounts on both sides of an equal sign, or matching up
amounts belonging to two different people, such as Susan
and Briana. The second strategy had to do with canceling out
equal amounts on both sides of an equal sign or amounts
belonging to two different people. Using these different
strategies, we can think of at least two different approaches
to dealing with the problem at hand. Figure 4 shows one possible approach.
In this approach, you cancel or match up Susan and Brianas boxes as well as Susans tube of candies and one of
Brianas tubes. At this point, we are left with both girls
loose candies and one of Brianas tubes. Assuming that both
girls have the same amount of candies, we think of the
amount of candies in Susans bag as two sub-amounts: the

Two students have the same amount of candies. Briana

has one box, two tubes, and seven loose candies. Susan
has one box, one tube, and 20 loose candies. If each box
has the same amount and each tube has the same
amount, can you figure out how much each tube holds?
What about each box?
Figure 2: The problem presented during the last lesson in
fourth grade.

Figure 4: One potential approach to the problem.

35

Figure 5: Another potential approach to the problem.

amount that corresponds to Brianas loose candies (7) and
what is left of Brianas candies the tube. Given that 20 7
= 13, we can assume that Brianas leftover tube has 13
candies.
Figure 5 shows another potential approach to the problem.
In this approach, the two boxes of candies are matched up,
as well as one of each of the girls tubes of candies. Then,
again assuming that both girls have the same amount of candies, we can assume that Susans loose candies 20 are
equal to Brianas loose candies 7 and her leftover tube,
thus 7 + x = 20 and the amount in the tube must be 13 candies. These are only two potential approaches. We envisioned children adopting one or the other approach, or a
combination of both, or some original way of dealing with
the problem of figuring out how many candies were contained in the tubes.
The researcher teaching this particular group of students
(David) generated a teaching and learning environment that
was conducive to childrens presentation of their own perspectives, ideas, and ways of representing problems. An
open-ended word problem was usually presented to the children at the beginning of each ninety-minute lesson.
First, childrens reactions to the problem were brainstormed. Then, children were asked to show, on paper, their
ideas about the problem and their suggested solutions, if
these were asked for. These notations were then shared by
the children at the front of the class. Usually, this involved
David selecting some representative examples to present to
the whole class. These group presentations then led to some
more group discussion about the problem and to some sort
of consensus about the problem and its solutions. David was
the teacher for this group of students only during our Early
Algebra interventions. At other times, the children had a
classroom teacher who was a participant observer to our
lessons. Our lessons responded to the needs presented by the
teacher, as well as the state curriculum frameworks that the
childrens curriculum was designed around.
Given the scenario at the front of the class, the students
started by discussing the problem. Aarielle recalled that this
problem was similar to the wallet problem (see above)
they had solved six weeks before. Kauthamy stated that
Susan had thirteen more candies in her bag than Briana did
in hers, and Albert observed that Briana had an extra tube
of candy. Thus, both Kauthamy and Albert started out by
highlighting the differences between Susan and Brianas
amounts; Susan has 13 more loose candies, and Briana has
36

an extra tube of candies. The combination of these initial

observations will prove key in the childrens dealing with
the problem.
When David asked if they could figure out how many
candies there were in a tube or in a box, most of the students answered that they could not. However, less than
fourteen minutes into the class, Albert continued to elaborate
on his initial idea and he explained that Brianas tubes had to
have thirteen candies in them so that one of her tubes plus
the seven loose candies could be equal to Susans twenty
loose candies! At once, other children in the class agreed
with Albert.
Up to this moment, the children in the class focused,
spontaneously, on the amount of candies contained in the
tubes. Thus, they were focusing on the unknown amounts on
both sides of the equal sign, as defined by an algebraic
expression of the problem. At this point, however, Cristian
noted that it did not matter how many candies were in the
boxes, thus pointing, for the first time, to the variable nature
of the amount of candies in the boxes. The amount of candies contained in the boxes was variable, and could be given
any possible value as long as both boxes each contained the
same number of candies. Cristian also explained that if
Brianas tube had 13 candies, then these candies added to the
7 loose candies would make it 20, plus 13 more for her second tube would be 33. For Susan, she has 13 in her tube
plus the 20 loose candies, which is 33. He said that because
of this, the two girls total amounts are equal no matter how
many candies are contained in the boxes. Figure 6 shows
Cristians thinking in a diagrammatic way.

Figure 6: Cristians thinking about the problem.

Mariah, however, unsure of Alberts reasoning, asked him to
explain why he thought there were thirteen candies in each
tube. He answered that the amount in a tube plus the seven
loose candies that Briana had, had to be equal to Susans
twenty loose candies. Mariah then asked about Brianas
second tube and Albert assured her that his proposal would
still work because Susan also had one tube. Figure 7 shows
Alberts thinking in a diagrammatic way.
Carissa further explained that the candies in Susans bag
the loose candies made up for the extra tube that Briana
had. A little later, Carissa explained that Briana and Susan
had the same amount of candies and thus Susans bag, that
had twenty loose candies, was really like thirteen plus seven.
She also explained that, when comparing the loose candies

Nancys written work (see Figure 9) is an example of an

iconic notation. She first worked with the amounts given for
the loose candies (twenty and seven) and correctly used the
difference of thirteen between these two amounts as the value
for what was inside the tubes, showing one tube on Susans
table and both of Brianas tubes as having thirteen in them.

Figure 7: Alberts thinking about the problem.

that each one of the girls had, Susan ended up having thirteen extra candies; those thirteen extra candies that Susan
had matched up exactly with the amount in Brianas extra
tube. Thus, Susans extra candies when compared with
Brianas loose candies 13 could be placed in an extra tube
for her, making Briana and Susan each have two tubes of
candies, one box, and 7 loose candies. Figure 8 shows
Carissas thinking in a diagrammatic way.

Figure 9: Nancy.

Figure 8: Carissas thinking about the problem.

David (the instructor) asked how many candies in Susans
bag (the loose candies) made up for Brianas extra tube and
Albert replied thirteen, which would leave seven extra candies (the amount of loose candies Susan had). A few minutes
later, David asked, How do we know that the tubes have
thirteen and that the girls are holding the same amount if we
havent peeked in the tubes yet? Cristian replied that this is
called algebra and Briana and Mariah explained that they
used algebra to subtract and make educated guesses. David
challenged the children to prove that there were indeed thirteen candies in a tube. Cristian explained that we could use
N to stand for a tube, and the class as a whole agreed that a
different letter should be used to stand for the boxes.
Subsequently, the students sat down to work on ways of
representing the problem in writing. Each of the students in
the class produced their written account of the problem.
Although most of the children in this class of eighteen students made iconic notations for the problem (78%), one third
of them included an equation in their notation and more than
one third (39%) included a letter to stand for one or more of
the unknown values.

Although Nancy acknowledged that Susan started out

with twenty loose candies, on the table she showed her as
having seven and thirteen just like Briana. One interesting feature of Nancys work is the question mark that she
placed on the two boxes the amount of candies in the
boxes was unknown (hence the question marks) and would
and could remain unknown to the very end of the problem.
In our longitudinal study, this was not the first time that we
had observed children using question marks to represent
unknown amounts.
Ramns written work (see Figure 10) is also interesting in
the way in which he was able to integrate both iconic and
algebra notations (N + 7). He consistently used the N to

Figure 10: Ramn.

37

show what was in the tubes of candies. In addition, he used

his notation to solve the problem and show his solution to
the problem. He represented Susans (1) and Brianas (2)
candies iconically, then matched what they had and crossed
out the matching amounts on both sides. He did not assign
a value for the boxes and appeared to have no problem
crossing them out since there was one on each side of his
notation. Through his matching, he arrived at the conclusion that, in order to have equal amounts of candy, Susan
must be left with seven and thirteen candies (totaling twenty
candies) and Briana must be left with N + 7, making N be
thirteen candies.

Cristian set Susans and Brianas amounts equal and

matched the elements on both sides of the equation.
Once the group as a whole met to discuss the problem and
the students written notations, David wrote on the board
that T is the amount in each tube and B is the amount in each
box. He then asked the class, How much did Briana have?
The class confidently called out:
2T + B + 7.
He then asked, How much did Susan have? and the class
called out:
T + B + 20.
When David asked whether these expressions could be
simplified, Albert suggested matching up the Bs. David did
so and crossed them out, saying:
Now we have 2T + 7 and on the other side we have T
+ 20. How could we simplify them further?

Figure 11: Albert.

Carissa, turning to the scenario shown on the tables at the

front of the class, suggested putting seven candies in Susans
bag and leaving thirteen out in a pretend tube. David did so
and Aarielle wrote on the board, breaking up the twenty into
thirteen and seven and matching up the two sevens. David
erased the 7 from the board to leave 2T = T + 13. Cristian
suggested matching two tubes to leave T = 13. David did this
with the actual tubes and recorded it on the board: So T
has to be 13. Lets count the tube candies. Subsequently,
Kauthamy counted the candies and found thirteen. The children shouted out, Hooray!, expressing their excitement at
their solution to the problem.

Figure 11 shows that Albert used an equation to represent

the problem. He used both N and Z as the unknowns. He
started by using N to represent the amounts in the tubes and
in the boxes but soon used Z for the amount of candy in the
tubes. After matching the equal amounts, he appeared to
have used the letters interchangeably as he finally reached
the equation 20 = N + 7.

Interview results
At the end of the school year, one to four weeks after the last
class, we individually interviewed the children on a series of
problems. In the last part of the interview, children were
asked to represent in writing and to solve the problem shown
in Figure 13.
Harold has some money. Sally has four times as much
money as Harold. Harold earns $18.00 more dollars.
Now he has the same amount as Sally. Can you figure out
how much money Harold has altogether? What about
Sally?
Figure 13: The interview problem.

Figure 12: Cristian.

Figure 12 shows a very sophisticated notation by Cristian.
Although similar to Alberts, Cristians notations are of
added relevance given the explanation written out at the side
of the equations that he matched up:
I broke 20 into 7 and 13. Then matched 7 and 7. Then
broke 2N to N and N and matched them. Then 13 = N.

38

Of the eighteen children from this class who were interviewed, ten represented Harolds initial amount as N, X or H,
and Sallys amount as N 4. For Harolds amount after earning 18 more dollars, eight children wrote N + 18. Four
children wrote the full equation N + 18 = N 4 and eight
children correctly solved the problem. However, only one
systematically used the method of matching up equivalent
terms on both sides of the equal sign to simplify the equation. Another child, when prompted, correctly explained this
method. Apparently, as the children worked in their written
notations, they easily inferred that Harolds starting amount
was 6. As Albert stated, I thought about six because it just
popped in my head.

Discussion
The lesson, and childrens reactions to the problem presented, suggest a number of issues related to algebra in the
mathematics curriculum. First of all, we have presented
evidence of young, elementary-school children doing algebra. The kind of problem they engaged with could be
represented by an expression that included unknown and
variable amounts on both sides of the equal sign, thus satisfying the constraints set for a particular vision of algebra as
involving the use of algebraic syntax.
One third of the children used an algebraic expression to
represent the problem, and over a third of the children
included a letter to stand for one or more of the unknown
values. Thus, we return to the question posed at the beginning of this paper: if such young children can succeed at
working and understanding algebra, why is it that so many
adolescents have trouble with algebra?
The children we were working with had been exposed to
from six to eight early algebra activities each semester for
the previous year and a half. Thus, we could argue that it was
this exposure that enabled them to approach the problem
successfully, to use algebraic expressions, and letters in their
notations. However, this still leaves two issues exposed: the
students are young, and doing at most 24 activities over the
course of a year and a half is not significant.
So we come back to the same argument presented at the
beginning of this paper: the location of the problem cannot be
found in the students themselves. If younger students were
able to work with algebra, older students could potentially
as well. It is most likely the type of instruction that the students are exposed to that is impacting on their performance in
algebra. If students were exposed to an algebrafied mathematics curriculum from the onset of their education, it is
likely that by the time they are adolescents, they would be
able to deal with a lot more complex mathematics.
The kinds of activities we developed over the last six
weeks of our longitudinal study were not simple or easy for
the students. Nevertheless, they were able to deal with the
challenges we proposed and, at the end of only six meetings
on equations, many were able to represent, discuss meaningfully and analyze problems involving unknown amounts
on both sides of an equality. In the classroom, at least a third
of the students in this class could represent the problem as
an equation, solve the equation, and meaningfully explain
why they could manipulate the elements in the equation. In
the interviews, more than half of the children correctly represented the amounts in the problem using letters to stand
for unknown amounts.
Our results suggest that dealing with equations is not
beyond ten-year-old students mathematical understanding
and that much more could be achieved if the same kind of
activities become part of the daily mathematics classes
offered to elementary school children.
Notes
[1] This article is part of a larger longitudinal study sponsored by the
National Science Foundation (Grant #9909591, awarded to D. Carraher and
A. Schliemann, see earlyalgebra.terc.edu).
[2] We use the term notation to refer to written, external representations (see
Lee and Karmiloff-Smith, 1996, who also state that these notations belong
to notational systems with which they share a set of rules).

References
Bodanskii, F. (1991) The formation of an algebraic method of problemsolving in primary school children, in Davydov, V. (ed.), Survey of
applied Soviet research in school mathematics education, the University of Chicago, Soviet studies in mathematics education 6, Psychological abilities of primary school children in learning mathematics,
Reston, VA, NCTM, pp. 275-338.
Booth, L. (1988) Childrens difficulties in beginning algebra, in Coxford, A. and Shulte, A. (eds), The ideas of algebra, K-12: 1988 yearbook,
Reston, VA, NCTM, pp. 20-32.
Brito-Lima, A. and da Rocha Falco, J. (1997) Early development of algebraic representation among 6-13 year-old children: the importance of
didactic contract, in Pehkonen, E. (ed.), Proceedings of the 21st Conference of the International Group for the Psychology of Mathematics
Education, Lahti, Finland, University of Helsinki, Lahti Research and
Training Centre, 2, pp. 201-208.
Brizuela, B. (2004) Mathematical development in young children: exploring notations, New York, NY, Teachers College Press.
Brown, L. and Coles, A. (2001) Natural algebraic activity, in Chick, H.,
Stacey, K., Vincent, J. and Vincent, J. (eds), Proceedings of the 12th
Study Conference of the International Commission on Mathematical
Instruction (ICMI): The future of the teaching and learning of algebra 1,
Melbourne, Australia, Department of Science and Mathematics Education, The University of Melbourne, pp. 120-127.
Carpenter, T. and Franke, M. (2001) Developing algebraic reasoning in the
elementary school: generalization and proof, in Chick, H., Stacey, K.,
Vincent, J. and Vincent, J. (eds), Proceedings of the 12th Study Conference of the International Commission on Mathematical Instruction
(ICMI): The future of the teaching and learning of algebra 1, Melbourne, Australia, Department of Science and Mathematics Education,
The University of Melbourne, pp. 155-162.
Carraher, D., Schliemann, A. and Brizuela, B. (2001) Can young students
operate on unknowns?, in van den Heuvel-Panhuisen, M. (ed.), Proceedings of the 25th Conference of the International Group for the
Psychology of Mathematics Education, Utrecht, The Netherlands,
Freudenthal Institute, 1, pp. 130-140.
Carraher, D., Schliemann, A. and Brizuela, B. (2003) Treating operations
as functions, in Carraher, D., Nemirovsky, R. and DiMattia, C. (eds),
Media and meaning. CD-ROM issue of Monographs for the Journal of
Research in Mathematics Education.
Cobb, P. (2000) From representations to symbolizing: introductory comments on semiotics and mathematical learning, in Cobb, P., Yackel, E.
and McClain, K. (eds), Symbolizing and communicating in mathematics classrooms: perspectives on discourse, tools and instructional
design, Mahwah, NJ, Lawrence Erlbaum, pp. 17-36.
Collis, K. (1975) The development of formal reasoning, Newcastle, Australia, University of Newcastle.
Crawford, A. (2001) Developing algebraic thinking: past, present, and
future, in Chick, H., Stacey, K., Vincent, J. and Vincent, J. (eds), Proceedings of the 12th Study Conference of the International Commission
on Mathematical Instruction (ICMI): The future of the teaching and
learning of algebra 1, Melbourne, Australia, Department of Science
and Mathematics Education, The University of Melbourne, pp. 192-198.
Davis, R. (1985) ICME-5 report: algebraic thinking in the early grades,
Journal of Mathematical Behavior 4(2), 195-208.
Davydov, V. (ed.) (1991, original work published 1969) Survey of applied
Soviet research in school mathematics education, The University of
Chicago 6, Psychological abilities of primary school children in learning mathematics, Reston, VA, NCTM, pp. 275-338).
Dougherty, B. (2003) Voyaging from theory to practice in learning: Measure Up, in Pateman N., Dougherty, B. and Zilliox, J. (eds), Proceedings
of the 2003 Joint Meeting of PME and PME-NA 1, Honolulu, HI, University of Hawaii, pp. 17-23.
Filloy, E. and Rojano, T. (1989) Solving equations: the transition from
arithmetic to algebra, For the Learning of Mathematics 9(2), 19-25.
Fujii, T. and Stephens, M. (2001) Fostering an understanding of algebraic
generalization through numerical expressions: the role of quasi-variable, in Chick, H., Stacey, K., Vincent, J. and Vincent, J. (eds),
Proceedings of the 12th Study Conference of the International Commission on Mathematical Instruction (ICMI): The future of the teaching and
learning of algebra 1, Melbourne, Australia, Department of Science
and Mathematics Education, The University of Melbourne, pp. 258-264.
Henry, V. (2001) An examination of educational practices and assumptions

39

regarding algebra instruction in the United States, in Chick, H., Stacey,

K., Vincent, J. and Vincent, J. (eds), Proceedings of the 12th Study
Conference of the International Commission on Mathematical Instruction (ICMI): The future of the teaching and learning of algebra 1,
Melbourne, Australia, Department of Science and Mathematics Education, The University of Melbourne, pp. 296-304.
Herscovics, N. and Linchevski, L. (1994) A cognitive gap between arithmetic and algebra, Educational Studies in Mathematics 27(1), 59-78.
Kaput, J. and Blanton, M. (2001), Algebrafying the elementary mathematics experience Part I: transforming task structures, in Chick, H.,
Stacey, K., Vincent, J. and Vincent, J. (eds), Proceedings of the 12th
Study Conference of the International Commission on Mathematical
Instruction (ICMI): The future of the teaching and learning of algebra 1,
Melbourne, Australia, Department of Science and Mathematics Education, The University of Melbourne, pp. 344-350.
Lee, K. and Karmiloff-Smith, A. (1996) The development of external symbol systems: the child as notator, in Goelman, R. and Kit-Fong Au, T.
(eds), Perceptual and cognitive development, San Diego, CA, Academic Press, pp. 185-211.
Lerner, D. and Sadovsky P. (1994) El sistema de numeracin: un problema
didctico [The number system: a didactical problem], in Parra, C. and
Saiz, I. (eds), Didctica de matemticas: aportes y reflexiones, Buenos
Aires, Argentina, Paids, pp. 93-184.
Linchevski, L. (2001) Operating on the unknowns: what does it really
mean?, in van den Heuvel-Panhuisen, M. (ed.), Proceedings of the 25th
Conference of the International Group for the Psychology of Mathematics Education, Utrecht, The Netherlands, Freudenthal Institute, 1, pp.
141-144.
MacGregor, M. (2001) Does learning algebra benefit most people?, in
Chick, H., Stacey, K., Vincent, J. and Vincent, J. (eds), Proceedings of
the 12th Study Conference of the International Commission on Mathematical Instruction (ICMI): The future of the teaching and learning of
algebra 2, Melbourne, Australia, Department of Science and Mathe-

matics Education, The University of Melbourne, pp. 405-411.

National Council of Teachers of Mathematics (2000) Principles and standards for school mathematics Reston, VA, NCTM.
Piaget, J. (1964) Development and learning, in Ripple, R. and Rockcastle,
V. (eds) Piaget rediscovered: a report of the conference on cognitive
studies and curriculum development, Ithaca, NY, Cornell University
Press, pp. 7-20.
Schifter, D. (1999) Reasoning about operations: early algebraic thinking in
grades K-6, in Stiff, L. and Curcio, F. (eds), Developing mathematical
reasoning in grades K-12: 1999 yearbook, Reston, VA, NCTM, pp. 62-81.
Schliemann, A. and Carraher, D. (2002) The evolution of mathematical
reasoning: everyday versus idealized reasoning, Developmental Review
22(2), 242-266.
Schoenfeld, A. (1995) Report of working group 1, in LaCampagne, C.
(ed.), The algebra initiative colloquium 2: Working group papers, Washington, DC, US Department of Education, OERI, pp. 11-18.
Sfard, A. and Linchevski, L. (1994) The gains and pitfalls of reification the
case of algebra, Educational Studies in Mathematics 26(2-3), 191-228.
Sutherland, R. (2002) A comparative study of algebra curricula, London,
UK, Qualifications and Curriculum Authority (QCA).
Vergnaud, G. (1988) Long terme et court terme dans lapprentissage de
lalgebre, in Laborde, C. (ed.), Actes du premier colloque franco-allemand de didactique des mathmatiques et de linformatique, Paris,
France, La Pense Sauvage, pp. 189-199.
Vygotsky, L. (1978) Mind in society, Cambridge, MA, Harvard University
Press.
Warren, E. (2001) Algebraic understanding: the importance of learning in
the early years, in Chick, H., Stacey, K., Vincent, J. and Vincent, J.
(eds), Proceedings of the 12th Study Conference of the International
Commission on Mathematical Instruction (ICMI): The future of the
teaching and learning of algebra 2, Melbourne, Australia, Department
of Science and Mathematics Education, The University of Melbourne,
pp. 633-640.

But what about me? Suppose I am a blind man, and I use a stick. I go tap, tap, tap. Where
do I start? Is my mental system bounded at the handle of the stick? Is it bounded by my skin?
Does it start halfway up the stick? Does it start at the tip of the stick? But these are nonsense
questions. The stick is a pathway along which transforms of difference are being transmitted.
The way to delineate the system is to draw the limiting line in such a way that you do not cut
any of these pathways in ways which leave things inexplicable. If what you are trying to
explain is a given piece of behaviour, such as the locomotion of the blind man, then, for this
purpose, you will need the street, the stick, the man; the street, the stick, and so on, round and
round.
But when the blind man sits down to eat his lunch, his stick and its messages will no longer
be relevant if it is his eating that you want to understand.
(Bateson, G. (2000, first edition 1972) Form, substance and difference, in Steps to an
ecology of mind, Chicago and London, The University of Chicago Press, p. 465.)

40