ANALYSIS OF THE UNDERLYING COGNITIVE

ACTIVITY IN THE RESOLUTION OF A TASK ON

DERIVABILITY OF THE ABSOLUTE-VALUE
FUNCTION: TWO THEORETICAL PERSPECTIVES

Luis R. Pino-Fan, Ismenia Guzmán, Vicenç Font, and Raymond Duval

This paper presents a study of networking of theories between the theory
of registers of semiotic representation (TRSR) and the onto-semiotic
approach of mathematical cognition and instruction (OSA). The results
obtained show complementarities between these two theoretical
perspectives, which might allow more detailed analysis of the students’
performance.
Keywords: Absolute value; Cognitive analysis; Derivative; Networking of
theories; Onto-semiotic approach; Theory of registers of semiotic representation

Análisis de la actividad cognitiva subyacente en la resolución de una

tarea sobre la derivabilidad de la función valor absoluto: dos perspectivas
teóricas
En este artículo se presenta un estudio de networking of theories, entre
la teoría de los registros de representación semióticos (TRRS) y el
enfoque onto-semiótico de la cognición e instrucción matemáticos
(OSA). Los resultados obtenidos revelan complementariedades entre
estas dos perspectivas teóricas cuya aplicación simultánea permitiría
hacer análisis más pormenorizados de las producciones de los
estudiantes.
Términos clave: Análisis cognitivo; Conexión de redes de teorías; Derivada;
Enfoque onto-semiótico; Teoría de los registros de representación semiótica;
Valor absoluto

Pino-Fan, L., Guzmán, I., Font, V., & Duval, R. (2017). Analysis of the underlying cognitive
activity in the resolution of a task on derivability of the absolute-value function: Two theoretical
perspectives. PNA, 11(2), 97-124.
L. Pino-Fan et al. 98

BACKGROUND

One of the main concerns of the research community on the mathematics

education is determining which are the difficulties that learners face on their way
to understanding, and therefore, learning, mathematical notions. This interest is
reflected in the fact that, over decades, one of the main focuses of research within
our scientific discipline has been the characteristics of the learner’s cognitive
activity. Furthermore, one of the most popular theories in the field is the theory
of registers of semiotic representation (TRSR) (Duval, 1995), which focuses on
the study of the cognitive activity of students when solving mathematical
problems, providing notions that make it possible to analyze and comprehend
how subjects use and link the different types of material representations and the
role that these representations have in the comprehension of mathematical
concepts.
On the other hand, an emerging theoretical model in the mathematics
education, which has been gaining more and more importance worldwide, is the
onto-semiotic approach (OSA) to mathematical cognition and instruction
(Godino, Batanero, & Font, 2007), which is a model that tries to articulate
several perspectives of the discipline and dimensions involved in the processes of
teaching and learning mathematics. Out of the many dimensions, our main
interest in this document is what is known as the cognitive facet, which
emphasizes the different types of mathematical objects involved in mathematical
practices developed in order to solve a certain mathematical task, and also in the
connections that subjects establishamong such objects and the meanings that they
give them in terms of the context in which these are used.
In this research, we aim at carrying out a comparative study between these
two theoretical approaches, the theory of registers of semiotic representation and
the onto-semiotic approach, which allows carrying out cognitive analysis from
the subjects’ performance. In order to conduct this study, and following the
proposed methodology for the works within the framework of the networking of
theories, we analysed the performance of a future high school teacher in a task
related to the differentiability of the absolute-value function.
In the following section, theoretical and methodological notions of the study
will be introduced, beginning with the features of the networking of theories that
were used, and then, the description of the theoretical and methodological tools
proposed by the TRSR and OSA for cognitive analysis. In the third section, the
task and the response protocol that will be the basis for the comparative study of
the theoretical perspectives is presented. In the fourth section, a detailed
development of the cognitive analysis from the TRSR and OSA perspectives is
done. In the fifth section, both perspectives of analysis are compared and the
approximations and complementarities between the two theoretical

PNA 11(2)
Analysis of the Underlying Cognitive… 99

methodologies used are performed. Finally, the conclusions of the study are
presented.

THEORETICAL AND METHODOLOGICAL NOTIONS

In what follows we summarize some aspects of the two theories involved in this
study. But first, we describe the features of the methodology of networking of
theories.

Networking of Theories
Currently, there are several theoretical positions that allow conducting cognitive
analyses—of students, prospective teachers or teachers—depending on what is
desired to observe and which is the concerned mathematical notion (Asiala et al.,
1996; Duval, 2006; Godino et al., 2007). However, the complex nature of the
subjects’ learning phenomena has directed research groups to make efforts to
revise and find possible complementarities between theoretical and
methodological approaches that allow providing more detailed and precise
explanations of such learning processes—e.g., Font, Trigueros, Badillo, and
Rubio (2016), carry out a comparative study between the APOS theory and the
OSA.
The idea of investigating the networking of theories is not new, as evidenced
in the works presented in the working group since the Fifth Congress of
European Research in Mathematics Education (CERME 5) to the past CERME 9,
on different approaches and theoretical perspectives in mathematics education
research. Artigue, Bartolini-Bussi, Dreyfus, Gray, and Prediger (2005) points out
that, as a research community, we need to be aware that discussion among
researchers from different research communities is insufficient to achieve
networking. Collaboration among teams using different theories with different
underlying assumptions is called for in order to identify the issues and the
questions. In general, research studiesthat have been performed in this area, have
explored ways of handling the diversity of theories in order to better grasp the
complexity of learning and teaching processes, and understand how theories can
or cannot be connected in a manner that respects their underlying assumptions.
In this regards, there are different strategies and methods to deal with this
type of studies. For example, Prediger, Bikner-Ahsbahs, and Arzarello (2008),
describe different connecting strategies and methods for articulating theories,
which range from completely ignoring other theoretical perspectives on the one
extreme end, to globally unifying different approaches on the other. As
intermediate strategies, the authors mention the need for making one’s own
perspective understandable and for understanding other perspectives, and the
strategies of comparing and contrasting different approaches, coordinating and
combining perspectives, and achieving local integration and synthesis. For our

PNA 11(2)
L. Pino-Fan et al. 100

study, we have utilized diverse strategies, within the framework of the methods
of networking of theories, which will be described below.
Firstly, a team of four researchers was formed, who are also the authors of
this document, two of which—the second and fourth author—have both vast
experience and deep knowledge in the use of the TRSR. The other two authors—
the first and third—have substantial experience in using the OSA. The first and
third authors use OSA mainly, but possess knowledge of TRSR too, and the
second author uses primarily TRSR but also has sufficient knowledge of OSA.
Therefore, an important first phase in networking theories is achieved: the need
to really understand the other theory.
Several authors have showed interest in determining the aspects that
characterize a theory in order to be able to classify it and compare it (e.g.,
Bikner-Ahsbahs & Prediger, 2010; Radford, 2008). According to Radford
(2008), essential elements of a theory include principles, methods, and
paradigmatic research questions. The principles of each theory entail a
positioning, explicit or implicit, about the nature of mathematical objects. For
this reason, the second step has been set to determine how both theories model
mathematical activity and what is their positioning, explicit or implicit, on the
nature of mathematical objects. This second step allows seeing the differences
and similarities between both theories and also providing a prior general idea on
how they can be coordinated.
In order to keep moving forward, once this first comparison between the
theories has been made, we have applied one of the basic principles of
networking of theories: to ensure that the job of connecting the theories be as
precise as possible. According to this principle, the third step was selecting a
specific mathematical object—the derivative—as context of reflection. The
reason for choosing the derivative over another topic is because it is a
mathematical object on which relevant research has been done by both theories,
utilizing the basic theoretical notions of each.
Once the mathematical object derivative had been selected, as fourth step we
chose a task where such object was used and one answer that, regarding the
application of the task to a sample of students, was provided by one of the
students. The selection of both the task and the answer has been done,
principally, by the first two authors and then agreed with the other two authors;
and it was selected due to the complex mathematical activity provided by the
student, which we refer to as Juliette. In such activity, Juliette shows difficulties
that are taken into account in the analysis conducted with two theoretical
perspectives.
As fifth step, we have analyzed the answer given by Juliette, from two
theoretical perspectives. On the one hand, we looked at the solution to the task
from the perspective of TRSR. This analysis has been basically carried out by the
two authors that use TRSR. On the other hand, we looked at the solution to the

PNA 11(2)
Analysis of the Underlying Cognitive… 101

task from the perspective of OSA. This analysis has been basically carried out by
the two authors that use OSA.
As sixth step, in this article we focused on the comparison of the principles
used in the cognitive analysis from both theoretical perspectives, and not on the
comparison of methods and paradigmatic general research questions between
TRSR and OSA. This step has been jointly carried out by the four authors in the
section on comparison of analysis.

Theory of Registers of Semiotic Representation (TRSR)

In cognitive psychology, the notion of representation plays an important role
regarding the acquisition and the treatment of an individual’s knowledge. As
Duval (1995) points out, “there’s no knowledge that can be mobilized by an
individual without a representation activity” (p. 15). However, in mathematics,
what matters are the various semiotic systems used to represent numbers,
functions, geometrical properties, etc. And unlike other areas of knowledge, any
mathematical activity always requires substituting some semiotic representation
for another, no matter the semiotic systems that are mobilized. This leads to
distinguish two quite different kinds of cognitive operations, which are known as
conversion and treatment: to substitute one semiotic representation for another,
only by changing the semiotic system mobilized; and to substitute two semiotics
representations within the same semiotic system.
Registers of semiotic representation are all the semiotic representations that
are used in mathematics for computing, deducing, solving—mathematically—
problems. They can be classified into four types according two criteria (Duval,
2006). First, the semiotic representations that are produced are either discursive
representations—numerical or algebraic expressions, definitions, descriptions,
among others—or non-discursive representation—geometrical figures, graphs,
diagrams. Second, their substitution can either be set out in algorithms—
numerical or algebraic expressions, graphs—or not be set out in algorithms—
language natural, heuristic exploration of geometrical figures. Treatments are
always specific to the type of register mobilized.
The cognitive analysis of mathematical activity in terms of registers, and
therefore, in terms of conversions and treatments, is based on three ideas that are
described below.
¨ There are as many different semiotic representations of the same
mathematical object, as semiotic registers used in mathematics.
¨ Each different semiotic representation of the same mathematical object
does not explicitly state the same properties of the object being
represented. What is being explicitly stated is the content of the
representation.
¨ The content of semiotic representations must never be confused with the
mathematical objects that these represent.

PNA 11(2)
L. Pino-Fan et al. 102

Take, for example, the mathematical object linear function. First, it can be
represented by mobilizing any of the three registers below (see Figure 1).

Figure 1. Task on recognition of the object linear function

Second, the content of a semiotic representation changes completely when we
jump from a register to another—the horizontal arrows. So, there is a cognitive
distance between the respective contents of two semiotic representations of the
same object. This cognitive distance varies according the type of the start register
chosen, for example, to give the data of a problem, and the register to mobilize in
order to solve the problem or to control the relevancy of the answer.
Third, when the student has achievedthe coordination between at least two
registers—all horizontal arrows between two columns—they are able to
recognize immediately the corresponding semiotic representation in another
register. Otherwise they confuse the mathematical object represented with the
particular content of the semiotic representation given and they remain blocked,
because there are no links at all between the registers.
This recognition is not a question of concept acquisition or definition. It
requires becoming aware of what arethe qualitative features from the two
respective contents that have to be matched. That is why specific tasks of
experimental exploration are needed—all vertical arrows in Figure 1. In these
tasks, the relevant variations are the qualitative variations and not the numerical
values variations.

PNA 11(2)
Analysis of the Underlying Cognitive… 103

From a mathematical point of view, what matters is treatment, not

conversion, because computing, deducing, depends on substitutions which are
made within the same register by applying properties. Treatment always depends
on the semiotic possibilities of internal substitutions or transformations of
representations that the register mobilized provides. Therefore, mathematical
treatments are cognitively more complex in the use of the register of natural
language and those registers that allow visualizing—graphs, geometrical shapes,
etc. That is why specific tasks are required in order to make students become
aware of the way of reasoning and visualizing in mathematics (Duval, 2008).
Now, in the reality of the mathematical work, conversions and treatments are
never separated, because mathematical activity always mobilizes, explicitly or
implicitly, two kinds of registers. In other words, the real mathematical activity
consists of conversions and treatments performed alternatively or in parallel—in
geometry, for example, even if, from a mathematical point of view, only
treatments are scientifically relevant. Therefore, the necessary steps in the
cognitive analysis of any mathematical activity are the following.
¨ Separating all the conversions that are, explicitly or implicitly, required
from the treatment.
¨ Coding the couple of registers mobilized for each conversion—start
register and arrival register.
¨ Making explicit the successive substitutions to perform mainly for the
treatment in natural language or in visualization registers.
Thus, we get a grid of all the cognitively heterogeneous tasks underlying a
mathematical activity that we want to introduce for teaching or for experimental
purposes. This grid allows to reach two purpouses that we describe below.
¨ To perform an accurate diagnosis of the difficulties and incomprehension
points encountered by students which are a barrier to success.
¨ To compare the student’s production with the information gathered in
other similar activities during different periods of time and to assess the
capacity of transferring to new situations.
A reliable and controllable interpretation of the students’ oral/written/drawn
production shall be based on comparison with other data gathered according to
the same device of observation and with one variation well-identified. This is
what allows determining the scope of the outcomes of a research.

The Onto-Semiotic Approach to Cognition and Mathematical Instruction

(OSA)
The OSA is a theoretical and methodological framework that has been developed
since 1994 by Godino and colleagues and has been described in several other
works (e.g., Font, Godino, & Gallardo, 2013). This theoretical framework arises
from the field of research of mathematics education and aims at articulating the
diverse dimensions—epistemic, cognitive, affective, interactional, mediational

PNA 11(2)
L. Pino-Fan et al. 104

and ecologic—that are implicit in the processes of teaching and learning

mathematics (Godino et al., 2007). Relevant for this work are the epistemic facet,
which makes reference to the distribution, during the time of teaching, of the
components of the institutional meaning (problems, linguistic elements, concepts,
propositions, procedures and arguments), and the cognitive facet, which refers to
the development of personal meanings—learning.
The notion of system of practices plays an important role for the teaching and
learning of mathematics within the OSA. Godino and Batanero (1994) refer to
the system of practices as “any performance or manifestation—linguistic or
not—carried out by someone in order to solve mathematical problems, to
communicate the solution to others, to validate the solution and to generalize it to
other contexts and problems” (p. 334). Font et al. (2013) point out that
mathematical practices can be conceptualized as the combination of an operative
practice, through which mathematical texts can be read and produced, and a
discursive practice, which allows reflecting on operative practices. These
practices can be carried out by one person—system of personal practices—or
shared within an institution—system of institutional practices.
Within the OSA, certain pragmatism is adopted since mathematical objects
are considered as entities that emerge from the systems of practices carried out in
a field of problems (Godino & Batanero, 1994). Font et al. (2013) put it this way:
“Our ontological proposal originates from mathematical practices, and these
become the basic context from which individuals gain experience and
mathematical objects emerge. Consequently, the object gains a status originated
from the practices that precede it” (p. 104). Ostensive objects—symbols, graphs,
etc.—and non-ostensive objects—concepts, propositions, etc.—intervene in
mathematical practices, which we evoke while doing mathematics and are
represented in a textual, oral, graphic, symbolic and even gestural way. New
objects emerge from the systems of operative and discursive mathematical
practices and these show their organization and structure (Godino et al., 2007). If
the systems of practices are shared within the core of an institution, then the
emerging objects will be considered as institutional objects, while, on the other
hand, if such systems correspond to one person, then these will be considered as
personal objects. The emergence of a personal object is progressive during the
history of a subject, as a consequence of experience and learning, while the
emergence of an institutional object is progressive over time.
In order to offer a finer and more pragmatic way to analyze the mathematical
practices developed in connection to certain problems, OSA introduces a
typology of primary mathematical entities—or primary mathematical objects,
that intervene in the systems of practices: (a) situations-problems (extra-
mathematical applications, exercises…); (b) linguistic elements (terms,
expressions, notations, graphs…) in diverse registers (written, oral, gestural…);
(c) concepts/definitions (introduced through definitions or descriptions: line,
point, number, function, derivative…); (d) propositions/properties (statements

PNA 11(2)
Analysis of the Underlying Cognitive… 105

about concepts); (e) procedures (algorithms, operations, calculation

techniques…); and (f) arguments (statements used to validate or explain
propositions and procedures, deductive or of another type).
Situation-problems are the origin or reason for the activity; language
represents the remaining entities and serves as an instrument for the action;
arguments justify the procedures and propositions that relate the concepts. These
primary mathematical objects are connected with each other, forming intervening
networks of objects, emerging from the systems of practices, which in OSA are
known as configurations. These configurations can be socio-epistemic (networks
of institutional objects) or cognitive (networks of personal objects).
Each one of the primary mathematical objects can be considered from
different dual facets or dimensions: personal–institutional, unitary–systemic,
expression–content, ostensive–non-ostensive and exemplar–type. Godino, Font,
Wilhelmi, and Lurduy (2011) point out that both of these dualities and primary
mathematical objects can be analyzed from a process-product perspective, which
entails the following processes: (a) institutionalization–personalization, (b)
generalization–particularization, (c) decomposition or analysis–composition or
reification, (d) materialization–idealization, and (e) representation–signification.
The emergence of primary mathematical objects, pointed out before, is linked,
respectively, to processes of problematization, communication, definition,
algorithmization, enunciation, and argumentation.
This way, the meaning of mathematical objects in OSA is, basically,
conceived in two ways: (a) from a pragmatic-anthropological perspective, which
deals with the relativity of the context—or system of mathematical practices—in
which these are used, in other words, as emerging and in the sense that is
assigned to them in the practices or systems of practices—institutional or
personal—in which these are mobilized; and (b) in terms of semiotic functions, a
notion that allows to make relations among the several entities in a referential
and operational way. According to Hjemslev (1943) and Eco (1976), a semiotic
functionis the correspondence or dependent relationship (or function) that is
established by a subject (person or institution) between an antecedent
(expression, signifier) and a consequent—content or meaning, according to a
criteria or correspondence code (rules, habits, agreements...). The content—o
consequent—of a semiotic function, and therefore, the meaning can be a personal
or institutional object, unitary or systemic, ostensive or non-ostensive (Godino et
al., 2007); similarly, this object can be a linguistic element, a definition, a
proposition, a procedure, an argument or problem (Font et al., 2013). According
to Peirce’s semiotic (1978), the OSA assumes that both the expression—
antecedent of a semiotic function—and the content—consequent of a semiotic
function—could be any kind of entity (primary mathematical object or process).

PNA 11(2)
L. Pino-Fan et al. 106

THE TASK AND ITS RESOLUTION: THE CASE OF JULIETTE

The task addressed in this study (Figure 2), has been the object of study in
several works (Pino-Fan, 2014; Tsamir, Rasslan, & Dreyfus, 2006), and
demands, from who solves it, the mobilization of representations (graphic,
symbolic and verbal), and argumentations that justify procedures.

Figure 2. Task on derivability of the absolute-value function

The selected protocol of resolution is the one given by Juliette to the task above
(Figure 3). The case of Juliette arises in connection to a research carried
conducted by the first author of this document (Pino-Fan, 2014), which is
oriented towards the implementation of a questionnaire to analyze partial aspects
of the knowledge of a sample of future teachers of mathematics.
As it can be observed in Figure 3, Juliette had some serious difficulties to
solve the task assigned. Her answer was selected from among other 93 answers,
due to the complexity of the cognitive activity reflected in her mathematical
practice.

PNA 11(2)
Analysis of the Underlying Cognitive… 107

Figure 3. Juliette’s solution to the task on derivability of the absolute-value

function

ANALYSIS OF THE PROTOCOL FROM TWO THEORETICAL

PERSPECTIVES
In this section, we develop a detailed cognitive analysis from the TRSR and OSA
perspectives.

Analysis from the TRSR Perspective

The formulation of the task contains elements of the verbal, graphic and
symbolic registers. The questions related to the symbolic and graphic
representation of the absolute-value function, combine the verbal register and
symbolic notations. The answers to these questions are expected to be given in
the form of any of the registers mentioned before—verbal, graphic, or symbolic.
In general, Juliette’s answers for the items (a), (b), and (c), show that she
knows the definition of the absolute-value function, and that she can express it in
the symbolic register. But regarding the derivative function, she shows
deficiencies, because even though she answers that if the graph of the function
presents a corner or peak on x = 0 then the function is not derivable, in her
upcoming arguments some confusions are perceived regarding the domain of the
f ¢ function. In the symbolic register, she represents the absolute-value function
by parts and for the non-negative values of x she writes the f derivative using
symbols. She recognizes f ¢(2) as image for x = 2 , but she is not successful
regarding x = 0 . Details per item are presented below.

PNA 11(2)
L. Pino-Fan et al. 108

¨ For item (a), the domain f ( x) = x is required. Juliette, being in the

graphic register, visualizes the typical edge of the graph of the absolute-
value function and she perceptively associates it with the non-
differentiability of the function. She does not make the domain of f ¢
explicit, and states that the function is not differentiable. She clearly
associates the top point or peak of the graph with the fact that the function
can not be derived in its domain. In other words, she does an incorrect
generalization. On the other hand, she represents the absolute-value
function in the symbolic register, expressing it by parts and stating that if
fwas differentiable, it would be so in its entire domain. In Juliette’s answer
to item a), an apparent separation between the graphic and symbolic
registers is perceived because she answers through the visualization of the
graph first, and then, she performs a treatment of the algebraic expression
of the absolute-value function through the symbolic register.
¨ For item (b) it is required to calculate f ¢(2) . Here, we can appreciate a
new contradiction in her answer. First, she answers that “ f ¢(2) cannot be
calculated”, and justifies her answer by citing her answer to item (a).
Then, she adds a hypothesis, “if it was differentiable, the graphic
representation would be a point on the cartesian plane”. It is not clear
whether the graphic representation of the function f ' is a point, or she is
only referring to x = 2 . In any case, we appreciate that she recognizes, in
isolation, that the notation f ¢(2) represents the ordinate of a point in the
cartesian plane, since she draws it but does not justifies it. It could be
stated that Juliette does a simple passage from symbolic to graphic.
¨ For item (c) the question is if it is possible to calculate f ¢(0) . Juliette is in
the symbolic register and suggests a new hypothesis, “if we take a
function of the type f ( x) = ì x, x ³ 0
í- x, x < 0 the derivative would be f ' ( x) = 1
î
[considering x ³ 0 ]”, and then draws the graphic representation of the
image of zero, f ¢(0) = 1. Although her treatment is coherent because she
justifies her answer deriving for the values x ³ 0 , obtains f ¢(0) = 1 and
graphs the corresponding point, she clearly does not master the absolute-
value function and does not articulate its algebraic representation, by
parts, with its graphic representation.
In general, the symbolic register has prevailed over the graphic register in
Juliette’s answers. She only used the graphic register to graph points, but she
failed to correctly read the graph of the absolute-value function. The question of
item (c) was a key question and required the articulation of the symbolic and
graphic registers, but despite the treatment and passages that can be observed in

PNA 11(2)
Analysis of the Underlying Cognitive… 109

her answers, the lack of mastering of her mathematical knowledge did not allow
her coordinate the registers at play—graphic and symbolic.

Analysis from the OSA Perspective

We have organized the analysis in two levels. A first level where the
mathematical practice carried out by Juliette is described in general terms—a
fairly more global analysis; and a second level where, in a meticulous way,
detailed information about primary mathematical objects, their meanings and the
processes mobilized in her practice, as well as the way in which she relates to
them—cognitive configuration of objects and processes—is provided.

Mathematical Practice
Firstly, we observe that Juliette begins her practice based on a visual justification
to answer, although wrongly, subtask (a), pointing out the existence of a peak at
the point of domain of the function x = 0 . From the beginning of her practice, we
can observe that Juliette confuses the non-derivability—local—at a point of
domain of the absolute-value function with, her misconception of, non-
derivability of the function—global. Later, Juliette writes the symbolic
definition, by parts, of the absolute-value function. We could say that, in a certain
way, such definition is correct, however, she does not make crucial
considerations, like for example, that the point of domain of the function x = 0
belongs to both f ( x) = x and f ( x) = - x . This fact leads her to a cognitive
conflict that is shown in her sentence “If it is considered as a function of the
type... the function would be differentiable in the whole domain, in other words,
(¥,-¥)”. This cognitive conflict generated from her visual interpretation of the
graph of the function—the function is not derivable since it has a peak in x = 0 —
in contraposition to her interpretation of the symbolic definition, by parts, of the
function—she considers that f ( x) = x exclusively for x ³ 0 , is what leads her to
give incorrect answers to the other subtasks.

Cognitive Configuration of Objects and Processes

Juliette mobilizes in her practice a series of primary mathematical objects—
linguistic elements, concepts/definitions, properties/propositions, procedures and
arguments—and processes—emerging—which are detailed below (see Table 1,
Table 2, Table 3, Table 4, and Table 5). Similarly, as part of the configuration of
objects and processes, the previous objects and processes that Juliette must
interpret and understand before starting her practice are identified.

PNA 11(2)
L. Pino-Fan et al. 110

Table 1
Linguistic Elements
Mathematical objects Meanings
Previous
LE1: f ( x) = x Symbolic representation of the absolute-value function.

LE2: “Graph of the Graphic-cartesian representation of the absolute-value

function” (Figure 1) function with domain (-5, 5). Also, all the real numbers can
be inferred as the domain of the absolute-value function, just
like Juliette does.
LE3: f ¢(2) Symbolic representation that denotes the derivative of the
absolute-value function at the point of domain x = 2 .
LE4: f ¢(0) Symbolic representation that denotes the derivative of the
absolute-value function at the point of domain x = 0 .
Emerging
x³0 Symbolic expression that determines the definition, by parts,
LE5: f ( x) = ì
í
x,
î- x , x<0 of the absolute-value function.

LE6: (- ¥, ¥) Symbolic-notational expression that refers, according to

Juliette, to the domain of the derived function.
LE7: f ¢( x) = 1 Symbolic expression that Juliette uses to denote the
derivative of the function at the point of domain x = 2 . This
expression is also used by Juliette to represent the derivative
of the function at the point x = 0 .
LE8: Graphic representation that denotes the derivative of the
function at the point of domain x = 2 .

LE9: Graphic representation that Juliette uses to refer to the

derivative of the function at the point of domain x = 0 , a
point at (0,1) of the cartesian plane.

Note. LE=Linguistic element.

Table 2 shows concepts and definitions related to mathematical practice carried
out by Juliette.

PNA 11(2)
Analysis of the Underlying Cognitive… 111

Table 2
Concepts/Definitions
Mathematical Objects Meanings
Previous
CD1: Function Particularized as absolute-value function and referred to
graphically as well as symbolically.
CD2: Derivative of a Particularized, with the sub-task (a), as the derivative of
function and its domain the absolute-value function.
CD3: Derivative at a point Specifically, at the points of domain of the function
x = 2 and x = 0 .
Emerging
CD4: Absolute value Defined symbolically as:
ì x, x ³ 0
x =í
î- x , x < 0
Juliette extrapolated this implicit definition of absolute-
value to her definition of absolute-value function.
CD5: Domain Of the derivative of the function, referred by Juliette as
(- ¥, ¥).
Note. CD=Concept/definition.
Table 3 shows properties/propositions that Juliette mobilizes in her practice.

Table 3
Properties/Propositions
Mathematical Meanings
Objects
Previous
PP1: “Examine the function Proposition that establishes a previous process of
f ( x) = x and its graph” enunciation-representation, which allows the linking of
the symbolic representation of the function with its
graphic representation.

PNA 11(2)
L. Pino-Fan et al. 112

Table 3
Properties/Propositions
Mathematical Meanings
Objects
Emerging
PP2: “...the function This proposition supposes the partial answer that
f ( x) = x is not Juliette provides to sub-task (a).
differentiable...”
PP3: “If it is considered as a Proposition that accounts for a definition process—
function of the type definition of the function by parts. This proposition also
ì x, x ³ 0 refers to: (a) a process of algorithmization, which allows
f ( x) = í the to derive, implicitly, the function by parts (this process
î- x , x < 0
of algorithmization refers to a procedure); and (b) a
function would be process of argumentation, in which a special treatment
differentiable in the whole for the derivative of the function at the point of domain
domain, in other words, x = 0 is not considered. This proposition also supposes
(- ¥, ¥)”. a partial answer by Juliette to subtask (a).
PP4: “... f ¢(2) can not be Partial answer to sub-task (b). Proposition that accounts
calculated...” for a process of argumentation of a graphic-visual type,
in which Juliette presupposes that, since the value
function is not differentiable because it has peaks, then
the derivative at the point of domain x = 2 , cannot be
calculated.
PP5: “... and if it was Partial answer to sub-task (b). Proposition that shows
differentiable [the absolute- evidence of the procedure of deriving by parts the
value function] then, the function. This proposition is reaffirmed with PP6.
graphic representation would
be a point on the cartesian
plane...”

PNA 11(2)
Analysis of the Underlying Cognitive… 113

Table 3
Properties/Propositions
Mathematical Meanings
Objects
PP6: “Considering the type This proposition could be decomposed into two. The
of function ‘LE5’, the first one would make reference to the processes of
derivative would be ‘LE7’ algorithmization and argumentation of PP3 and PP4. A
then it would represent a second one would refer to a process of representation in
point ‘LE8’” which Juliette links LE7 and LE8.
PP7: “...if the function ‘LE5’ Answer to sub-task c). This proposition refers to a
is considered, the derivative process of algorithmization that leads to the procedure
would be ‘LE7’ and the of deriving the function by parts, and to a process of
graphic representation would argumentation, based on LE5 and the process of
be ‘LE9’” algorithmization, that allows Juliette to find the
derivative for x = 0 . Likewise, this proposition also
makes reference to a process of representation in which
Juliette links LE7 and LE9.
Note. PP=Property/proposition.
We could also analyze the meaning that Juliette gives to the sentence that lies
before PP7, “As mentioned before [her answer to sub-task (b)]...”, as a
proposition that suggests as a first partial answer to sub-task c), that f ¢(0) cannot
be calculated because the graph of the function has a peak—as in the case of sub-
task (b). This is relevant because the two propositions (PP4 and PP5) that make
reference to partial answers to sub-task (b), show contradictions in the answers
that Juliette provides to sub-task (b).Thus, a cognitive conflict arises within her
that leads her to argue that the derivative in x = 2 cannot be calculated if based
on the graphic-visual aspects only, but can indeed be calculated if based on
purely symbolic-algorithmic aspects.The same occurs with her answer to sub-
task (c).

PNA 11(2)
L. Pino-Fan et al. 114

Table 4
Procedures
Mathematical
Objects Meanings
Emerging
P1: Graphic-Visual Procedures that allow Juliette to provide answers to
sub-tasks (a), (b) and (c), from her interpretation of the
graphof the absolute-value function. In the case of
Juliette’s answer, these kinds of procedures are
connected with the type of arguments A1. By means of
these types of procedures Juliette achieves propositions
such as PP2 and PP4.
P2: Calculation of the Procedures that Juliette utilizes to provide answers,
derivative from LE5 through propositions PP3, PP5 and PP6, to sub-tasks
(a), (b) and (c) respectively. Basically, this procedure
consists of the derivative by parts of the function, from
the definition CD4 that is also mentioned in LE5.
Note. P=Procedure.
The description of the arguments that Juliette mobilizes in her practice are
displayed in Table 5.
Table 5
Arguments
Mathematical
Objects Meanings
Emerging
A1: Graphic-Visual Argument centered on the visual consideration of the
“…because the peak in the graph of the function. This argument is
graph has a peak at the one Juliette uses to point out that the function is
the point x = 0 ” not differentiable and therefore, it is not possible to
calculate the derivative at the points of domain x = 2
and x = 0 .
A2: Symbolic- Argument presented by Juliette in which the process
Algorithmic of definition of the absolute-value function
(symbolically: CD4, LE5) and the process of
algorithmization that allows obtaining the procedure
of calculation of the derivative by parts (P2), are
considered.
Note. A=Argument.

PNA 11(2)
Analysis of the Underlying Cognitive… 115

As explained in a preliminary section (on the OSA to cognition and mathematical

instruction), the notion of meaning is a key notion within the OSA and it can
basically be conceived in two ways. In this sense and complementary to the
description of primary mathematical objects and processes that Juliette mobilizes
in her practice, we can identify the relevant semiotic functions that she
establishes to each one of the items of the task.
For item (a) Juliette establishes three semiotic functions. The first one has a
graphic-visual nature because, from the absolute-value graph—LE2, antecedent,
the proposition PP2—consequent—is enunciated, arguing such proposition in a
visual way through the peak in the graph—argument A1, correspondence code.
The content—or consequent—of this first semiotic function constitutes,
according to Juliette’s practice, a partial answer to item (a) “...the function is not
differentiable...”
The second partial answer to item (a) reflects the establishment of two
semiotic functions. The first of them accounts for a manipulation of the symbolic
representation of the function, that allows moving from the antecedent LE1 to the
consequent LE5 by means of the implicit definition of absolute-value—CD4,
correspondence code. The second one, allows us perceiving the symbolic-
algorithmic nature of Juliette’s second partial answer because, from the way she
defines the absolute-value function—CD4, antecedent, and by procedure P2—
correspondence code, she achieves the proposition PP3—consequent—that
constitutes her partial answer: “...the function is differentiable in the whole
domain...”.
With the two partial solutions that Juliette provides as answers to item (a), it
becomes evident that there is an apparent disconnection between the graphic-
visual interpretation and treatment that she performs in order to give her first
partial answer, and the symbolic-algorithmic manipulation and interpretation that
leads her to give her second partial answer.
In the same way, for item (b) Juliette provides two partial solutions, one of
graphic-visual nature and another one of symbolic-algorithmic nature. In the first
partial solution, the establishment of a semiotic function can be identified. Here,
the antecedent is the proposition PP2 given in her solution to item (a). Thus,
since Juliette considers that the function is not differentiable at any point of the
domain—because there is a peak in the graph—argument A1, correspondence
code—then “... f ¢(2) can not be calculated...”—PP4, consequent.
In the second partial solution to item (b), the establishment of two
concatenated semiotic functions can be identified. For the first one, Juliette starts
in the way she defines the absolute-value function—CD4, antecedent), and
through a procedure of deriving the positive part of the function
f ( x) = x, "x, x ³ 0 —P2, correspondence code, she concludes that the derivative
at the point of domain is 1, which she represents with LE7—consequent. For the
second semiotic function, LE7 is established as antecedent, and through the

PNA 11(2)
L. Pino-Fan et al. 116

proposition PP5—correspondence code, provides the graphic representation—

LE8, consequent—of LE7.
Finally, for item (c) it is possible to identify two concatenated semiotic
functions. For the first one, Juliette once again starts from the definition CD4—
antecedent, and through the procedure P2—correspondence code, of deriving
f ( x) = x since 0 ³ 0 , she concludes that the derivative at the point of domain is
1, which she represents with LE7—consequent. The second semiotic function
has the LE7 as antecedent and LE9 as consequent.
In general, it is possible that Juliette does not make a connection between the
graphic and symbolic representations of the function, even though such
connection is established in the formulation of the task by means of the a priori,
or institutional, semiotic function in which mathematical objects that had been
previously enunciated in Tables 1, 2 and 3 are mobilized. This a priori semiotic
function connects symbolic—LE1, antecedent—and graphic—LE2,
consequent—representations of the function, through a proposition PP1—
correspondence code—imposed as if it was a norm.

COMPARISON OF THE ANALYSIS

As seen in the previous section, the analysis conducted with both TRSR and
OSA, show deficiencies in Juliette’s mathematical activity, related to the lack of
connection of the interpretations and treatments that she makes in the graphic and
symbolic representations of the absolute-value function. However, there are some
qualifications regarding the analysis carried out and the results obtained that are
intrinsic to the principles and notions of each theoretical perspective. The Table 6
presents in detail the general methodologyof analysis used by both theoretical
perspectives.

Table 6
Methodology for Cognitive Analysis According to TRSR and OSA
Theory of Register of Semiotic
Representation (TRSR) Onto-Semiotic Approach (OSA)
Identification of the registers involved in The proposed task is decomposed into
the formulation of the task. basic units of analysis, and based on this
decomposition, primary mathematical
objects, their meanings and previous
processes are identified. Semiotic
functions established a priori are
identified, if there are any.

PNA 11(2)
Analysis of the Underlying Cognitive… 117

Table 6
Methodology for Cognitive Analysis According to TRSR and OSA
Theory of Register of Semiotic
Representation (TRSR) Onto-Semiotic Approach (OSA)
It focuses on the analysis of the It focuses the analysis on the subject’s
mathematical activity of the individual, mathematical practice.
focusing on the registers that are at play
when solving a task.
Identification of significant units within Describing the subject’s mathematical
the representations put into play in each practice in general terms, identifying
register. basic units of analysis.
Identification and characterization of the Identifying and describing the
subject’s cognitive operations and configuration—or sub configurations—
characterization of the cognitive of primary mathematical objects—
operations of treatments and linguistic elements, concepts,
conversions/passages in the subject’s propositions, procedures and arguments,
activities. processes and their meanings, involved
In the treatments, specify the type of in the subject’s mathematical practice.
register utilized from the subject’s Identifying and reconstructing the
production, determining: if it is a visual semiotic functions established by the
recognition, a calculation, or a subject, which provide meaning to the
utilization of a discursive or algebraic mathematical objects mobilized during
definition. his/her practice and give sense to the
In the conversions, specify the sense of elements that form the configuration
these in terms of congruence or non- described in the previous step.
congruence of the representations at
stake. In this stage, the pair of registers
mobilized and the sense of the
conversion should be indicated.
Furthermore, from a mathematical point
of view, it should be pointed out if the
conversion is explicit or implicit.
Studying conversions in terms of Studying coherence, from a
congruence—or incongruence—of the mathematical point of view, regarding
articulation—or no articulation—that a the connection—or lack of—that a
subject performs between the different subject makes between diverse sub-
registers of representation involved. configurations and, in consequence,
between the different semiotic functions
that the subject establishes within
his/her practice.

PNA 11(2)
L. Pino-Fan et al. 118

A first key aspect that is important to distinguish is that, while the analysis from
the OSA perspective focused on the subjects’ mathematical practices, and
mathematical objects, processes and their meanings, that emerge from such
practices, the TRSR focused its analysis primarily on the registers of
representation that the subject mobilizes in his/her productions. In this way, the
methodology proposed by TRSR can be considered as more “global”, in the
sense that the subjects’ cognitive activity is analyzed without performing
valuations from a mathematical point of view, as it is done with the tools of
OSA. For example, with TRSR, it is possible to draw conclusions such as there
are deficiencies in the subject’s knowledge because he/she does not carry out
conversion, passage or treatment. However, contradictions can indeed be
detected in the subject’s cognitive activity and whether the task was successfully
solved or not can also be determined. On the other hand, with OSA it is possible
to provide a detailed explanation, from a mathematical point of view, of what are
the knowledge deficiencies the subject has in relation to a certain problem, by
identifying and describing in a detailed way, the mathematical objects, processes
and their meanings, involved in his/her mathematical practice.
Yet, it must be pointed out that in OSA there is not systematization for the
analysis of linguistic elements. As a part of the methodology proposed by OSA,
language signs—linguistic elements—can be identified, which the subject uses to
express cognitive activity. However, whether such linguistic elements make
reference to representations of objects in a same register or in different ones is
not made explicit in the analysis; in other words, the linguistic elements of OSA
would correspond to representations of objects in different registers. In this way,
a linguistic element encompasses different languages—verbal, figurative,
symbolic... These different languages could make reference both to registers of
semiotic representation and semiotic systems.
In this regard, TRSR makes a clear distinction between register of semiotic
representation and semiotic system. As discussed before, a register is a semiotic
system that fulfills three conditions: (a) traces or significant units of
representation that belong to a system must be clearly identifiable and should not
allow contradictions; (b) it allows internal treatments in each register; and (c) it
allows passages or conversions from one register to another. Thus, we can
observe how the notion of register of semiotic representation of TRSR,
complements and enriches the notion of linguistic elements of OSA, by making a
very clear distinction between register and semiotic system, and systematizing
the analysis of such registers. Having said that, it is quite undeniable that in the
methodologies of analysis proposed by both theoretical approaches, it is agreed
that language is the most relevant semiotic system of them all. In this sense, we
could add that language is a multifunctional system. Another example of
multifunctional system is the figural register, since geometries—flat,
spherical…—do not constitute a register, but a mathematical framework—or a

PNA 11(2)
Analysis of the Underlying Cognitive… 119

framework for other disciplines, in which several registers can work. Therefore,
it is important not to confuse register with framework.
Despite the systematization that TRSR provides for the analysis of registers
of representation, by means of the study of treatments and conversions, some
relevant aspect must be taken into account for the improvement of cognitive
analysis and provide more detailed explanations, from a mathematical point of
view, about the subject’s cognitive activity to which we have access to through
his/her mathematical practice. As mentioned above, since it is a global theory
and it focused on a subject’s cognitive activity—and in his/her ability to mobilize
diverse registers of representation, the TRSR focuses its attention on the study of
registers of representation mobilized by such subject, leaving the content and the
meaning of the mathematical objects being represented implicit. For example, if
we consider the answer given by Juliette to item (c), with the TRSR methodology
of analysis it can be concludedthat she performs a treatment in the symbolic
register that allows passing from the expression f ( x) = ì x, x ³ 0
í- x, x < 0 to the
î
expression f ¢( x) = 1, then she performs a direct passage that allows her to move
from the latter expression—given in the symbolic register—to the graph of a
point on the cartesian plane—graphic register. In general, we can observe how
often conversions/passages always have as starting point and as point of arrival a
register of representation, that could as well be the same—in the case of
treatments—or different—in the case of conversions/passages. But, how does
Juliette’s answer relate to the way she conceives the definition of absolute-value
function? What explanation can be provided to the fact that Juliette had
expressed the derivative of the absolute-value function at x = 0 as f ¢( x) = 1
instead of f ¢(0) = 1 —although we know that such answer is mathematically
incorrect?
The notion of semiotic function, introduced by OSA, allows giving a better
explanation about what is Juliette’s cognitive activity to give her answer and
allows responding to the two previous questions. As evidenced with the analysis
performed with the OSA perspective, Juliette establishes two semiotic functions
as part of her answer to item (c). As described in the previous section, the first
semiotic function is established from the way Juliette defines the absolute-value
function (antecedent), considering x = 0 as part of the domain of the line segment
f ( x) = x . Afterwards, through the implicit procedure of derivation of f ( x) = x
—correspondence rule, she finally obtains f ¢(0) = 1—consequent. It is possible
to observe that, in order to establish this first semiotic function, the meaning that
Juliette gives to the mathematical object definition of absolute-value function is
emphasized, leaving the representation of such mathematical object implicit. The
semiotic function that we have just discussed has a definition as antecedent, an
implicit procedure as correspondence code, and a representation as consequent—
that lies in the same register than the antecedent. Something similar can be said

PNA 11(2)
L. Pino-Fan et al. 120

about the second semiotic function, but the difference is that it has a
representation in the symbolic register as antecedent, a proposition in the verbal
register as correspondence code, and a representation in the graphic register as
consequent. Thus, we can notice how the semiotic function of OSA, reinterprets
and enriches the notions of treatments and conversions/passages of TRSR, by
emphasizing the meaning—personal or institutional—given to the mathematical
content of the objects represented and contemplating that both antecedent and
consequent of the function can be languages—representations,
concepts/definitions, properties/propositions, procedures or arguments.
Finally, it is important to point out that similar analyses can be carried out for
the expected solutions—from an expert or institutional point of view—from the
two theoretical perspectives—cognitive analysis of the answer of the expert in
TRSR and epistemic analysis in OSA—analysis of expected institutional
meanings. However, for length reasons and because it would be redundant with
what has already been described in this document, we do not present such a priori
analysis in this document. In any case, the last stage in the methodological
approach of both perspectives (Table 6) makes reference to the study of the
proximity of the subject’s knowledge in relation to the knowledge that was
expected to be mobilized in his/her practice.

FINAL REFLECTIONS
In this paper, we discussed the relation between representations and the
underlying mathematical activity during the development of a task—about
derivative. In order to examine such a relation, it is necessary to rely on models
which analyze students’ mathematical activity. In the literature, most models are
cognitive-based. They consider that learning a mathematical concept and its
application occurs if various appropriate internal representations are developed
and integrated together with functional relations among them. Other few models
are semiotic and pragmatic-based. The pragmatic approach also gives importance
to the use of diverse representations; however, different reasons are provided in
comparison to those from the cognitive approach. On the one hand, the cognitive
approach primarily explores representations from a representational perspective.
On the other hand, the pragmatic approach emphasizes the instrumental
dimension, namely, what can be done with a representation.
We investigate this relation, first, from a cognitive perspective (TRSR) and
second, from a pragmatic and semiotic perspective (OSA). The notions of
semiotic system and register of semiotic representation of the TRSR are essential
for the comprehension of the cognitive activity needed to solve a task, while that
OSA provides a level of analysis of the subject’s cognitive activity that shows
mathematical objects that are involved in the processes of treatment and
conversion/passages between registers of semiotic representation. This level of

PNA 11(2)
Analysis of the Underlying Cognitive… 121

analysis complements the analysis carried out using the tools of TRSR, because
with the tools of configuration of objects and processes and semiotic function,
the contents of representations become explicit and are used as part of such
cognitive activity. The registers of representation are implicitly involved in
semiotic functions; however, these emphasize the mathematical content of the
representation. The relationship between notions of mathematical objects and its
meanings—as considered in OSA—and semiotic representation—as considered
in TRSR, are essential for the analysis and characterization of mathematical
knowledge.
The results of the comparison of analysis presented in comparison of the
analysis section show that between these two theoretical perspectives there are
complementarities that would allow performing more precise and finer cognitive
analysis, from the subjects’ production. As a result of these complementarities,
we thought it was pertinent to propose the following methodology to develop
cognitive analysis, integrating the contributions of both perspectives:
¨ To center the analysis on subjects’ production, since it is through it that
we can get closer to their cognitive activity.
¨ To decompose the subject’s production in significant units of analysis.
¨ To identify in each significant unit, the language signs that the subject
utilizes to express his/her knowledge. It is important to specify if these
signs make reference to a semiotic system or to a register of
representation, and also, to which type does it refer to.
¨ To study and describe if the representations—language signs—that the
subject utilizes to show his/her knowledge make reference to
concepts/definitions, properties/propositions, procedures or arguments—
identification of configurations or sub configurations of objects and
processes. The meaning that the subject gives to each of such
mathematical objects (concepts, properties…) must be described.
¨ To identify and reconstruct the semiotic functions that the subject
establishes as part of his/her practice. Determine if such semiotic
functions make reference to cognitive operations of treatments or
conversions/passages.
¨ To study coherence, from a mathematical point of view, in relation to the
connections—or lack of—that the subject makes among the diverse sub-
configurations and, in consequence, among the diverse semiotic functions
that he/she establishes during his/her practice. Similarly, to study the
semiotic functions that make reference to the conversions in terms of
congruence—or non-congruence—of the articulation—or no
articulation—that a subject performs among the different registers of
representation involved.
The above-mentioned methodology can also be utilized for the performance of
respective a priori analysis, and which will unveil expected mathematical

PNA 11(2)
L. Pino-Fan et al. 122

knowledge, which is indispensable to determine the proximity of the knowledge

shown by subjects—based on the latter point of the previously mentioned
methodology—in relation to the expected knowledge. Besides, the methodology
is flexible since it can be adapted both to the productions of a subject regarding
diverse types of tasks, and the mathematical objects that such tasks demand.
Thus, the combination of the two frameworks contributes to literature by
extending our understanding of the relationship between representations and the
underlying mathematical activity during the development of tasks that provide
better explanations about the aspects that make it possible or impossible to
comprehend mathematical notions.

ACKNOWLEDGEMENTS
This study has been performed in the framework of research projects
FONDECYT Nº 11150014 (Funded by CONICYT, Chile), and EDU2015-
64646-P (MINECO, Spain).

REFERENCES
Artigue, M., Bartolini-Bussi, M., Dreyfus, T., Gray, E., & Prediger, S. (2005).
Different theoretical perspectives and approaches in research in mathematics
education. In M. Bosch (Ed.), Proceedings of the fourth Congress of the
European Society for Research in Mathematics Education (pp. 1239-1243).
San Feliu de Guixols, Spain: CERME.
Asiala, M., Brown, A., DeVries, D., Dubinsky, E., Mathews, D., & Thomas, K.
(1996). A framework for research and curriculum development in
undergraduate mathematics education. In J. Kaput, A. H. Schoenfeld, & E.
Dubinsky (Eds.), Research in collegiate mathematics education II (pp. 1-32).
Providence, RI: American Mathematical Society.
Bikner-Ahsbahs, A., & Prediger, S. (2010). Networking of theories¾an
approach for exploiting the diversity of theoretical approaches. In B. Sriraman
& L. English (Eds.), Theories of mathematics education: Seeking new
frontiers (pp.483-506). New York, NY: Springer.
Duval, R. (1995). Sémiosis et pensée: registres sémiotiques et apprentissages
intellectuels [Semiosis and human thought: Semiotic registers and intellectual
learning]. Berne, Switzerland: Peter Lang.
Duval, R. (2006). A cognitive analysis of problems of comprehension in a
learning of mathematics. Educational Studies in Mathematics, 61(1-2), 103-
131.
Duval, R. (2008). Eight problems for a semiotic approach in mathematics
education. In L. Radford, G. Schubring, & F. Seeger (Eds.), Semiotics in

PNA 11(2)
Analysis of the Underlying Cognitive… 123

Mathematics Education: Epistemology, history, classroom and culture (pp.

39-61). Rotterdam, The Netherlands: Sense Publishers.
Eco, U. (1976). Tratado de semiótica general [A theory of semiotics]. Barcelona,
Spain: Lumen.
Font, V., Godino, J. D., & Gallardo, J. (2013). The emergence of objects from
mathematical practices. Educational Studies in Mathematics, 82(1), 97-124.
Font, V., Trigueros, M., Badillo, E., & Rubio, N. (2016). Mathematical objects
through the lens of two different theoretical perspectives: APOS and OSA.
Educational Studies in Mathematics, 91(1), 107-122.
Godino, J. D., & Batanero, C. (1994). Significado institucional y personal de los
objetos matemáticos [Institutional and personal meanings of mathematical
objects]. Recherches en Didactique des Mathématiques, 14(3), 325-355.
Godino, J. D., Batanero, C., & Font, V. (2007). The onto-semiotic approach to
research in mathematics education. ZDM-The International Journal on
Mathematics Education, 39(1), 127-135.
Godino, J. D., Font, V., Wilhelmi, M., & Lurduy, O. (2011). Why is the learning
of elementary arithmetic concepts difficult? Semiotic tools for understanding
the nature of mathematical objects. Educational Studies in Mathematics,
77(2), 247-265.
Hjelmslev, L. (1943). Prolegomena to a theory of language. Madison, WI:
University of Wisconsin Press.
Peirce, C. S. (1978). The collected papers of Charles Sanders Peirce. Cambridge,
MA: The Belknap Press of Harvard University Press.
Pino-Fan, L. (2014). Evaluación de la faceta epistémica del conocimiento
didáctico-matemático de futuros profesores de bachillerato sobre la derivada
[Evaluation of epistemic facet of didactic-mathematical knowledge of future
high school teachers on the derivative function] (Unpublished doctoral thesis).
University of Granada, Spain.
Prediger, S., Bikner-Ahsbahs, A., & Arzarello, F. (2008). Networking strategies
and methods for connection theoretical approaches: First steps towards a
conceptual framework. ZDM, 40(2), 165-178.
Radford, L. (2008). Connecting theories in mathematics education: Challenges
and possibilities. ZDM, 40(2), 317-327.
Tsamir, P., Rasslan, S., & Dreyfus, T. (2006). Prospective teachers’ reactions to
right-or-wrong tasks: The case of derivatives of absolute value functions.
Journal of Mathematical Behavior, 25(3), 240-251.

PNA 11(2)
L. Pino-Fan et al. 124

Luis R. Pino-Fan Ismenia Guzmán Retamal

University of Los Lagos University of Los Lagos
[email protected] [email protected]
Vicenç Font Moll Raymond Duval
University of Barcelona University of Lille
[email protected] [email protected]

Received: December 2015. Accepted: April 2016.

Handle: http://hdl.handle.net/10481/44148

PNA 11(2)