Publishing Partner: Cambridge University Press CUP Extra Publisher Login

New from Oxford University Press!


Style, Mediation, and Change

Edited by Janus Mortensen, Nikolas Coupland, and Jacob Thogersen

Style, Mediation, and Change "Offers a coherent view of style as a unifying concept for the sociolinguistics of talking media."

New from Cambridge University Press!


Intonation and Prosodic Structure

By Caroline Féry

Intonation and Prosodic Structure "provides a state-of-the-art survey of intonation and prosodic structure."

The LINGUIST List is dedicated to providing information on language and language analysis, and to providing the discipline of linguistics with the infrastructure necessary to function in the digital world. LINGUIST is a free resource, run by linguistics students and faculty, and supported by your donations. Please support LINGUIST List during the 2017 Fund Drive.

Review of  Cognitive Semantics and Scientific Knowledge

Reviewer: Vanja Kljajevic
Book Title: Cognitive Semantics and Scientific Knowledge
Book Author: András Kertész
Publisher: John Benjamins
Linguistic Field(s): Semantics
Cognitive Science
Issue Number: 16.3073

Discuss this Review
Help on Posting

Date: Fri, 21 Oct 2005 12:09:20 -0700 (PDT)
From: Vanja Kljajevic
Subject: Cognitive Semantics and Scientific Knowledge

AUTHOR: Kertész, András
TITLE: Cognitive Semantics and Scientific Knowledge
SUBTITLE: Case studies in the cognitive science of science
SERIES: Converging Evidence in Language and Communication Research 4
PUBLISHER: John Benjamins
YEAR: 2004

Vanja Kljajevic, Institute of Cognitive Science, Carleton University,
Ottawa, Canada

András Kertész's book "Cognitive Semantics and Scientific Knowledge" is an
attempt to answer the question "How should we analyze scientific
knowledge?". Although the book is organized in a way that brings cognitive
semantics in focus, it actually addresses the problem that neither
traditional analytic philosophy nor more recently developed naturalized
philosophy of science based on behavioral psychology can explain some
basic questions pertaining to scientific knowledge. As a solution, Kertész
proposes to develop 'cognitive science of science' -- i.e. a version of
the naturalized philosophy of science which would be based on linguistics
as one of the disciplines that constitute cognitive science. More
specifically, Kertész chooses to apply the methods of cognitive semantics,
i.e. the cognitive theory of metaphor and the two-level approach to
meaning to the domain of scientific knowledge. This choice is based on an
assumption that not only psychology, but other disciplines contributing to
cognitive science (e.g. linguistics) as well as their sub-disciplines
(cognitive semantics) can borrow their methodology to philosophy of

By narrowing down a more general question "What the case would be if the
methods of cognitive semantics were applied to scientific knowledge?" to a
few specific hypotheses and research questions carefully formulated in
terms of the scope of the two cognitive semantics theories, Kertész
addresses this main question at different levels in each of the four parts
of the book. In the first part, 'Preliminaries', we find that Kertész's
approach was inspired by cognitivism on one side, and the naturalized
philosophy, on the other. Cognitivism postulates that certain behaviors
(e.g. language) can be explained only by appeal to 'internal cognitive
processes'. These internal processes consist of mental representations of
information and a finite set of rules that operate on them. The mind's
capacity to automatically receive, store, manipulate, and output
INFORMATION depends on its receiving, storing, manipulating and outputting
REPRESENTATIONS of that information (Von Eckardt 1995).

Note that cognitivism is still a theory of COGNITIVE states and processes,
yet to learn how to tackle phenomena such as understanding, sensations,
emotions, moods, skills, consciousness, or creativity. The concept of
understanding is particularly elusive, because it exhibits different
paradigms, such as 'insightful understanding' in science (Haugeland 1998),
which differs from our understanding of the world we live in, which in
turn differs from our capacity to understand language. Since the purpose
of scientific endeavor is to understand phenomena, it is interesting to
observe the efforts to determine to what extent cognitivism, which cannot
fully grasp the process of understanding, can contribute to our
understanding of the underlying mechanisms of understanding of scientific

Kertész believes that cognitivism, combined with the naturalized
philosophy of science, has the potential to solve the problems of
theoretical terms, the problem of the criterion of truth, and Zeno
paradox, among others. The major limit of the traditional, analytic
philosophy of science was its concern to JUSTIFY its object of inquiry,
based on the principles of rationality and a priori assumptions. The
naturalized philosophy of science, an analogue of the naturalized
epistemology, aims to DESCRIBE and EXPLAIN scientific knowledge by using
empirical, a posteriori evidence. While NATURALISM for Quine
means 'behaviourism', SCIENCE is a synonym for natural sciences, and
SCIENTIFIC KNOWLEDGE should be explained in terms of behavioral
psychology, Kertész opts for a weak version of naturalism and a less rigid
concept of science: explanations, after all, are not all alike, and
cognitive science has already resulted in empirically legitimate (i.e.
testable) explanations of phenomena.

In the second part of the book, 'Prospects: Theoretical Terms', Kertész
tests the hypothesis that the cognitive theory of metaphor and the two-
level approach to meaning can solve the problem of theoretical terms
(Chapter 3). The author asks 'what the case would be' (p. 65) if the two
metascientific extensions of cognitive semantics, the cognitive theory of
metaphor (Chapter 4) and the two-level approach to meaning (Chapter 5),
were involved in solving the problem of theoretical terms, which are
introduced to build theories, explain observable facts, and provide
understanding of phenomena. However, theoretical terms are only
hypothetical constructs, which do not refer to 'observable reality'. Yet,
it is precisely the theoretical terms that, via the method of explication,
establish the relationship between a theory and that reality. The early
logical empiricism, which was focused on the semantic (referential) aspect
of theoretical terms, failed to capture the dichotomy. A more plausible
approach, Kertész suggests, would perhaps attempt a pragmatic explanation,
focusing on the context-dependency of theoretical terms, with a specific
scientific theory functioning as a context.

Following the current trend to reconstruct the problem of theoretical
terms 'as an essentially linguistic issue' (p. 57), Kertész analyzes the
terms of the Standard Theory (ST) and the Government and Binding Theory
(GB). The author conducts the analysis in three steps: first he analyzes
the structure of theoretical terms, then the structure of scientific
explanations, and finally the relationship between theoretical terms and
the context (pragmatics). According to the holistic approach, the
theoretical terms such as for example 'tree', 'node', 'cycle',
or 'daughter', are part of metaphorical expressions, which in turn are
based on metaphorical concepts. The explanations are structured in such a
way that the explication process focuses on the relation holding between
the theoretical term and the underlying metaphorical concept. Finally, the
pragmatic aspects of theoretical terms are already part of metaphorical
concepts and scenarios in which they are employed.

On the other side, the modular approach postulates that the interpretation
of a lexical item (or a theoretical term) in a certain context is the
mapping of its semantic representation onto the level of its conceptual
structure (hence the name 'two-level approach'). The relationship between
the conceptual and the semantic representations entails that theoretical
terms, like other lexical items, are semantically underdetermined.
However, several conceptual principles regulate interpretation of a
lexical item in a certain context (the principle of conceptual shift,
conceptual specification and conceptual selection), enabling a correct
mapping of semantic representations onto the conceptual representations
required by the context. Thus, unlike the traditional analytic philosophy,
which is interested in the relationship between the EXPLICANS and
EXPLICANDUM, the cognitive science of science focuses on the CONCEPTUAL
aspects of theoretical terms.

Part III of the book, 'Prospects: Sociological Extensions', explores some
elements of the sociological perspective on scientific knowledge (Chapter
7), proposing sociological extensions of the modular (Chapter 8) and the
holistic approach (Chapter 9), and testing whether the two extensions can
capture the interactions between the conceptual and social factors that
shape scientific knowledge. The sociological extension of the modular
approach is a case study aiming to show that 'theoretical contents' of ST
and GB are based on interaction of the conceptual and the motivational
modules. These modules, the argument goes, have universal principles and

For example, a universal principle of the conceptual module is the
relevance principle, adapted from Sperber and Wilson's relevance theory,
stating: "Every explanation is put forward in a context ct under the
assumption of its maximal relevance with respect to ct." While this
principle, among many other pragmatic principles, can easily be applied to
scientific theories, that is not the case with the universal principle of
the so-called motivational module, which is probably 'the weakest link' of
the sociological extension. It states that the social interests select 'I-
properties' of scientific theories. In other words, the social interests
determine which conceptual properties of an object will become elements of
its conceptual representation. As an example, Kertész comments on the
social interests that governed the development of ST and GB, i.e.
systematic description and systematic restriction of grammar,
respectively. These social interests are reflected in the motivational
principles underlying the two theories (the principle of systematicity /
simplicity and the principle of restrictiveness), with the motivational
element functioning as a part of the conceptual structures in these

The sociological extension of the holistic approach is a case study about
the relationship between the conceptual and social factors in the theories
of AIDS, emphasizing the role of conversational analysis in these theories
and suggesting that metascientific reflexion of the cognitive theory of
metaphor could contribute to object-scientific research (e.g. conscious
use of metaphors).

Chapter 10 is an attempt to justify the application of the two cognitive
semantic theories to scientific knowledge with regard to the following:
double-facedness of scientific knowledge, its uniqueness, the explication
of the term 'theory', the self-application and the pragmatics of
scientific discourse. In brief, unlike the analytic philosophy of science,
which cannot explain the relationship between these universal principles
and the principles of specific scientific theories, Kertész claims that
the two-level approach explains this relationship in terms of modularity,
and the cognitive theory of metaphor does that in terms of conceptual
structure. Unlike the analytic philosophy of science, which assumes that
scientific knowledge is unique, both cognitive semantic theories integrate
this knowledge into a 'wider context of human behavior' (p. 135, 139).
Further, the author claims, the two theories offer a new approach to the
term 'theory' by considering scientific theory as a set of representations
(modularism) and a conceptual framework within which different conceptual
domains interact (holism). Finally, both approaches are self-applicable
(p. 136, 141) and both approaches are compatible with the idea that
scientific research / knowledge should include not only logical,
syntactic, semantic, but also 'communicative, textual and social factors'

Part IV of the book, 'Limits', explains why the two cognitive semantic
theories fall short of their goal when applied to scientific knowledge. By
addressing the problem of normativity of metascientific reflexion, as well
as drawbacks of plausible reasoning and fallacies in the two approaches
(Chapter 11), circularity in them (Chapter 12), and their inability to
solve, without first empirically reformulating, problems such as the
criterion of truth or Zeno's paradox (Chapter 13), Kertész concludes the
book with an emphasis on the limits of the two cognitive semantic theories
(Chapter 14), pointing out to further questions for the cognitive science
of science (Chapter 15).


The idea of widening the scope both of the naturalized philosophy of
science and of cognitive semantics deserves credit. However, Kertész shows
that this goal cannot be achieved by accepting the main hypotheses of the
book. The hypotheses are not only overgeneralized, as pointed by the
author, they are also inconsistent at several levels. For example, it is
not always clear whether Kertész talks about Fodorian or Chomskian or the
modularity of the two-level approach, although they differ in important
ways. Also, throughout the book Kertész develops case studies as thought
experiments, with terminology used in pre-explicative sense. If we recall
that the basic concepts of the theories whose methodology the author
adopts in search for the cognitive science of science also lack precise
definitions (e.g. concept, conceptual domain, metaphorical representation,
etc.), then it becomes clear that the naturalized philosophy of science
based on the two semantic theories lacks the potential to coherently and
consistently EXPLAIN phenomena, evolving into a better alternative to the
analytic philosophy of science.

This informationally dense book requires a patient reader. Chapters are
structured around the main hypotheses and research questions, but the book
also provides ample background information, sometimes incorporated into
the text, but more often presented as notes and quotations. Those who
write about interdisciplinary topics know how difficult it is to
appropriately incorporate background information and will perhaps tolerate
text's fragmentation. Other readers may feel occasionally frustrated by a
half-page or longer quotations that very often are not necessary.

Finally, Kertész's stand on the concept of cognitive architecture is not
clear. This is problematic, given that explanations of the relationship
between the scientific knowledge and knowledge of language should stem
from the conception of the functional architecture within which the two
entities relate in a clear way. Kertész's proposal stems from relating
these two entities methodologically, not theoretically per se.
Methodological choices can affect theory in unwanted ways. In this case,
application of linguistics methodology to scientific knowledge has
theoretical consequences: an incorrect assumption that the two types of
knowledge belong to the same type of knowledge. However, that is not the
case. Note that LANGUAGE, the object of linguistics and semantics, can be
studied as either KNOWLEDGE of language (competence) or as USE of that
knowledge (performance). Scientific knowledge, on the other hand, is
always explicit. By choosing to apply the methods of two cognitive
semantic theories to scientific knowledge, Kertész dismisses the
distinction between the tacit or implicit knowledge, such as knowledge of
language, and explicit knowledge, such as scientific knowledge.

Also, there is growing neuropsychological evidence indicating that the two
types of knowledge employ different cognitive resources. While
deterioration of episodic memory is one of the first symptoms of
Alzheimer's dementia, followed by gradual deterioration of semantic
knowledge, semantic dementia wipes out a person's semantic knowledge (i.e.
knowledge about the world and the mental lexicon), leaving out only those
aspects that are tightly related to everyday experience. On the other
side, aphasic syndromes affect language in different ways, without in
principle affecting either episodic or semantic memory.

Thus, knowledge of language and scientific knowledge are parallel neither
methodologically nor theoretically. Why, then, use the methods of
linguistics to explain the nature of scientific knowledge? While it is
certainly interesting to follow the author's reasoning on 'what the case
would be' if one applied modularism of a certain type or semantic holism
in analysis of scientific knowledge, the book could have benefited from a
clearer picture on the author's understanding of the relationship between
language and scientific knowledge.

This brings out the main objection to this book. It pertains to the
author's assumption that since linguistics is a contributing discipline of
cognitive science, and cognitive semantics is part of linguistics,
cognitive semantics is also part of cognitive science. Based on this
assumption Kertész proposes employing the methods of cognitive semantics
in analysis of scientific knowledge (object of philosophy of science,
which is also 'part' of cognitive science). This reflects the so
called 'weak', 'less ambitious' concept of cognitive science (Gardner
1985) according to which cognitive science is just a 'collaboration' among
the following disciplines: neuroscience, psychology, linguistics,
artificial intelligence and philosophy. Although this concept is present
in most dictionaries and textbooks on cognitive science, it does not
correctly reflect the state of affairs. Cognitive science is a scientific
paradigm (Kuhn 1962), based on specific theoretical and methodological
assumptions (Von Eckardt 1995, Encyclopaedia of Cognitive Science 2003),
which guide research towards coherent and consistent explanations of
phenomena. In other words, cognitive science is not the sum of the five
disciplines, but rather their crossing point. This means that, contrary to
the 'weak' view, not all the content of the contributing disciplines
necessarily constitutes cognitive science.

Does it mean, then, that it is too early to talk about the cognitive
science of science? The answer to this question is perhaps in the 'strong'
version of cognitive science as a basis for the naturalized philosophy of
science. That means acceptance of cognitive science as a new paradigm,
which is emerging from the crossing of the five disciplines. In other
words, instead of 'borrowing' the methods from the contributing sub-
disciplines of cognitive science, the cognitive science of science would
explain the underlying mechanisms of scientific knowledge by a theory that
could consistently explain conceptualization at all levels, starting from
the word-level and text, and capturing the mapping processes between the
levels (conceptual vs. semantic), across the domains (source vs. target),
and between the elements of cognition (perception vs. conception
interactivity). Developing such a theory, which could be done within the
context of Jackendoff's representational modularity, would perhaps bail
out the idea of cognitive science of science.


Bierwisch, M. "How Much Space Gets into Language?" Ms.

Bierwisch, M. & Schreuder, R. 1992. From Concepts to Lexical
Items. "Cognition" 42, pp. 23-60.

"Encyclopaedia of Cognitive Science". 2003. Edited by Lynn Nadel. London,
England: Nature Pub. Group.

Gardner, H. 1985. "The Mind's New Science". New York: Basic Books.

Kuhn, T. 1970. "The Structure of Scientific Theories". Chicago: University
of Chicago Press.

Lakoff, G. & Johnson, M. 1980. Metaphors We Live By. Chicago: The
University of Chicago Press.

Von Eckardt, B. 1995. "What Is Cognitive Science?" Cambridge, MA: The MIT


Vanja Kljajevic received her doctorate in Cognitive Science from Carleton
University, Ottawa. Her research interests include language disorders,
processing of complex syntax, syntactic working memory, and philosophy of
psychology. She is currently involved in projects related to human-
computer interaction and verbal communication of emotions.