Chances are that you might have
heard or read something lately about evolutionary psychology. This still very
young discipline has given rise to a booming research industry which regularly
makes news headlines even in the non-scientific press. The goal of evolutionary
psychology, as with any brand of psychology, is to discover and understand the
design of the human mind. But fuse psychology with evolution as evolutionary
psychologists do, and you will have a major controversy at your hand. But why
is this?

According to evolutionary
psychologists, taking evolutionary theory seriously will lead to a major
reframing of how we should think about our human mind. Whereas traditional
cognitive scientists assume that the human mind is mainly driven by generalized
cognitive abilities which are basically jack-of-all-trades and applicable to
all knowledge domains, evolutionary psychologists on the contrary assume that
our minds are largely composed of vast arrays of cognitive modules, i.e.
cognitive capacities that are specialized and dedicated to solve very limited
problems and nothing else besides. A much better theoretical model to capture
the human mind than the general-problem solvers of early artificial
intelligence researchers therefore is the Swiss Army knife. However, contrary
to the Swiss Army knife, the human mind did not have an intelligent designer —
the structure of our mind is completely due to blind evolutionary forces.
Natural processes which took thousand of years of mutation and selection
crafted its highly complex design. This is why the popular evolutionary
psychologists’ saying has it that our anatomically modern bodies house a mind
well adapted by natural selection to an ancient environment (the Pleistocene).
No wonder that «traditional» cognitive scientists have raised criticisms
against such a profane view of the origins of our mind.

In Reconstructing Reason and
Representation, Murray Clark attempts to do two things. First, he wants to
clarify and evaluate the empirical and conceptual credentials of evolutionary
psychology.  Second, he wants to assess the implications of evolutionary
psychology for some issues in epistemology, philosophy of science, and
philosophy of mind (p. 1). As Clark correctly points out, this is not an easy
task given that evolutionary psychology is still in its infancy and that there
is no consensus among evolutionary psychologists themselves as to the central
theories or methods of their discipline. To circumvent many of these problems,
Clark decides to focus on the work of the discipline’s best-known advocates,
Leda Cosmides and John Tooby.

Clark’s basic theoretical postulate
is "that we ought to allow epistemology to go modular and view knowledge
as a set of natural kinds housed in a massively modular mind. That is,
knowledge is not a univocal concept to be clarified by a priori analysis but an
empirically discovered phenomenon, like water, to be elucidated using the
results of science and made consistent with other scientific results." (p.
1-2). He wishes to propose nothing more but also nothing less than a new
approach to naturalized epistemology updated by results derived from
evolutionary psychology. The list of topics Clark touches on his way is
impressive:

Chapter 1 attempts to clarify and
systematize ideas about the massive modular mind derived by Cosmides and Tooby’s
work. Chapter 2 takes up foundational worries that Jerry Fodor has recently
voiced against the whole massive modularity project, and tries to show moreover
that these are unfounded. In particular, Fodor has argued that evolutionary
psychology cannot accommodate global, abductive inference within the context of
local, computational processes as realized in cognitive modules. Clark wants to
show that Fodor has misrepresented evolutionary psychology and thus has argued
against a straw man all along. Chapter 3 attempts to work out an account of
misrepresentation based on a modular mind that is capable of solving the
disjunction problem of misrepresentation in accounts of naturalized
epistemology. Chapter 4 then attempts to show that reliable processes have been
selected for because of their indirect connection to true belief during the
Pleistocene period of our ancestral history; this argument is aimed
particularly at Steven Stich’s argument against reliable inferential systems.
In chapter 5, Clark then argues for a version of naturalized epistemology that
is based on reliabilist and externalist conceptions of epistemic justification
and knowledge, and defends it against general philosophical criticism against
the project of naturalization. Finally, in chapter 6 Clark directly moves to a
defense of his postulations that knowledge is a set of natural kinds, that it
should be studied empirically and not just by an a priori appeal to reason
alone.

The list is impressive, but is the
book up to its ambitious project? There is no question about the need of
detailed studies to assess the philosophic implications of evolutionary
psychology. Most studies to date have concentrated on several and detailed
aspects of evolutionary psychology, but very few investigations have
concentrated on drawing out the implications of this research framework in any
systematic way. Given the need of investigations to assess the implications of
evolutionary psychology, it is unfortunate that Reconstructing Reason and
Representation does not deliver what it promises. Many chapters in the book
fall short on their aims because the arguments developed therein are not as
effective as they should be against worries put up against evolutionary
psychology. I do not have the space here to go into details, but will highlight
just one shortcoming.

In chapter 2, Clark takes on Fodor’s
arguments against evolutionary psychology. In them Fodor argues that there is a
serious problem every modular mind will face, namely the so-called "input
problem": how does the modular mind know where to send incoming
information? How does the modular mind manage to assign specific information to
the proper cognitive module? There are, so it seems, only two possible ways to
do it. Either there is a single mechanism that takes representations at large
as its input domain and assigns some representations to one cognitive module
and others to other modules — in which case this mechanism would have to be
less modular than the cognitive modules it has to serve (but that option is
ruled out by someone advocating a massively modular mind). Or there are
specific mechanisms with distinct input information, one of which (M1) assigns
its input information to one cognitive module (C1) and the other of which (M2)
assigns its different input information to a different cognitive module (C2).
But this option would not help either since it faces a regress. The question
would come up again as to the nature of the mechanism assigning representations
to either (M1) or (M2): is it a single general mechanism or are there different
mechanisms, etc. Now Clark’s solution to the input problem simply reads, "If
there is a cheater detection module, then thought, conscious or not, is not
required, because the module is triggered automatically by the contextual cues
in the environment." (p. 25)

However, this will not do as a
solution since the input problem is precisely not about thought, conscious or
not. The problem is generated because there are no reliable perceptual cues
marking out each of the various specific domains to the corresponding proper
modules (such as cheater detection).  Fodor’s guess is that in the case of
cheater detection such assigning mechanisms will inevitably have to involve
judgments that draw on the whole resource of the mind (and thus they will be
general). The contextual cues of social exchange situations such as cheating
need intelligent interpretation; a kind of interpretation modules would not be
able to deliver based solely on their limited access to information within the
mind.

As the cited sentence makes clear,
Clark does not even address the central issues of Fodor’s input problem. Sadly,
many passages in the book show similar shortcomings in its arguments against
prominent skeptics of evolutionary psychology or naturalized epistemology
generally. However, this notwithstanding, Clark’s book has lots of merits and
is worthwhile to read. Reconstructing Reason and Representation is the
first book to systematically address philosophic implications of evolutionary
psychology by explicitly connecting issues in evolutionary psychology with
traditional problems in epistemology. There is a great deal to learn from its
discussion, and the book has definitely set the agenda for future work to come.
But it also shares some of the weaknesses of first attempts — future investigations
will have to take up the challenge and lead it to better results. 

© 2005 Harry Witzthum

Harry Witzthum,
Ph.D. did his doctorate in Philosophy at the University of Sheffield (UK). His
research interests comprise the philosophy of mind and psychology, philosophy
of language, and cognitive science.  He currently lives in Switzerland.

Categories: Philosophical, Psychology