(ebook english) Pinker,
[ Pobierz całość w formacie PDF ]
So How
Does
the Mind Work?
STEVEN PINKER
Abstract: In my book How the Mind Works, I defended the theory that the
human mind is a naturally selected system of organs of computation. Jerry Fodor claims
that ‘the mind doesn’t work that way’ (in a book with that title) because (1) Turing
Machines cannot duplicate humans’ ability to perform abduction (inference to the best
explanation); (2) though a massively modular system could succeed at abduction, such a
system is implausible on other grounds; and (3) evolution adds nothing to our under-
standing of the mind. In this review I show that these arguments are flawed. First, my
claim that the mind is a computational system is different from the claim Fodor attacks
(that the mind has the architecture of a Turing Machine); therefore the practical
limitations of Turing Machines are irrelevant. Second, Fodor identifies abduction
with the cumulative accomplishments of the scientific community over millennia.
This is very different from the accomplishments of human common sense, so the
supposed gap between human cognition and computational models may be illusory.
Third, my claim about biological specialization, as seen in organ systems, is distinct
from Fodor’s own notion of encapsulated modules, so the limitations of the latter are
irrelevant. Fourth, Fodor’s arguments dismissing of the relevance of evolution to
psychology are unsound.
In 2000 Jerry Fodor published a book called The Mind Doesn’t Work That Way
(hereafter: TMDWTW). The way that the mind doesn’t work, according to Fodor,
is the way that I said the mind does work in my book How the Mind Works
(HTMW).
1
This essay is a response to Fodor, and one might think its title might be
Yes, It Does! But for reasons that soon become clear, a more fitting title might be
No One Ever Said it Did.
Fodor calls the theory in How the Mind Works the New Synthesis. It combines
the key idea of the cognitive revolution of the 1950s and 1960s—that the mind
is a computational system—with the key idea of the new evolutionary biology of
the 1960s and 1970s—that signs of design in the natural world are products of
the natural selection of replicating entities, namely genes. This synthesis, some-
times known as evolutionary psychology, often incorporates a third idea, namely
that the mind is not a single entity but is composed of a number of faculties
specialized for solving different adaptive problems. In sum, the mind is a system
Supported by NIH grant HD 18381. I thank Clark Barrett, Arthur Charlesworth, Helena Cronin,
Dan Dennett, Rebecca Goldstein, and John Tooby for invaluable comments.
Address for Correspondence: Department of Psychology, William James Hall 970, Harvard
University, Cambridge MA 02138.
Email: pinker@wjh.harvard.edu
1
Fodor discusses HTMW together with a second book, Henry Plotkin’s Evolution in Mind
(Plotkin, 1997), which is similar in approach. But Fodor focuses on HTMW, as will I.
Mind & Language, Vol. 20 No. 1 February 2005, pp. 1–24.
#
BlackwellPublishingLtd. 2005,9600GarsingtonRoad,Oxford,OX42DQ,UKand350MainStreet,Malden,MA02148,USA.
2 S. Pinker
of organs of computation that enabled our ancestors to survive and reproduce in
the physical and social worlds in which our species spent most of its evolutionary
history.
Readers who are familiar with Fodor’s contributions to cognitive science but
who have not read TMDWTW might be puzzled to learn that Fodor begs to differ
so categorically. The first major theme of HTMW is computation, and Fodor,
more than anyone, has defended what he calls the computational theory of mind:
that thinking is a form of computation. The second major theme is specialization,
and Fodor’s most influential book is called The Modularity of Mind, a defense of
the idea that the mind is composed of distinct faculties rather than a single general-
purpose learning device or intelligent algorithm. The third theme is evolution,
the source of innate biological structure, and Fodor, like many evolutionary
psychologists, is willing to posit far more innate structure than is commonly
accepted in contemporary philosophy and psychology. So it is surprising that
Fodor insists that HTMW is wrong, wrong, wrong. Fodor and I must disagree
on how the concepts of computation, faculty psychology (specialization), and
innate biological organization should be applied to explaining the mind. This
essay will be organized accordingly.
The Concept of Computation in
How the Mind Works
According to HTMW (pp. 24–27; chap. 2), mental life consists of information-
processing or computation. Beliefs are a kind of information, thinking a kind of
computation, and emotions, motives, and desires are a kind of feedback mechanism
in which an agent senses the difference between a current state and goal state and
executes operations designed to reduce the difference. ‘Computation’ in this
context does not refer to what a commercially available digital computer does
but to a more generic notion of mechanical rationality, a concept that Fodor
himself has done much to elucidate (Fodor, 1968; 1975; 1981; 1994).
In this conception, a computational system is one in which knowledge and goals
are represented as patterns in bits of matter (‘representations’). The system is
designed in such a way that one representation causes another to come into
existence; and these changes mirror the laws of some normatively valid system
like logic, statistics, or laws of cause and effect in the world. The design of the
system thus ensures that if the old representations were accurate, the new ones are
accurate as well. Deriving new accurate beliefs from old ones in pursuit of a goal is
not a bad definition of ‘intelligence’, so a principal advantage of the computational
theory of mind (CTM) is that it explains how a hunk of matter (a brain or a
computer) can be intelligent.
CTM has other selling points. It bridges the world of mind and matter,
dissolving the ancient paradox of how seemingly ethereal entities like reasons,
intentions, meanings, and beliefs can interact with the physical world.
#
Blackwell Publishing Ltd. 2005
So How Does the Mind Work? 3
It motivates the science of cognitive psychology, in which experimenters char-
acterize the mind’s information structures and processes (arrays for images, tree
structures for sentences, networks for long-term memory, and so on). Since
computational systems can have complex conditions, loops, branches, and filters
which result in subtle, situationally appropriate behavior, the CTM allows the
mind to be characterized as a kind of biological mechanism without calling to
mind the knee-jerk reflexes and coarse drives and imperatives that have made
people recoil from the very idea. Finally, mental life—internal representations
and processes—appears to be more lawful and universal than overt behavior,
which can vary with circumstances. This is behind Chomsky’s idea that there is
a single Universal Grammar that applies to all the world’s languages despite their
differences in overt words and constructions. Much of HTMW extends this idea
to other areas of human psychology, such as the emotions, social and sexual
relations, and humor.
Fodor, as I have acknowledged, deserves credit for capturing the sense of
‘computation’ in which it can sensibly be said that the mind is a kind of computer.
That sense—in which a system’s state transitions map onto logical relationships, or,
as Fodor often puts it, the components of the system have both causal and semantic
properties—says nothing about binary digits, program counters, register operations,
stored programs, or any of the other particulars of the machines that process our
email or compute our taxes and which are improbable characterizations of a human
brain. The beauty of Fodor’s original formulation is that it embraces a variety of
systems that we might call ‘computational’, including ones that perform parallel
computation, analogue computation (as in slide rules and adding machines), and
fuzzy computation (in which graded physical variables represent the degree to
which something is true, or the probability that something is true, and the physical
transitions are designed to mirror operations in probability theory or fuzzy logic
rather than in classical logic). Any adequate characterization of the concept of
‘computation’ should embrace these possibilities. After all, the term digital computer
is not redundant, and the terms analogue computer and parallel computer are not self-
contradictory.
At the same time, the computational theory of mind is by no means empty
or necessary. It can be distinguished from the traditional belief that intelligence
comes from an immaterial substance, the soul. It differs from the claim that
intelligence is made possible only by specific biochemical properties of neural
tissue. It differs from the assertion that mental life can be understood only in
terms of first-person present-tense subjective experience. And it differs from
the claim that intelligence can be understood only by considering what mental
states refer to in the world, or by examining the incarnate person embedded in
a physical and social context. Fodor emphasizes the idea that the representa-
tions in a computational system are syntactic: they are composed of parts in
some arrangement, and the causal mechanisms of the system are sensitive to the
identity and arrangement of those parts rather than to what they refer to in the
world.
#
Blackwell Publishing Ltd. 2005
4 S. Pinker
The Concept of Specialization in
How the Mind Works
HTMW does not try to account for all of human behavior using a few general-
purpose principles such as a large brain, culture, language, socialization, learn-
ing, complexity, self-organization, or neural-network dynamics. Rather, the
mind is said to embrace subsystems dedicated to particular kinds of reasoning
or goals (pp. 27–31). Our intelligence, for example, consists of faculties
dedicated to reasoning about space, number, probability, logic, physical objects,
living things, artifacts, and minds. Our affective repertoire comprises emotions
pertaining to the physical world, such as fear and disgust, and emotions
pertaining to the social and moral worlds, such as trust, sympathy, gratitude,
guilt, anger, and humor. Our social relationships are organized by distinct
psychologies applied to our children, parents, siblings, other kin, mates, sex
partners,friends,enemies,rivals,tradingpartners,andstrangers.Wearealso
equipped with communicative interfaces, most prominently language, gesture,
vocal calls, and facial expressions.
The intended analogy is to the body, which is composed of systems divided into
organs assembled from tissues built out of cells. Our ‘organs of computation’,
therefore, are not like discrete chips laid out on a board with a few solder tracks
connecting them. Just as some kinds of tissue, like the epithelium, are used (with
modifications) in many organs, and some organs, like the blood and the skin,
interact with the rest of the body across an extensive interface, some kinds of
specialized thoughts and emotions may serve as constituents that are combined into
different assemblies. The concept of an artifact, for example—an object fashioned
by an intelligent agent to bring about a goal—combines the concept of an object
from intuitive physics with the concept of a goal from intuitive psychology. The
psychology of sibling relations embraces the emotion of affection (also directed
toward mates and friends), an extra feeling of solidarity triggered by perceived
kinship, and a version of disgust pinned to the thought of having sexual relations
with the sibling.
This kind of faculty psychology has numerous advantages. It is consistent with
models of cognitive faculties such as language, spatial cognition, and audition that
require specialized machinery (nouns and verbs, allocentric and egocentric frames
of reference, and pitch and timbre, respectively). It is supported by the existence of
neurological and genetic disorders that target these faculties unevenly, such as
a difficulty in recognizing faces (and facelike shapes) but not other objects, or a
difficulty in reasoning about minds but not about objects or pictures. Finally, a
faculty psychology is necessary to account for many of the complex but systematic
patterns in human thought and emotion. The fact that we love our siblings but
don’t want to have sex with them, and may want to have sex with attractive
strangers without necessarily loving them, is inexplicable by a theory of social
psychology that doesn’t distinguish among kinds of human relationships but
appeals only to global drives like ‘positive affect’.
#
Blackwell Publishing Ltd. 2005
So How Does the Mind Work? 5
The Appeal to Evolution in
How the Mind Works
Evolution is the third key idea in HTMW (pp. 21–24; chap. 3). The organs of
computation that make up the human mind are not tailored to solve arbitrary
computational problems but only those that increased the reproductive chances of
our ancestors living as foragers in pre-state societies.
One advantage of invoking evolution is that it provides psychology with
explanatory adequacy. It helps account for why we have the specializations we
do: why children learn spoken language instinctively but written language only
with instruction and effort, why the system for recalling memories satisfies many of
the specifications of an optimal information-retrieval system, why our preferred
sexual partners are nonsiblings who show signs of health and fertility. More
generally, it explains why the human psyche has specific features that could not
be predicted from the mere proposition that the brain engages in computation.
Evolutionary psychology also helps to explain many instances of error, irration-
ality, and illusion—why we gamble, eat junk food, fall for visual illusions, obsess
over celebrities, and fear snakes and heights more than hair dryers near bathtubs or
driving without a seatbelt. The nature of the explanation is that there can be a
mismatch between the ancestral environment to which our minds are evolution-
arily adapted and the current environment in which we find ourselves.
The most general attraction of a synthesis between cognitive science and
evolutionary psychology is that it continues the process of the unification of
putatively incommensurable metaphysical realms that has been the major thrust
of science for four centuries (Tooby and Cosmides, 1992; Wilson, 1998). Newton
united the sublunary and superlunary spheres, Lyell united the formative past and
static present, W¨ hler united living tissue and nonliving chemistry, and Darwin,
Mendel, and Watson and Crick united seeming teleological design in organisms
with ordinary processes of forward causation. In the same way, the idea that the
human mind is an evolved computer aims to bridge the last major chasm in human
knowledge, that between matter and mind, biology and culture, nature and
society, the natural sciences and the humanities. This consilience promises not
only a more parsimonious metaphysics but greater depth and explanatory power
for the disciplines that study the mind and its products. Hypotheses about psycho-
logical function cannot be conjured up by whim but must be compatible with
evolutionary biology and in some cases may be deduced from it.
I turn now to how each of these themes is treated in Fodor’s critique of HTMW.
The Concept of Computation in
The Mind Doesn’t Work that Way
In TMDWTW, Fodor argues that he never meant that all of the mind could be
explained as a kind of computation. On the contrary, there is a key thing that a
human mind can do but which a computational system cannot do. I will discuss
#
Blackwell Publishing Ltd. 2005
[ Pobierz całość w formacie PDF ]