Textbook Ebook Explanation and Integration in Mind and Brain Science 1St Edition David M Kaplan All Chapter PDF
Textbook Ebook Explanation and Integration in Mind and Brain Science 1St Edition David M Kaplan All Chapter PDF
Textbook Ebook Explanation and Integration in Mind and Brain Science 1St Edition David M Kaplan All Chapter PDF
Explanation and
Integration in Mind
and Brain Science
edited by
David M. Kaplan
1
OUP CORRECTED PROOF – FINAL, 11/06/2017, SPi
3
Great Clarendon Street, Oxford, ox2 6dp,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© the several contributors 2017
The moral rights of the authors have been asserted
First Edition published in 2017
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2017956292
ISBN 978–0–19–968550–9
Printed and bound by
CPI Group (UK) Ltd, Croydon, CR0 4YY
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
OUP CORRECTED PROOF – FINAL, 11/06/2017, SPi
Contents
Index 257
OUP CORRECTED PROOF – FINAL, 11/06/2017, SPi
OUP CORRECTED PROOF – FINAL, 11/06/2017, SPi
List of Figures
List of Contributors
1
Integrating Mind
and Brain Science
A Field Guide
David M. Kaplan
1. Introduction
Long a topic of discussion among philosophers and scientists alike, there is growing
appreciation that understanding the complex relationship between neuroscience and
psychological science is of fundamental importance to achieving progress across these
scientific domains. Is the relationship between them one of complete autonomy or
independence—like two great ships passing in the night? Or is the relationship a
reductive one of total dependence—where one is subordinate to the other? Or perhaps
the correct picture is one of mutually beneficial interaction and integration—lying
somewhere in between these two extremes? One primary strategy for addressing this
issue, and one that occupies center stage in this volume, involves understanding
the nature of explanation in these different domains. Representative questions taken
up by various chapters in this volume include: Are the explanatory patterns employed
across these domains similar or different in kind? If their explanatory frameworks do
in fact differ, to what extent do they inform and constrain each other? And finally, how
should answers to these and other related questions shape our thinking about the pros-
pects for integrating mind and brain science?
Several decades ago, during the heyday of computational cognitive psychology, the
prevailing view was that the sciences of the mind and brain enjoy a considerable degree
of independence or autonomy from one another—with respect to their theories, their
research methods, and the phenomena they elect to investigate (e.g., Fodor 1974; Johnson-
Laird 1983; Lachman et al. 1979; Newell and Simon 1972; Pylyshyn 1984; Simon 1979). In
an expression of the mainstream perspective in the field at the time, the psychologist
Philip Johnson-Laird proposes that “[t]he mind can be studied independently from
the brain. Psychology (the study of the programs) can be pursued independently
from neurophysiology (the study of the machine code)” (Johnson-Laird 1983, 9).
OUP CORRECTED PROOF – FINAL, 11/08/2017, SPi
2 DAVID M. KAPLAN
In the intervening decades, the doctrine of disciplinary autonomy has fallen on hard
times. Today, it is far from being the universally held or even dominant view. In fact,
given the emergence of cognitive neuroscience as a new scientific field formed precisely
at the interface between these disciplines, one might reasonably wonder whether the
consensus has now shifted in exactly the opposite direction—towards a view of complete
disciplinary integration and interdependence rather than autonomy. In the inaugural
issue of the Journal of Cognitive Neuroscience, then editor Michael Gazzaniga writes:
In the past 10 years, there have been many developments in sciences concerned with the study of
mind. Perhaps the most noteworthy is the gradual realization that the sub-disciplines committed
to the effort such as cognitive science, neuroscience, computer science and philosophy should not
exist alone and that each has much to gain by interacting. Those cognitive scientists interested in
a deeper understanding of how the human mind works now believe that it is maximally fruitful
to propose models of cognitive processes that can be assessed in neurobiologic terms. Likewise, it
is no longer useful for neuroscientists to propose brain mechanisms underlying psychological
processes without actually coming to grips with the complexities of psychological processes
involved in any particular mental capacity being examined. (Gazzaniga 1989, 2)
From the outset, contributors to the cognitive neuroscience movement have explicitly
recognized the interdisciplinary and integrative nature of the field, which is unified
by the common goal of trying to decipher how the mind and brain work (Boone and
Piccinini 2016; Churchland and Sejnowski 1988). Despite the rapidly growing influ-
ence of cognitive neuroscience and cognate fields such as computational neuroscience,
some researchers continue to maintain that neuroscience is largely or completely
irrelevant to understanding cognition (e.g., Fodor 1997; Gallistel and King 2009). Others
maintain that psychology is (or ought to be) a tightly integrated part of the broader
scientific enterprise to discover and elucidate the multi-level mechanisms underlying
mind and cognition (e.g., Boone and Piccinini 2016; Piccinini and Craver 2011).
Hence, the debate over an autonomous psychology remains incompletely settled.
The objective of this chapter is to provide a field guide to some of the key issues that
have shaped and continue to influence the debate about explanation and integration
across the mind and brain sciences. Along the way, many of the central proposals
defended in the individual chapters will be introduced and important similarities and
differences between them will be highlighted. Since questions on this topic have a long
track record of philosophical and scientific engagement, providing some of the broader
historical and theoretical context will facilitate a deeper appreciation of the contribu-
tions each individual chapter makes to these important and ongoing debates.
it will prove useful to have working definitions of these key concepts, which recur
throughout this introductory chapter as well as the volume more generally.
First, consider the notion of autonomy. Generally speaking, autonomy implies
independence from external influence, control, or constraint. The Tibet Autonomous
Region is, at least according to the Chinese Government, appropriately so called
because it is free of direct external control from Beijing. Autonomous robotic
vehicles are appropriately so called because they are capable of sensing and navigat-
ing in their environments without reliance on direct human input or control. In a
similar manner, scientific disciplines may also be autonomous from one another.
Following Piccinini and Craver (2011), we might provisionally define one discip-
line as being autonomous from another when at least one of the following c onditions
is satisfied:
(a) they can independently select which phenomenon to investigate,
(b) they can independently select which methods to use,
(c) they can independently select which theoretical vocabulary to apply,
(d) the laws/theories/explanations from one discipline are not reducible to the
laws/theories/explanations of the other discipline, or
(e) evidence from one discipline does not exert any direct constraints on the laws/
theories/explanations of the other discipline.
Importantly, this characterization of autonomy is flexible and admits of degrees.
A scientific discipline can in principle completely or partially satisfy one or more of
these conditions (a–e), and consequently can be completely or partially autonomous
with respect to another discipline. At one extreme, a discipline may only incompletely
or partially satisfy a single condition, comprising a minimal form of a utonomy. At
the other extreme, a discipline may completely satisfy all conditions, instantiating a
maximal form of autonomy (at least with respect to identified conditions a–e).
The notion of distinctness is closely related, but logically weaker. Disciplines exhibit
distinctness if they investigate different kinds of phenomena, use different kinds of
methods, or construct different kinds of explanations. The last of these is most relevant
in the context of the present volume. As we will see, the thesis of the explanatory dis-
tinctness of neuroscience and psychology—roughly, that they employ characteristically
different kinds of explanation—is a key premise in a number of recent arguments for
the autonomy of psychology.
It is important to distinguish between autonomy and distinctness because one can
obtain without the other. Generally speaking, distinctness is a necessary but insufficient
condition for autonomy (for additional discussion, see Piccinini and Craver 2011).
Without distinctness there is clearly no scope for autonomy. If two disciplines investi-
gate the same phenomena, in an important sense, they cannot independently select which
phenomenon to investigate. They are instead constrained or bound to investigate the
same phenomena. Similarly, if two disciplines employ the same methods or theoretical
vocabularies, in an important sense, they cannot independently select which methods
OUP CORRECTED PROOF – FINAL, 11/08/2017, SPi
4 DAVID M. KAPLAN
or theoretical vocabularies to use. They are bound to use the same across the disciplines.
Although distinctness is required for autonomy, it does not entail it. Two or more
things can be distinct yet be mutually dependent or interdependent. Consider a simple
example. The Earth is distinct from the Sun, yet these systems influence one another in
a multitude of ways (e.g., gravitationally and thermodynamically). They are distinct,
but not autonomous in any interesting sense of the word. Similarly, a scientific field or
discipline may have its own distinct laws, principles, and theories, yet these may turn
out to be reducible to or evidentially constrained by those of another discipline. Even
though distinctness does not entail autonomy, as will be discussed shortly, they are
often endorsed as a package deal.
(that which does the explaining) appears as the premise set, which includes statements
characterizing the relevant empirical conditions under which the phenomenon
obtains (initial conditions) and at least one general law required for the derivation of
the explanandum. According to the view, good scientific explanations are those in
which the explanans provides strong or conclusive evidential grounds for expecting
the explanandum-phenomenon to occur (Hempel 1965).
In its most general formulation, the covering law account is intended to apply
uniformly to the explanation of spatiotemporally restricted events such as the explo-
sion of the Space Shuttle Challenger, as well as the explanation of general regularities
or laws such as the explanation of Kepler’s laws of planetary motion in terms of more
basic laws of Newtonian mechanics. A derivation of one or more sets of laws (com-
prising a theory) from another set of laws (comprising another theory) is known as
an intertheoretic reduction. According to the covering law account, intertheoretic
reduction comprises a special case of deductive-nomological explanation.
Nagel (1961) developed these ideas into an explicit model of theory reduction, pro-
posing that a theory (or law) from a higher-level science such as psychology can be
reduced to, and thereby explained by, a theory (or law) from a lower-level science
such as neuroscience or biology just in case (a suitably axiomatized version of) the
higher-level theory can be logically derived from (a suitably axiomatized version of)
the lower-level theory. Since the terminology employed in both the reduced and
reducing theories will invariably differ in some way, so-called bridge principles or
rules of correspondence are required to establish links between the terms of the two
theories. For example, a bridge principle might connect terms from thermodynamics
such as “heat” with those of statistical mechanics such as “mean molecular energy.”
Finally, because the reduced theory will typically only apply over a restricted part
of the domain of the reducing theory or at certain limits, boundary conditions that set
the appropriate range for the reduction are often required in order for the derivation
to be successful. The theory reduction model can be represented schematically as
follows (Bechtel 2008, 131):
Lower-level laws (in the basic, reducing science)
Bridge principles
Boundary conditions
_______________________________
∴ Higher-level laws (in the secondary, reduced science).
Oppenheim and Putnam (1958) famously argue that the Logical Positivists’ grand
vision of scientific unification can finally be achieved, at least in principle, by reveal-
ing the derivability relationships between the theories of the sciences. They start by
assuming that each scientific discipline occupies a different level within a single
global hierarchy. The Oppenheim–Putnam framework then involves an iterated
sequence of reductions (so-called micro-reductions) starting with the reduction of
some higher-level theory to the next lowest-level theory, which in turn is reduced to
OUP CORRECTED PROOF – FINAL, 11/08/2017, SPi
6 DAVID M. KAPLAN
the next lowest-level theory, and so on, until the level of fundamental physical theory
is eventually reached. As Fodor succinctly puts it: “all true theories in the special sci-
ences should reduce to physical theories in the long run” (1974, 97). Oppenheim and
Putnam’s general framework entails a specific conception of how psychology will
eventually reduce to neuroscience and beyond:
It is not absurd to suppose that psychological laws will eventually be explained in terms of the
behavior or individual neurons in the brain; that the behavior of individual cells—including
neurons—may eventually be explained in terms of their biochemical constitution; and that the
behavior of molecules—including the macro-molecules that make up living cells—may even-
tually be explained in terms of atomic physics. If this is achieved, then psychological laws will
have, in principle, been reduced to laws of atomic physics. (Oppenheim and Putnam 1958, 7)
Although many philosophers once held out high hopes for reductive successes of this
kind, few are so optimistic today. The theory reduction account faces challenges along
several fronts including those raised about its adequacy as a general account of the
relations between the sciences and as a specific account of the relation between neuro-
science and psychology. Its descriptive adequacy as a general account of reduction in
science has been called into question as it has proved exceedingly difficult to locate real
examples that satisfy the account even in domains thought to be paradigmatic such as
physics (e.g., Sklar 1967). Other general issues concern its oversimplified or otherwise
inaccurate portrayal of the relationships between the various sciences including the
relationships between the theories, concepts, and explanandum phenomena of those
sciences (e.g., Bickle 1998; Churchland 1989; Feyerabend 1962; Schaffner 1967, 1969;
Wimsatt 2007). Yet, it is the specific challenges that stand out as most relevant for
present purposes.
One primary reason for heightened skepticism about theory reduction as an
adequate account of the specific relationship between neuroscience and psychology is
the conspicuous absence of laws or lawlike generalizations in these sciences. This is what
Rosenberg (2001), in the context of biology, aptly refers to as the “nomological vac-
uum.” Since unification is supposed to be achieved by deriving the laws of psychology
from the laws of neuroscience (or some other lower-level science such as biophysics),
clearly a precondition for such unification is the availability of laws at both the level of
reduced and reducing theories. If the theoretical knowledge of a given discipline can-
not be specified in terms of a set of laws (an assumption that mechanists and others
reject), there is simply no scope for unification along these lines. Yet, despite decades of
effort to identify genuine lawful generalizations in psychology or neuroscience of the
sort one finds in other scientific disciplines such as physics, few if any candidate laws
have been revealed.
In their chapter, Martin Roth and Robert Cummins echo similar criticisms about the
“nomic conception of science” at the heart of the covering law framework. As Cummins
puts it in his earlier and highly influential work on the nature of psychological
explanation: “Forcing psychological explanation into the subsumptivist [covering
OUP CORRECTED PROOF – FINAL, 11/08/2017, SPi
law] mold made it continuous with the rest of science only at the price of making it
appear trivial or senseless” (Cummins 1983, 27). In their chapter, Roth and Cummins
identify one source of confusion underwriting views that attribute an explanatory
role to laws in psychological science. Building on previous work by Cummins (2000),
they indicate how the term “law” in psychology is often (confusingly) used by
researchers working in the field to refer to effects (i.e., robust patterns or regularities),
which are the targets of explanation rather than explanations in themselves. For
example, Fitts’ law describes but does not explain the widely observed tradeoff
between speed and accuracy in skilled human motor behavior. The Weber–Fechner
law describes but does not explain how the just-noticeable difference between two
stimuli is proportional to the magnitude of the stimuli. Nevertheless, someone
might be tempted to try to read off the nomological character of psychological science
(and the explanatory role of psychological laws) from the mere appearance of the
word “law” in these instances. Yet these nominal laws, which simply describe effects
or phenomena to be explained, do not satisfy any of the standardly accepted criteria
for lawhood such as being exceptionless, having wide scope, etc., and are thus poorly
suited to play the required role in covering law explanation and theory reduction.
Roth and Cummins instead maintain that psychological laws are better understood
as capturing the explananda for psychological science rather than the explanans,
and argue that, appearances notwithstanding, psychological explanations do not
involve subsumption under laws. Their efforts to expose how the nomic character
of psychology is largely illusory places additional pressure on efforts to recruit the
covering law framework to shed light on the nature of psychological explanation
and reduction.
Another reason many participants in this debate see intertheoretic reduction as
a problematic way to achieve unification among the scientific disciplines is that
successful reduction renders the laws and theories of the higher-level (reduced) science
expendable in principle. Since all of the laws and all observational consequences of the
higher-level (reduced) theory can be derived directly from information contained in
the lower-level theory, the resulting picture is one in which the higher-level sciences
in principle provide no distinctive, non-redundant explanatory contribution over and
above that made by the lower-level science. As Fodor puts it, reductionism has “the
curious consequence that the more special sciences succeed, the more they ought to
disappear” (1974, 97). In practice, however, higher-level sciences might retain their
usefulness temporarily until the lower-level sciences become theoretically mature
enough to support the reductions on their own, or they might play heuristic roles such
as revealing the regularities or phenomena that the lower-level sciences seek to explain.
Hence, even hard-core reductionists such as John Bickle can admit that “psychological
causal explanations still play important heuristic roles in generating and testing
neurobiological hypotheses” (author’s emphasis; Bickle 2003, 178). But this picture
will nevertheless appear deeply unsatisfying to those who seek to secure a long-term
explanatory role for psychological science. For these and other reasons, using theory
OUP CORRECTED PROOF – FINAL, 11/08/2017, SPi
8 DAVID M. KAPLAN
reduction as the basis for an account of the relationship between psychology and
neuroscience has appeared unpromising to many.
The traditional theory reduction framework thus offers one potential strategy for
unifying or integrating psychological and brain science, but one that is fraught with
problems. The well-known considerations rehearsed above indicate that the pros-
pects for achieving unification by reduction are either extremely dim due to the lack
of explanatory laws in psychology and neuroscience, or else reduction can succeed
but in doing so would impose unbearably heavy costs by rendering psychology
explanatorily inert and obsolete. Neither of these paths appears particularly promis-
ing. This has consequently inspired a search for alternative ways of characterizing
the relationship between the sciences of the mind and the brain that do not bottom
out in theory reduction, including those that manage to secure some degree of
autonomy for psychology.
Before moving on, it is worth pausing briefly to describe another reductionist
account—importantly distinct from the theory reduction account—that has received
considerable attention in recent decades. This is the “ruthless reductionism” account
advocated primarily by John Bickle (2003, 2006). Bickle’s account rejects a number of
core assumptions of the theory reduction view including that laws are central to reduc-
tion, and that successful reductions of the concepts and kinds posited by higher-level
theories to those of some basic lower-level theory proceeds via a sequence of step-wise
reductions. According to ruthless reductionism, reductions can instead be “direct” (i.e.,
without any intermediate steps) such as the “reductions of psychological concepts and
kinds to molecular-biological mechanisms and pathways” (Bickle 2006, 412). Bickle
argues that researchers in lower-level neuroscience such as cellular and molecular
neuroscience accomplish these direct reductions by experimentally intervening at the
cellular or molecular level and producing detectable effects at the level of the phenom-
enon to be explained (the behavioral or psychological level). Accordingly, there is a
path for reduction that skips over any intermediary levels.
Despite its role in the broader debate, ruthless reductionism exhibits many similar
problems to traditional theory reduction accounts. In particular, it treats higher-level
explanations in psychology (and even higher-level fields within neuroscience includ-
ing cognitive neuroscience and systems neuroscience) as expendable in principle,
and therefore fails to secure a permanent role for explanations developed at these
higher levels. It therefore fails to exemplify the type of “middle ground” integrative
views about the relationship between psychology and neuroscience emphasized in
this volume.
3.2 Autonomy and multiple realizability
Another traditional response that philosophers have given is to argue that psychology
exhibits a similar kind of autonomy with respect to “lower-level” sciences such as
neuroscience in the sense that their theories or explanations are unconstrained by
evidence about neural implementation.
OUP CORRECTED PROOF – FINAL, 11/08/2017, SPi
Many early defenses of the autonomy of psychology and other higher-level sciences
involved appeals to multiple realizability in order to deny the possibility of reducing
the theories or laws of the higher-level science to those of the lower-level science
(e.g., Fodor 1974, 1997; Putnam 1975). These views emerged as direct responses to the
traditional theory reduction model and its negative implications for the independent
status of psychology and the rest of the special sciences.
Recall that according to the classical theory reduction model, successful intertheo-
retic reduction requires a specification of appropriate bridge principles and boundary
conditions (Nagel 1961). Bridge principles establish systematic mappings or identities
between the terms of the two theories, and are essential for the reduction to go through.
Anti-reductionists therefore naturally gravitate towards these bridge principles in their
attacks, claiming that bridge principles will generally be unavailable given that the
events picked out by special science predicates or terms (e.g., functionally defined terms
such as “money” or “pain”) will be “wildly disjunctive” from the perspective of lower-
level sciences such as physics (Fodor 1974, 103). In other words, the enterprise to build
bridge principles connecting the vocabularies or predicates of the higher- and lower-
level sciences in an orderly, one-to-one manner breaks down because higher-level phe-
nomena are often multiply realized by heterogeneous sets of lower-level realizers. Put
somewhat differently, multiple realizability entails that the predicates of some higher-
level science will cross-classify phenomena picked out by predicates from a lower-level
science. The one-to-many mapping from the psychological to the neurobiological (or
physical) implied by multiple realizability renders the bridge principle building enter-
prise at the heart of the theory reduction model a non-starter. Since the establishment of
bridge principles is a necessary condition for classical intertheoretic reduction, multiple
realizability directly implies the irreducibility and autonomy of psychology.
This line of argument has connections to functionalist and computationalist views
in the philosophy of mind, which also depend on a notion of multiple realizability.
According to one influential version of computationalism, cognition is identified with
digital computation over symbolic representations (Newell and Simon 1976; Anderson
1996; Johnson-Laird 1983; Pylyshyn 1984). Proponents of computationalism have
long maintained that psychology can accomplish its explanatory objectives without
reliance on evidence from neuroscience about underlying neural mechanisms.
Multiple realizability is taken to justify a theoretically principled neglect of neurosci-
entific data based on the alleged close analogy between psychological processes and
running software (e.g., executing programs) on a digital computer, and the multiple
realizability of the former on the latter. According to the analogy, the brain merely pro-
vides the particular hardware on which the cognitive programs (e.g., software) happen
to run, but the same software could in principle be implemented in indefinitely many
other hardware platforms. For this reason, the brain is deemed a mere implementation
of the software. If the goal is to understand the functional organization of the
software—the computations being performed—determining the hardware details is a
relatively unimportant step.
OUP CORRECTED PROOF – FINAL, 11/08/2017, SPi
10 DAVID M. KAPLAN
[W]e could, if we liked, require the taxonomies of the special sciences to correspond to the
taxonomy of physics [or neuroscience] by insisting upon distinctions between the natural
kinds postulated by the former wherever they turn out to correspond to distinct natural kinds
in the latter. (Fodor 1974, 112)
Xarquedas, 431
Xingú R., 383, 402
Yacuiba, 210, 220, 223, 243, 312
Yaguachi R., 127
Yaguarón R., 359
Yapurá R., 17, 406
Yaracuy, 54, 55, 66, 77, 81, 93
Yareta, 237
Yaritagua, 66
Yauli, 197, 200
Yauricocha, 197, 198
Yerba Mate, 324, 347, 393, 422
Yhú, 335
Ypoa Lake, 338, 340
Yucca, 142
Yungas, 213, 215, 224, 226, 234, 242, 243
Yungay, 166
Yurimaguas, 173
Yuruán, 108
Yuruary R., 89, 92, 93
Updated editions will replace the previous one—the old editions will
be renamed.
1.D. The copyright laws of the place where you are located also
govern what you can do with this work. Copyright laws in most
countries are in a constant state of change. If you are outside the
United States, check the laws of your country in addition to the terms
of this agreement before downloading, copying, displaying,
performing, distributing or creating derivative works based on this
work or any other Project Gutenberg™ work. The Foundation makes
no representations concerning the copyright status of any work in
any country other than the United States.