Professional Documents
Culture Documents
Ebook The New Behaviorism Foundations of Behavioral Science 3Rd Edition John Staddon Online PDF All Chapter
Ebook The New Behaviorism Foundations of Behavioral Science 3Rd Edition John Staddon Online PDF All Chapter
https://ebookmeta.com/product/foundations-of-behavioral-
neuroscience-10th-edition-neil-carlson/
https://ebookmeta.com/product/foundations-of-python-network-
programming-3rd-edition-brandon-rhodes-john-goerzen-2/
https://ebookmeta.com/product/foundations-of-python-network-
programming-3rd-edition-brandon-rhodes-john-goerzen/
https://ebookmeta.com/product/foundations-of-python-network-
programming-3rd-edition-brandon-rhodes-john-goerzen-3/
Concise Guide to Neuropsychiatry and Behavioral
Neurology 3rd ed 3rd Edition John Barry
https://ebookmeta.com/product/concise-guide-to-neuropsychiatry-
and-behavioral-neurology-3rd-ed-3rd-edition-john-barry/
https://ebookmeta.com/product/philosophy-of-social-science-the-
philosophical-foundations-of-social-thought-3rd-edition-ted-
benton/
https://ebookmeta.com/product/clinical-psychology-for-trainees-
foundations-of-science-informed-practice-3rd-edition-andrew-c-
page/
https://ebookmeta.com/product/foundations-of-hyperbolic-
manifolds-third-edition-3rd-ed-instructor-solution-manual-
solutions-john-g-ratcliffe/
https://ebookmeta.com/product/foundations-of-marketing-7th-
edition-john-fahy/
“The New Behaviorism is a feast of clear thinking and deep scholarship, one
deserving careful attention by students and scholars in any feld who are
curious about human behavior — and about why we understand ourselves so
poorly. Professor Staddon recasts some of the simplistic notions that launched
behavioral psychology in the early 1900s, challenging us to understand
behavior in all its glorious complexity while still avoiding the shoddy thinking
that has driven investigations of a ‘mental world’ for more than a century.”
Robert Epstein, American Institute for Behavioral Research and
Technology, USA
“B. F. Skinner returned to the public eye after tech companies began to
use his conditioning techniques to glue users, expectedly waiting for the
next Like, to their social networks. John Staddon ofers a deep look into
Skinner’s thinking, why he rejected freedom and dignity, how he made
pigeons superstitious, and why he believed that human behavior should be
controlled. The New Behaviorism provides a well-written and well-reasoned
analysis of the potential and limits of behaviourism, new and old. It is a
marvellous guide to understanding the uneasy relation between behaviorism
and the rest of psychology.”
Gerd Gigerenzer, Max Planck Institute for
Human Development, Berlin
Third Edition
John Staddon
Third edition published 2021
by Routledge
605 Third Avenue, New York, NY 10158
and by Routledge
2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2021 John Staddon
The right of John Staddon to be identifed as author of this work has
been asserted by him in accordance with sections 77 and 78 of the
Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or
reproduced or utilised in any form or by any electronic, mechanical,
or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or retrieval
system, without permission in writing from the publishers.
Trademark notice: Product or corporate names may be trademarks
or registered trademarks, and are used only for identifcation and
explanation without intent to infringe.
First edition published by Psychology Press 2001
Second edition published by Psychology Press 2014
Library of Congress Cataloging-in-Publication Data
Names: Staddon, J. E. R., author.
Title: The new behaviorism : foundations of behavioral science / John Staddon.
Description: Third edition. | New York, NY : Routledge, 2021. | Includes
bibliographical references and index. | Summary: “This ground-breaking book
presents a brief history of behaviorism, along with a critical analysis of radical
behaviorism, its philosophy and its applications to social issues. This third edition
is much expanded and includes a new chapter on experimental method as well
as longer sections on the philosophy of behaviorism. It ofers experimental and
theoretical examples of a new approach to behavioral science. It provides an
alternative philosophical and empirical foundation for a psychology that has rather
lost its way. The mission of the book is to help steer experimental psychology
away from its current undisciplined indulgence in “mental life” toward the core
of science, which is an economical description of nature. The author argues that
parsimony - the elementary philosophical distinction between private and public
events, even biology, evolution and animal psychology are all ignored by much
contemporary cognitive psychology. The failings of radical behaviorism as well
as a philosophically defective cognitive psychology point to the need for a new
theoretical behaviorism, which can deal with problems such as “consciousness”
that have been either ignored, evaded or muddled by existing approaches. This
new behaviorism provides a unifed framework for the science of behavior that
can be applied both to the laboratory and to broader practical issues such as law
and punishment, the health-care system, and teaching”— Provided by publisher.
Identifers: LCCN 2020053775 (print) | LCCN 2020053776 (ebook) |
ISBN 9780367745813 (hardback) | ISBN 9780367745806 (paperback) |
ISBN 9781003158578 (ebook)
Subjects: LCSH: Behaviorism (Psychology)
Classifcation: LCC BF199 .S76 2021 (print) | LCC BF199 (ebook) |
DDC 150.19/43—dc23
LC record available at https://lccn.loc.gov/2020053775
LC ebook record available at https://lccn.loc.gov/2020053776
ISBN: 978-0-367-74581-3 (hbk)
ISBN: 978-0-367-74580-6 (pbk)
ISBN: 978-1-003-15857-8 (ebk)
Typeset in Bembo
by Apex CoVantage, LLC
To DB
9 Taylor & Francis
Taylor & Francis Group
http://taylorandfra ncis.com
Contents
PART I
History 1
PART II
Explanation 53
4 Behaviorist Theory 55
5 Radical Theory 74
6 Variation and Selection 87
7 Behavior–Evolution Parallels 108
8 Experimental Methods 117
9 What Is Rational? 134
10 Philosophy of Behaviorism 147
11 Free Will, Behaviorism, and Utopia 161
12 Values 176
13 Behaviorism and Mental Life 185
x Contents
14 Cognitivism and the New Behaviorism 191
15 The New Behaviorism 209
16 Internal States: The Logic of Historical Systems 219
17 Consciousness and Behaviorism 233
18 Three Domains 240
PART III
Behaviorism and Society 257
Index 305
Preface to the Third Edition
Many people were kind enough to comment on chapters and on talks I have
given on some of the topics of the book. I thank in particular John Malone,
Nancy Innis, Jennifer Higa, Max Hocutt, Kent Berridge, Geofrey Hall,
Eugene Moss, and Peter Killeen. I thank the late Peter Harzem for the
long Pavlov quotation in Chapter 18 and for detailed comments on the
draft of the frst edition. I am especially grateful to Armando Machado (for
the frst edition) and Jérémie Jozefowiez (for the frst two editions), whose
thoughtful comments made me think very hard about many tricky points.
I thank the fve commentators on the frst edition in the July 2004 issue of
the Journal of the Experimental Analysis of Behavior and two friends, Charles
Hosler and Ralph Heinz, who read and commented on the book. I also
thank Sheri Sipka for her meticulous, efcient, and thoughtful editorial
help. I gratefully acknowledge research support from the Alexander von
Humboldt Foundation, the National Institutes of Health and the National
Science Foundation. Without the long-term support provided by a Senior
Research Scientist Award from National Institute of Mental Health, and
Duke University’s willingness to cooperate by relieving me of most teaching
and administrative responsibilities, this work could never have been under
taken. I thank Jack Marr, Clive Wynne, and Howard Rachlin for very help
ful comments on the second edition; and especial thanks to Alan Silberberg
for a very careful reading of the manuscript of this edition. I am grateful for
many years of informative discussion of several of the topics in the book
with Kevin Hoover, Bruce Caldwell, and other members of the History of
Political Economy center at Duke. Of course, all errors and misjudgments
are entirely my responsibility.
9 Taylor & Francis
Taylor & Francis Group
http://taylorandfra ncis.com
Part I
History
9 Taylor & Francis
Taylor & Francis Group
http://taylorandfra ncis.com
1 The Psychology of the
“Other One”
Figure 1.1 William James (1842–1910), philosopher, pioneer psychologist, and promoter
of the philosophy of pragmatism, at Harvard during the late 1890s. There were
two James brothers. One wrote delicate convoluted prose; the writing of the
other was vivid and direct. Surprisingly, it was the psychologist, William, who
was the easier writer.
The Psychology of the “Other One” 5
of the Weber–Fechner law, which relates the variability of a sensory judg
ment to the magnitude of the physical stimulus4 (Box 1.1). Weber–Fechner
is one of the few real laws in psychology. For very many sensory dimensions,
variability of judgment, measured as standard deviation or “just-noticeable
diference” (JND), is proportional to the mean. For example, suppose that
someone is able to correctly detect that weights of 100 g and 105 g are dif
ferent just 50% of the time—so the JND is 5%, then the Weber–Fechner
relation says that he will also be able to tell the diference between 1,000 g
and 1,050 g just 50% of the time.
Marginal Utility
A similar curve is used by economists in functional theories of eco
nomic behavior. The curve relates the amount of a good, something
like pizza, cash, or real estate, to the value it has for the owner, its
utility. The fact that the curve is negatively accelerated means that
constant increments in the good yield smaller and smaller amounts
of additional utility (the equivalent to ΔI in the Weber–Fechner law),
that is, diminishing marginal utility. Like the Weber–Fechner curve,
the curve is derived by inference from empirical results. Many (but
far from all!) equilibrium (stable) choice patterns in real markets can
be explained by assuming that people choose so as to equate marginal
utilities—as this usually maximizes their total utility. Given that mar
ginal utility is assumed to decline with amount, this assumption can
explain, for example, why someone who has much milk but no bread
will be willing to exchange some milk for some bread with someone
who likewise has a bread surplus. Diminishing marginal utility is one
way to explain why people, unlike koala bears, are omnivores rather
than single-foodies.
Origins
What is behaviorism, and how did it begin? The word was made famous
by famboyant Johns Hopkins psychologist John Broadus Watson (1878–
1958). Watson, a South Carolina native, after a rambunctious teenagerdom,
8 History
was a prodigious academic success until he had problems with his love life.
Divorced in 1920 after his wife discovered he was having an afair with his
research assistant Rosalie Rayner (a scandalous event in those days), Wat
son was fred from Johns Hopkins. Plus ça change—not that anybody would
much notice nowadays, especially as Watson soon married Rosalie. The next
year he joined the large advertising agency J. Walter Thompson, at a much-
enhanced salary, where he very successfully followed the advertising version
of the Star Trek motto: boldly creating needs where none existed before.12
His success in this new feld was no accident. The major fgures in the softer
sciences have almost invariably been people with a gift for rhetoric, the art of
verbal persuasion. They may have other skills (e.g., Skinner was a wonderful
experimenter), but their ability to found schools usually owes at least as much
to their ability to persuade and organize as to their purely scientifc talents.
Watson’s impact on psychology began with a nineteen-page article in the
theoretical journal Psychological Review in 1913,13 although the basic idea had
been foating around for a decade or more.14 In the next few years, Watson
followed up with several books advocating behaviorism. He was reacting
against the doctrine of introspection, the idea that the basic data of psychol
ogy could be gathered from one’s own consciousness or reports of the con
sciousness of others. Émigré German behaviorist Max Meyer (1873–1967)
made this distinction between private and public events explicit in the title
of his introductory textbook Psychology of the Other One (1921).
The Unconscious
The idea that people have direct access to the causes of their own behavior
has been thoroughly discredited. Introspection, as it is called, is an unreliable
and imperfect guide. Take the “tip-of-the-tongue” phenomenon. Everyone
has had the experience of being unable to recall the name of something or
someone that you “know you know.” The experience is quite frequent in
older people—that’s why old-folks groups often sport name tags, even if the
members are well known to one another. Usually, after a while, the sought
name comes to mind—the failure to recall is not permanent. But where
does it come from, and why? And how do you “know you know” (this is
called metacognition, about which more later). Introspection is no help in
fnding an answer to either question.
The great German physicist and physiologist Hermann von Helmholtz
(1821–1894; Figure 1.2) pointed out many years ago that perception
itself is a process of unconscious inference, in the sense that the brain presents
to consciousness, not the visual image that the eyes see but an inferred
view of the world. A given image or succession of two-dimensional
images can always be interpreted in several ways, each consistent with
a diferent view of reality. The brain picks one, automatically, based on
probabilities set by personal and evolutionary history. I show later that this
view of the brain as a categorizer provides a simple alternative to mental
istic accounts of consciousness.
The Psychology of the “Other One” 9
Figure 1.2 Herrmann von Helmholtz (1821–1894) German physicist, physician, and
philosopher. In psychology, he made contributions in vision and hearing,
in physics on optics and conservation of energy, and in the philosophy of
perception through his idea of “unconscious inference.”
Here was a man who could work out modular equations and theorems
of complex multiplication, to orders unheard of, whose mastery of con
tinued fractions was, on the formal side at any rate, beyond that of any
mathematician in the world. . . . All his results, new or old, right or
wrong, had been arrived at by a process of mingled argument, intuition,
and induction of which he was entirely unable to give any coherent
account.19
All too rarely do I fnd colleagues who will assent to the proposition
(which I fnd irresistible) that the very ground-rules of science, its con
cern only for public knowledge, preclude its fnding an explanation
for my consciousness, the one phenomenon of which I am absolutely
certain.25
But his emphasis was on the motor system, “covert speech,” and so on. As
we will see in later chapters, the limitations of consciousness are amply bal
anced by the autonomy of the unconscious.
B. F. Skinner, the predominant behaviorist of the twentieth century, was
less preoccupied with proprioception. But he nevertheless claimed that psy
chological theory should not include any “explanation of an observed fact
which appeals to events taking place somewhere else, at some other level
of observation, described in diferent terms, and measured, if at all, in dif
ferent dimensions”28 a position from which he occasionally deviated, as we
will see. As it stands, it is an odd view, since it would exclude from science
atomic theory, quantum theory, and genetics. For example, the law of mul
tiple proportions, a good behavioristic observation from nineteenth-century
chemistry, is explainable by reference to atoms and molecules, which are
not directly observed. The breeding patterns of peas were explained by
Mendel as a refection of the random combination of discrete hereditary
14 History
particles—genes—unobserved at that time and for many decades thereafter.
Much of physics is still a quest for entities postulated by theory but hitherto
unobserved. The detection of the Higgs boson is just the most recent in a
series of such eforts. It is a testimony to Skinner’s rhetorical skill, and the
dedication of his followers, that he managed to get away with a scientifc
position that, in efect, separates psychology from the most developed of the
“hard” sciences.
***
Behaviorism was a reaction against introspection as a scientifc method.
Watson and those who followed him were correct in this. Introspection
sheds only a very dim light on the true causes of behavior. Inaccessible to
conscious thought is the vast “dark matter” of the unconscious, responsible
for recollection, creativity, not to mention that “secret planner” whose hid
den motives complement conscious motivation.
But behaviorism also went too far in its attempts to simplify psychology.
Watson’s claim that thought is nothing but covert speech and Skinner’s puz
zling objection to theoretical entities that cannot be directly observed have
hobbled the development of behaviorism and allowed the rest of psychology
to backslide slowly into mentalism—familiar and readily compatible with
“folk psychology” but rich in philosophical error and a scientifc dead-end.
Notes
1. Roediger, R. (2004) What happened to behaviorism? American Psychological Society
Observer, 17(3), March.
2. Dawes, R. M. (1994) House of cards: Psychology and psychotherapy built on myth. New
York: The Free Press; Horgan, J. (1999) The undiscovered mind. New York: The
Free Press. “Psychology as a scientifc discipline can be seen as wallowing, perhaps
slowly disintegrating,” writes one commentator (Branch, M. [2014] Malignant side
efects of null-hypothesis signifcance testing. Theory & Psychology, 24[2], 256–277),
noting the fssiparous state of the American Psychological Association, which, as
of 2013, had 59 almost-independent divisions. This Balkanization is mirrored in
introductory psychology textbooks. See also, https://quillette.com/2018/10/07/
the-devolution-of-social-science/, and Machado, A., Lourenço, O., & Silva, F. J.
(2000) Facts, concepts, and theories: The shape of psychology’s epistemic triangle.
Behavior and Philosophy, 28(1/2), Spring–Fall.
3. R. I. Watson came up with eighteen psychological slices and dices: Watson, R. I.
(1971) Prescriptions as operative in the history of psychology. Journal of the History
of the Behavioral Sciences, 7, 311–322. The number of subdivisions has undoubtedly
increased since then.
4. See, for example, Dehaene, S. (2003) The neural basis of the Weber-Fechner law:
A logarithmic mental number line. Trends in Cognitive Sciences, 4(7), April, 145–147
and, for an account of how Weber–Fechner underlies S. S. Stevens’s power law,
see Staddon, J. E. R. (1978) Theory of behavioral power functions. Psychologi
cal Review, 85, 305–320. https://dukespace.lib.duke.edu/dspace/bitstream/handle/
10161/6003/Staddon1978.pdf?sequence=1.
5. See O’Donohue, W., & Kitchener, R. (Eds.) (1999) Handbook of behaviorism. New
York: Academic Press, for a collection of papers on the varieties of behaviorism.
The Psychology of the “Other One” 15
6. For examples of this kind of debate, check out Fodor and Staddon: https://psycrit.
com/w/A_dialogue_between_Fodor_and_Staddon, and Staddon, J. E. R. (2004)
A call to arms. The Behavior Analyst, 27, 117–118.
7. Amsel, A. (1989) Behaviorism, neobehaviorism and cognitivism in learning theory.
Hillsdale, NJ: Erlbaum, p. 1.
8. Kennedy, J. S. (1992) The new anthropomorphism. Cambridge: Cambridge University
Press.
9. Ridley, M. (1992) Animist debate. Nature, 359, 280.
10. Tudge, C. (1999) Chimps don’t talk, but they do cry. The New Statesman, August 2,
my italics.
11. http://en.wikipedia.org/wiki/Watson_(computer). The “Watson” here refers to an
IBM pioneer, not J. B. Watson, the founder of behaviorism. AlphaGo—the Movie
describes the machine’s 4–1 win over reigning champ Lee Sedol in 2016.
12. See Todd, J. T., & Morris, E. K. (Eds.) (1994) Modern perspectives on John B. Watson
and classical behaviorism. Westport, CT: Greenwood Press, for an account of Watson’s
life and work. For an account of his advertising career, see Kreshel, J. P. (2013) John
B. Watson at J. Walter Thompson: The legitimation of “science” in advertising.
Journal of Advertising, 19(2), 49–59.
13. Watson, J. B. (1913) Psychology as the behaviorist views it. Psychological Review, 20,
158–177, p. 158. https://psychclassics.yorku.ca/Watson/views.htm.
14. Mills, J. A. (1998) Control: A history of behavioral psychology. New York: New York
University Press.
15. Named after its inventor American ophthalmologist Adelbert Ames, Jr. and frst
built in 1934. There are many videos of Ames Rooms on the internet. Here is one
by famboyant psychologist Philip Zimbardo: https://www.youtube.com/watch?
reload=9&v=hCV2Ba5wrcs.
16. http://en.wikipedia.org/wiki/M%C3%BCller-Lyer_illusion.
17. For a survey, see Gregory, R. L. (2005) The Medawar lecture 2001: Knowledge
for vision: Vision for knowledge. Philosophical Transactions of the Royal Society B, 360,
1231–1251.
18. An excellent biography is Kanigel, R. (1981) The man who knew infnity. New York:
Macmillan.
19. Kanigel, R. (1981) The man who knew infnity. New York: Macmillan, p. 216.
20. Kanigel, R. (1981) The man who knew infnity. New York: Macmillan, p. 288.
21. Haward, L. UCL Magazine, Autumn, 1999.
22. Watson, J. B. (1913) Psychology as the behaviorist views it. Psychological Review, 20,
158–177, p. 158.
23. Consensus is a fawed defnition of “objectivity” or “truth,” of course. The scien
tifc majority is sometimes wrong, as in their initial rejection of Alfred Wegener’s
theory of tectonic drift and (according to some critics) the allegiance of the major
ity to the idea of catastrophic anthropogenic climate change; for example, Staddon,
J., & Morcombe, P. (2020) The case for carbon dioxide. Academic Questions, 33(2),
246–258. https://link.springer.com/article/10.1007%2Fs12129-020-09871-0.
24. Ziman, J. (1968) Public knowledge: The social dimension of science. Cambridge:
Cambridge University Press.
25. Pippard, B. (1992) Counsel of despair. Nature, 357, 29.
26. Here are Watson’s rather inconsistent comments on proprioception: “[The]
behaviorist . . . has never really held the view that thinking is merely the action of
language mechanisms.”
But then, in the next paragraph:
“A whole man thinks with his whole body in each and in every part. If he is
mutilated or if his organs are defective or lacking, he thinks with the remaining parts
left in his care: but surely he does everything else in exactly the same way. . . . [E]
veryone admits that [a tennis] player is using every cell in his body during the
16 History
game. Nevertheless if we sever a small group of muscles in his right arm his play
ing is reduced practically to that of a novice. This illustration serves us very well in
explaining why one emphasizes laryngeal processes in thinking. Surely we know
that the deaf and dumb use no such laryngeal processes, nor does the individual
whose larynx has been removed. Other bodily processes have to take on the func
tion of the larynx. Such functions are usually usurped by the fngers, hands, arms,
facial muscles, muscles of the head, etc. I have in another place emphasized the
extent to which fnger and hand movements are used by the deaf and dumb when
they are engaged in silent thinking. . . . It would be an easy experiment, but so far
as I know not hitherto tried, to bind the fngers and arms of such an individual and
then give him a problem in arithmetic, memorizing simple stanzas, and the like,
which have to be worked out without exteroceptive aid. It would be necessary
probably to tie down eye movements, were such a thing possible, and to restrain
even the head and intercostal muscles.” (Watson, J. B. [1920] Is thinking merely the
action of language mechanisms? British Journal of Psychology, 11, 87–104.)
27. Watson, J. B. (1916) The psychology of wish fulfllment. The Scientifc Monthly,
quoted in Malone, J. C. (2014) Did John B. Watson really “found” behaviorism?
The Behavior Analyst, 37(1). Malone makes a persuasive case for Watson’s early sym
pathy with Freud’s views on the unconscious.
28. Skinner, B. F. (2104/1953) Science and Human Behavior, p. 193. The B. F. Skinner
Foundation and The Free Press.
2 Neo-Behaviorism and
Learning Psychology
1.5
Value
0.5
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Input Output
sensory response. Each sensory cell inhibits its two neighbors accord
ing to the simple equation:
Feeding Dynamics
Clark L. Hull
Watson was one of the frst to put rats through mazes, but the technique was
exploited most extensively by Yale psychologist Clark L. Hull (1884–1952)
and his followers, most notably his chief disciple the tough-minded Kenneth
Spence (1907–1967). Hull was a late developer in psychology, but in his
middle years, he became acquainted with the theory of physics, in its classi
cal, Newtonian form. No great mathematician himself, he was nevertheless
excited by the axiomatic method and pursued it doggedly. His major con
tribution was a book, The Principles of Behavior (1943), in which he proposed
a mathematical formulation to explain learning in the white rat.
Hull was a catalyst rather than a great teacher. His weekly seminars at
Yale became famous, and his ideas were spread by those who attended them
to many other universities across the US. But his most prominent follow
ers were supervised by others, and few of his own students made names in
academic psychology.
Experimental Method
Hull’s between-group experimental method, termed Null Hypothesis Statistical
Test (NHST), is still the norm in most of psychology. Groups of subjects
are exposed to diferent conditions. The individuals in control and experi
mental groups are selected at random to make the groups as similar as pos
sible. (Randomization is rarely a problem in the animal laboratory, but when
the subjects14 are human, real randomization can be difcult.) Control and
experimental groups are then compared.
For Hull and his co-workers, the subjects were rats, and the conditions
usually related to the amount type or probability of reward—rats running,
one at a time, down a straight runway to food pellets of diferent sizes, for
Neo-Behaviorism and Learning Psychology 25
example. The performances of the individuals (e.g., their running speeds,
learning rates) are then averaged and compared statistically to see if the
independent variable (e.g., pellet size) has an efect. The method, as a way
of understanding processes that take place in individuals rather than in the
group, rests on assumptions about the essential uniformity of the subjects,
and what they were doing, that were not understood for many years. Indeed,
the limitations of what has come to be termed between-group methodology
have not been fully assimilated by the psychological community (see Chap
ter 8). The limitations of the competing, within-subject, method are some
what better understood, as we will see shortly, but its advantages have yet to
be fully exploited.
“Frustration” Efect
Neo-behaviorists used groups even when the conditions being compared
could be studied with single animals. An example is the so-called frustration
efect (FE), which was demonstrated in 1952 with hungry rats running in a
“double runway,” that is, a runway with two goal boxes: G1, a short way
from the start box and, the second, G2, some distance after that: a short frst
runway followed by a much longer second one. In training, the rat runs to
G1 and gets a bit of food; then he runs to G2 and gets another bit of food.
The experimenter measures how fast the rat runs right after G1, in the frst
part of the long second runway.
The rat always gets food in G2. The experiment is in two phases. In the
frst phase, he gets food in G1 on every trial. Thus, he learns to expect food
in the frst goal box. In the second phase, he gets food in the frst goal box
on only half the trials: rewarded on half, “frustrated” on half. The question
is, How fast does he run in the long runway after food and after no-food, in
the frst goal box? The answer is that after training, he runs faster when there
is no food in the mid-goal box compared to when there is food, especially in
the frst third of the long runway. This elevated response is the “FE,” a sup
posedly excitatory efect of nonreward (i.e., absence of an expected reward).
Even though the FE only involves a comparison within the same individual,
the original experiment15 used no less than eighteen rats.
Why a group experiment? There are at least two reasons. One is that
group experiments were the norm at that time. Amsel and Roussel were
not the only ones to use many animals when one or two should have suf
fced. The second reason is that the efect was relatively small and probably
only showed up reliably in the group average. (More on the hazards of aver
aging later.)
The FE (now termed the omission efect) was replicated much later in an
experiment with a handful of pigeons using an operant procedure that pro
duced much larger single-subject efects than the rat-runway setup. Using
this method, it was possible to show conclusively that the FE was not an
energizing efect of nonreward, a conceptually problematic idea in any
26 History
case, but an inhibitory efect of reward. The FE represents disinhibition,
not excitation.16
Like all the early behaviorists, Hull was a philosophical descendent of the
British Empiricist philosophers, John Locke, Bishop Berkeley, and the great
Scot David Hume, and inherited their belief in associations as the primary
psychological units. Hull was, for a while, much impressed by logical positiv
ism, the position of the Vienna Circle17 of philosophers, and others. He had
earlier been infuenced by the formalism of Isaac Newton’s Principia Math
ematica (1687). The latter two infuences led him to prefer a mathematical,
highly formal approach to theory. Like Skinner and many other American
scientists in the Edison tradition, he was a gifted gadgeteer, but unlike Skin
ner, he admired abstract theory more than practical machinery.
It is helpful to organize the theoretical ideas of Hull and Tolman in terms
of three questions: What data did they seek to explain? Where did their
theoretical concepts come from? and What form did their theories take?
Clark Hull was interested in explaining data from experiments with groups
of rats learning simple tasks under the action of food reward and with groups
of people learning to recall lists of nonsense syllables (i.e., trigrams such as
JIV that do not correspond to actual words) under verbal instructions. In the
rat experiments, the data to be explained were not the moment-by-moment
behavior of individuals but “functional relations” based on data from groups
of subjects. Much was made, for example, of the negatively accelerated learn
ing curve that relates a measure of performance on a learning task to learning
trials. A measure such as the average percentage of correct responses in a
two-choice task typically increases rapidly at frst and then more slowly as it
approaches 100% (Box 1.1 shows an example of such a curve in two other
contexts). Another standard function was the delay-of-reinforcement gradient,
which showed how performance declines as the time between the rewarded
(reinforced) response and the reinforcement is increased.
Hull’s theoretical concepts came from an associationist view of the causal
chain between stimulus and response and a Watsonian notion of “ideas” as
manifestations of overt or covert motor activity. The “association of ideas”
seems to have come originally from Greek philosophy. Aristotle suggested
that ideas that occur close together in time are automatically linked in the
mind, as are things that are similar. Temporal proximity and similarity are
long-established principles of association.
For Hull, each “response” was characterized by a fnal “strength,” called
reaction potential, which was, in turn, determined by intervening variables (a
term from Tolman, about whom more in a moment) such as habit strength
and generalization. The variability characteristic of behavior was accommo
dated by the assumption of behavioral oscillation of reaction potential—what
would now be termed a stochastic (variable-in-time) process. Each of these
intervening variables was assumed to be linked in some way to observables.
Experimental results were deduced as formal predictions from this rather
lumbering system.
Neo-Behaviorism and Learning Psychology 27
Figure 2.2 Time relations between a conditioned stimulus (e.g., a tone, conditional
stimulus [CS]: light lines) and an unconditioned stimulus (e.g., food, uncon
ditioned stimulus [US]: heavy lines) in some Pavlovian conditioning pro
cedures. In delay conditioning, for example, after one of these pairings has
been repeated a sufcient number of times, the response reliably made to the
US (e.g., salivation the conditioned response, or CR) occurs when the CS
is presented. In trace conditioning, the CR usually occurs after the CS and
close to the time when the US will occur. In temporal conditioning, each US
presentation is the efective CS for the next US.
Generalization
The idea that training with one stimulus and response will increase the ten
dency to respond to similar stimuli and to make similar responses was another
idea from Pavlov. Other concepts came from Sherringtonian refexology
28 History
Figure 2.3 C. S. Sherrington (1856–1952) was an English physiologist whose book The
Integrative Action of the Nervous System (1906) summarized most of what was
then known about the refex and exerted a strong infuence on the evolution
of American psychology, especially behaviorism. But, like some other famous
neurophysiologists, he ended up as no behaviorist—writing, in his eighties, a
Cartesian, dualist philosophical rumination, Man on His Nature (1940).
(Figure 2.3), for example, the idea of reception, a process modeled on what
was then known of receptor physiology. The idea of interaction between
stimulus efects, which was added to accommodate “whole-greater-than
sum-of-parts” efects made famous by the German Gestalt (form) psycholo
gists.18 And fnally the idea of a stimulus trace that persists after the physical
stimulus has ceased.
How were these ideas put together to explain learning? A detailed exposi
tion would take me well beyond the confnes of a short survey—and would,
in any case, be of largely historical interest. But some of the basic ideas
can be conveyed simply. Hull tied reinforcement, the strengthening efect of
reward on a response that produces it, to the reduction of physiological need.
This turned out to be a hard assumption to prove: many rewarding events,
especially in humans (fame, money, art, fashion), have no obvious link to
physiology. Even food, the prototypical reinforcer, is efective as a rein
forcer long before it has any nutritional consequences. The rat may well
repeat his bar pressing after he gets the very frst food pellet, long before his
Neo-Behaviorism and Learning Psychology 29
blood sugar or any other gross physiological measure shows any change. So
much for physiological need. Moreover, the theoretical efects of reinforcing
events can be described without specifying their physiological concomitants.
The concept of conditioned or secondary reinforcement was invented to bridge
the gap between physiologically efective reinforcers and “neutral” stimuli
which acquire reinforcing properties, but this idea is still problematic.19
How did Hull conceive of the efects of reinforcement? One efect was to
“condition” events occurring at the time the “goal” (e.g., food at the end of
the maze) is reached. The basic idea was that any pair of events, A–B, that
regularly occur in succession somehow become linked (just as the CS and
the US become linked) so that the occurrence of A will come to call up
B. This is just a physiological interpretation of Aristotle’s notion of associa
tion through succession and temporal contiguity, interpreted as post hoc, ergo
propter hoc—B follows A; hence, A caused B. Thus, stimuli near the goal,
traces of earlier stimuli and stimuli generated by movements, or “fractional
anticipatory” movements (proprioception again) may become conditioned
to the “goal response.” In this way, the goal response will “move forward”
and occur earlier and earlier. This improvement in performance is the visible
evidence that learning has occurred.
Hull defned parts of this complex process in a quantitative way, but the
details were never fully specifed by Hull or by his several successors, most
of whom flled in theoretical gaps as best they could with verbal arguments.
Numerous faws in the system were pointed out,20 and Hull’s successors
increasingly abandoned quantitative modeling in favor of enthusiastic data
gathering. The mathematical approach was carried forward largely by math
ematical learning theorists.21 A single, very infuential theoretical paper by two
neo-Hullian Yale psychologists, Robert Rescorla and Allan Wagner,22 also
led to several elaborations by Wagner and others and something of a theoreti
cal revival. Rescorla and Wagner showed that temporal contiguity is not by
itself sufcient to produce conditioning: the CS must in some sense predict the
US—precede it reliably. Models in the Hullian tradition continue to appear
in the literature on classical conditioning (also known as associative learning).
Hull’s other direction was the study of verbal learning. The nonsense-
syllable method he used was pioneered by German experimental psycholo
gist Hermann Ebbinghaus (1850–1909), the founder of the feld of verbal
learning. Hull and several co-workers in 1940 put together an extraordinary
volume describing their theory of how people learn nonsense syllables. The
title of this book, Mathematico-deductive Theory of Rote Learning,23 shows its
allegiance to positivistic philosophy of science, and the theory was presented
in a highly formal way, using the calculus of symbolic logic (Fitch, one of
the book’s co-authors, was a logician). Hull defended the book as follows:
“Its chief value consists in the large-scale pioneering demonstration of the
logico-empirical methodology in the feld of behavior” (p. xi).
To a modern reader, the “logico-empirical methodology” with its rhet
oric of mathematical theorems and corollaries, makes the work almost
30 History
unreadable. Nevertheless, the assumptions of the theory are relatively sim
ple—and important, because these ingredients, in various combinations,
are part of most subsequent psychological theories. The theory explains
the patterns of learning of nonsense syllables in a serial anticipation pro
cedure: each syllable in a short list of 5 to 20 syllables is presented for a
couple of seconds, and the subject’s task is to anticipate the next syllable.
Data of interest are the patterns of errors—such as the serial-position efect,
which is that syllables at the beginning (primacy) and end (recency) of the
list are better remembered than syllables in the middle—and the efects of
the temporal spacing between list presentations (e.g., in comparisons of
spaced vs. massed practice).
The serial-position efect was discovered by Hermann Ebbinghaus in
1885. A modern version of his experiment looks like this:24 subjects are
shown a series of ffteen words, then tested for recall of the words imme
diately or after 30 seconds. When tested immediately, people remembered
items at the beginning and end of the series better than those in the middle.
This is the serial position efect. Memory for words at the end of the list
fades when the test is delayed for 30 seconds or more. As a methodological
footnote, although modern experiments always use groups and averaged
data, Ebbinghaus used just one subject—himself. His pioneering experi
ment would not be publishable in a contemporary psychological journal.
The reader can decide whether this refects poorly on Ebbinghaus or on
modern journal practices.
It is also worth noting that the serial-position efect suggests that remem
bered items interfere with one another so that recall of items in the middle
are recalled less frequently than items at the beginning and end of a list—just
like the lateral-inhibition process illustrated in Box 2.1. There is more to serial
recall than this, but interference of this sort is undoubtedly part of the process.
The chief theoretical ingredient in explaining memory efects is the
quasi-physiological idea of a memory trace, supposedly left by each trigram
as the subject attends to it. Thus, the recency efect fades after 30 seconds
because the trace of the last item fades to the point that there is interference
with the traces of earlier items, which are now, relatively, as almost recent
as the last item. This works because memory traces fade rapidly at frst and
then more slowly25 so that the traces of two successive items become more
similar as time elapses.
Traces are of two kinds, excitatory and inhibitory. “Response strength,”
the tendency to recall a particular syllable, is proportional to the algebraic
sum of these two tendencies. A correct anticipation is presumed to occur
when the summed strength exceeds a threshold, which is subject to some
intrinsic variability. Thus, as the trace strengths of two items approach each
other, confusion between them will increase. These ingredients, a decaying
trace, excitatory and inhibitory contributions to a strength variable, stochas
ticity, and the idea of response threshold, were to form a permanent part of
the theoretical toolbox used by later learning theorists.
Another random document with
no related content on Scribd:
sœur, j’ai souvent entendu dire que si elle manquait d’expression,
elle avait un brillant doigté.
Hervé pensa, à part soi, que reprocher à Mlle de Trivières de
manquer de sentiment musical était une grande injustice, et il écouta
de nouveau.
Cette fois, le chant d’une belle voix de contralto se mariait aux
sons du piano.
Il reconnut la plainte désolée de Fortunio :
« Mademoiselle Perrin,
« La présente est pour vous apprendre en douceur, comme il me
l’a bien recommandé, que votre filleul et ami Plisson Victor, de la
quatrième, vient d’être gravement blessé… »
— Oh ! mon Dieu ! Qu’est-ce que je disais ? Continuez, ma sœur.
— On lui a coupé la jambe hier, la gauche, ça s’est très bien
passé ; il avait reçu dedans un éclat d’obus dans le gras de la… »
Lisez, mademoiselle Rose, il y a un mot…
— « De la cuisse ! » Ah ! mon pauvre Victor ! Ensuite, ma sœur ?
Est-ce qu’il donne des détails ?