The Cambridge Handbook of

Thinking and Reasoning

Edited by
Keith J. Holyoak
and
Robert G. Morrison
 CHAPTER 1

Thinking and Reasoning:

A Reader’s Guide

Keith J. Holyoak
Robert G. Morrison

“Cogito, ergo sum,” the French philosopher What Is Thinking?

René Descartes famously declared, “I think,
therefore I am.” Every normal human adult We can start to answer this question by look-
shares a sense that the ability to think, to rea- ing at the various ways the word “think-
son, is a part of their fundamental identity. ing” is used in everyday language. “I think
A person may be struck blind or deaf, yet that water is necessary for life” and “George
we still recognize his or her core cognitive thinks the Pope is a communist” both ex-
capacities as intact. Even loss of language, press beliefs (of varying degrees of appar-
the gift often claimed as the sine qua non ent plausibility), that is, explicit claims of
of homo sapiens, does not take away a per- what someone takes to be a truth about the
son’s essential humanness. Unlike language world. “Anne is sure to think of a solution”
ability, which is essentially unique to our carries us into the realm of problem solv-
species, the rudimentary ability to think and ing, the mental construction of an action
reason is apparent in nonhuman primates plan to achieve a goal. The complaint “Why
(see Call & Tomasello, Chap. 25 ); and yet it didn’t you think before you went ahead with
is thinking, not language, that lies closest to your half-baked scheme?” emphasizes that
the core of our individual identity. A person thinking can be a kind of foresight, a way
who loses language but can still make intel- of “seeing” the possible future.1 “What do
ligent decisions, as demonstrated by actions, you think about it?” calls for a judgment,
is viewed as mentally intact. In contrast, the an assessment of the desirability of an op-
kinds of brain damage that rob an individ- tion. Then there’s “Albert is lost in thought,”
ual of the capacity to think and reason are where thinking becomes some sort of mental
considered the harshest blows that can be meadow through which a person might me-
struck against a sense of personhood. Cogito, ander on a rainy afternoon, oblivious to the
ergo sum. world outside.

1
2 the cambridge handbook of thinking and reasoning

Rips and Conrad (1 989) elicited judg- state of affairs, some goal that motivates the
ments from college students about how var- thinker to perform mental work.
ious mentalistic terms relate to one another. Our definition thus includes quite a few
Using statistical techniques, the investigators stipulations, but notice also what is left out.
were able to summarize these relationships We do not claim that thinking necessarily
in two diagrams, shown in Figure 1 .1 . Fig- requires a human (higher-order primates,
ure 1 .1 (A) is a hierarchy of kinds, or cat- and perhaps some other species on this or
egories. Roughly, people believe planning other planets, have a claim to be considered
is a kind of deciding, which is a kind of thinkers) (see Call & Tomasello, Chap. 25 )
reasoning, which is a kind of conceptual- or even a sentient being. (The field of ar-
izing, which is a kind of thinking. People tificial intelligence may have been a disap-
also believe that thinking is part of con- pointment in its first half-century, but we
ceptualizing, which is part of remembering, are reluctant to define it away as an oxy-
which is part of reasoning, and so on [Fig- moron.) Nonetheless, our focus in this book
ure 1 .1 (B)]. The kinds ordering and the parts is on thinking by hominids with electro-
ordering are similar; most strikingly, “think- chemically powered brains. Thinking often
ing” is the most general term in both order- seems to be a conscious activity of which
ings – the grand superordinate of mental ac- the thinker is aware (cogito, ergo sum); how-
tivities, which permeates all the others. ever, consciousness is a thorny philosophi-
It is not easy to make the move from the cal puzzle, and some mental activities seem
free flow of everyday speech to scientific def- pretty much like thinking, except for being
initions of mental terms, but let us nonethe- implicit rather than explicit (see Litman &
less offer a preliminary definition of thinking Reber, Chap. 1 8). Finally, we do not claim
to suggest what this book is about: Thinking that thinking is inherently rational, optimal,
is the systematic transformation of mental rep- desirable, or even smart. A thorough history
resentations of knowledge to characterize ac- of human thinking will include quite a few
tual or possible states of the world, often in chapters on stupidity.
service of goals. Obviously, our definition in- The study of thinking includes several in-
troduces a plethora of terms with meanings terrelated subfields that reflect slightly dif-
that beg to be unpacked, but at which we can ferent perspectives on thinking. Reasoning,
only hint. A mental representation of knowl- which has a long tradition that springs from
edge is an internal description that can be philosophy and logic, places emphasis on the
manipulated to form other descriptions. To process of drawing inferences (conclusions)
count as thinking, the manipulations must from some initial information (premises). In
be systematic transformations governed by standard logic, an inference is deductive if the
certain constraints. Whether a logical deduc- truth of the premises guarantees the truth
tion or a creative leap, what we mean by of the conclusion by virtue of the argument
thinking is more than unconstrained associ- form. If the truth of the premises renders the
ations (with the caveat that thinking may in- truth of the conclusion more credible but
deed be disordered; see Bachman & Cannon, does not bestow certainty, the inference is
Chap. 21 ). The internal representations cre- called inductive.2 Judgment and decision mak-
ated by thinking describe states of some ex- ing involve assessment of the value of an
ternal world (a world that may include the option or the probability that it will yield
thinker as an object of self-reflection) – that a certain payoff ( judgment) coupled with
world might be our everyday one, or per- choice among alternatives (decision mak-
haps some imaginary construction obeying ing). Problem solving involves the construc-
the “laws” of magical realism. Often (not tion of a course of action that can achieve a
always – the daydreamer, and indeed the goal.
night dreamer, are also thinkers), thinking Although these distinct perspectives on
is directed toward achieving some desired thinking are useful in organizing the field
 thinking and reasoning: a reader’s guide 3

Figure 1 .1 . People’s conceptions of the relationships among terms for mental activities. A, Ordering
of “kinds.” B, Ordering of “parts.” (Adapted from Rips & Conrad, 1 989, with permission.)

(and this volume), these aspects of thinking eighteenth-century philosophers Immanuel

overlap in every conceivable way. To solve Kant (in Germany) and David Hume (in
a problem, one is likely to reason about the Scotland) laid the foundations for all subse-
consequences of possible actions and make quent work on the origins of causal knowl-
decisions to select among alternative actions. edge, perhaps the most central problem in
A logic problem, as the name implies, is a the study of thinking (see Buehner & Cheng,
problem to be solved (with the goal of de- Chap. 7). If we were to choose one phrase
riving or evaluating a possible conclusion). to set the stage for modern views of think-
Making a decision is often a problem that ing, it would be an observation of the British
requires reasoning. These subdivisions of the philosopher Thomas Hobbes, who, in 1 65 1 ,
field, like our preliminary definition of think- in his treatise Leviathan, proposed, “Rea-
ing, should be treated as guideposts, not soning is but reckoning.” “Reckoning” is an
destinations. odd term today, but in the seventeenth cen-
tury it meant computation, as in arithmetic
calculations.3
It was not until the twentieth century that
A Capsule History the psychology of thinking became a scien-
tific endeavor. The first half of the century
Thinking and reasoning, long the academic gave rise to many important pioneers who
province of philosophy, have over the past in very different ways laid the foundations
century emerged as core topics of empirical for the emergence of the modern field of
investigation and theoretical analysis in the thinking and reasoning. Foremost were the
modern fields known as cognitive psychol- Gestalt psychologists of Germany, who pro-
ogy, cognitive science, and cognitive neuro- vided deep insights into the nature of prob-
science. Before psychology was founded, the lem solving (see Novick & Bassok, Chap. 1 4).
4 the cambridge handbook of thinking and reasoning

Most notable of the Gestaltists were Karl predictions (much like an engineer might
Duncker and Max Wertheimer, students use a physical scale model of a bridge to
of human problem solving, and Wolfgang anticipate the effects of stress on the ac-
Köhler, a keen observer of problem solv- tual bridge intended to span a river).4 In the
ing by great apes (see Call & Tomasello, 1 960s and 1 970s, modern work on the psy-
Chap. 25 ). chology of reasoning began in Britain with
The pioneers of the early twentieth cen- the contributions of Peter Wason and his col-
tury also include Sigmund Freud, whose laborator Philip Johnson-Laird (see Evans,
complex and ever-controversial legacy in- Chap. 8).
cludes the notions that forms of thought The modern conception of thinking as
can be unconscious (see Litman & Reber, computation became prominent in the
Chap. 1 8) and that “cold” cognition is tan- 1 970s. In their classic treatment of human
gled up with “hot” emotion (see Molden & problem solving, Allen Newell and Herbert
Higgins, Chap. 1 3 ). As the founder of clini- Simon (1 972) showed that the computa-
cal psychology, Freud’s legacy also includes tional analysis of thinking (anticipated by
the ongoing integration of research on nor- Alan Turing, the father of computer science)
mal thinking with studies of thought disor- could yield important empirical and theo-
ders, such as schizophrenia (see Bachman & retical results. Like a program running on a
Cannon, Chap. 21 ). digital computer, a person thinking through
Other early pioneers in the early and a problem can be viewed as taking an in-
mid-twentieth century contributed to vari- put that represents initial conditions and a
ous fields of study that are now embraced goal, and applying a sequence of operations
within thinking and reasoning. Cognitive de- to reduce the difference between the initial
velopment continues to be influenced by the conditions and the goal. The work of Newell
early theories developed by the Swiss psy- and Simon established computer simulation
chologist Jean Piaget (see Halford, Chap. 22) as a standard method for analyzing human
and the Russian psychologist Lev Vygotsky thinking. Their work also highlighted the po-
(see Greenfield, Chap. 27). In the United tential of production systems (see Novick &
States, Charles Spearman was a leader in the Bassok, Chap. 1 4), which were subsequently
systematic study of individual differences in developed extensively as cognitive models
intelligence (see Sternberg, Chap. 3 1 ). In the by John Anderson and his colleagues (see
middle of the century, the Russian neurolo- Lovett & Anderson, Chap. 1 7).
gist Alexander Luria made immense contri- The 1 970s saw a wide range of major de-
butions to our understanding of how think- velopments that continue to shape the field.
ing depends on specific areas of the brain, Eleanor Rosch, building on earlier work by
anticipating the modern field of cognitive Jerome Bruner (Bruner, Goodnow, & Austin,
neuroscience (see Goel, Chap. 20). Around 1 95 6), addressed the fundamental question
the same time, in the United States, Herbert of why people have the categories they do,
Simon argued that the traditional rational and not other logically possible groupings of
model of economic theory should be re- objects (see Medin & Rips, Chap. 3 ). Rosch
placed with a framework that accounted for argued that natural categories often have
a variety of human resource constraints such fuzzy boundaries (a whale is an odd mam-
as bounded attention and memory capac- mal) but nonetheless have clear central ten-
ity and limited time (see LeBoeuf & Shafir, dencies or prototypes (people by and large
Chap. 1 1 , and Morrison, Chap. 1 9). This was agree that a bear makes a fine mammal).
one of the contributions that in 1 978 earned The psychology of human judgment was re-
Simon the Nobel Prize in Economics. shaped by the insights of Amos Tversky and
In 1 943 , the British psychologist Kenneth Daniel Kahneman, who identified simple
Craik sketched the fundamental notion that cognitive strategies, or heuristics, that people
a mental representation provides a kind of use to make judgments of frequency and
model of the world that can be “run” to make probability. Often quick and accurate, these
 thinking and reasoning: a reader’s guide 5

strategies can in some circumstances lead might be performed using numbers in ei-
to nonnormative judgments. After Tversky’s ther decimal or binary code, starting with
death in 1 996, this line of work was con- either the leftmost or rightmost digit. Fi-
tinued by Kahneman, who was awarded the nally, the level of implementation addresses
Nobel Prize in Economics in 2002. The cur- the question, “How are the representation
rent view of judgment, which has emerged and algorithm realized physically?” The cash
from 3 0 years of research, is summarized by register could be implemented as an elec-
Kahneman and Frederick (Chap. 1 2; also see tronic calculator, a mechanical adding ma-
LeBoeuf & Shafir, Chap. 1 1 ). (Goldstone and chine, or even a mental abacus in the mind of
Son, Chap. 2, review Tversky’s influential the clerk.
theory of similarity judgments.) In his book, Marr stressed the importance
In 1 982, a young vision scientist, David of the computational level of analysis, ar-
Marr, published a book called Vision. Largely guing that it could be seriously misleading
a technical treatment of visual perception, to focus prematurely on the more concrete
the book includes an opening chapter that levels of analysis for a cognitive task with-
lays out a larger vision – a vision of how out understanding the goal or nature of the
the science of mind should proceed. Marr mental computation.5 Sadly, Marr died of
distinguished three levels of analysis, which leukemia before Vision was published, and
he termed the level of computation, the level so we do not know how his thinking about
of representation and algorithm, and the level levels of analysis might have evolved. In
of implementation. Each level, according to very different ways, Marr’s conception of a
Marr, addresses different questions, which computational level of analysis is reflected
he illustrated with the example of a phys- in several chapters in this book (see espe-
ical device, the cash register. At Marr’s most cially Doumas & Hummel, Chap. 4; Buehner
abstract level, computation (not to be con- & Cheng, Chap. 7; Lovett & Anderson,
fused with computation of an algorithm on a Chap. 1 7).
computer), the basic questions are “What is In the most recent quarter-century, many
the goal that the cognitive process is meant other springs of research have fed into the
to accomplish?” and “What is the logic of the river of thinking and reasoning, including
mapping from the input to the output the field of analogy (see Holyoak, Chap. 6),
that distinguishes this mapping from other neural network models (see Doumas &
input–output mappings?” A cash register, Hummel, Chap. 4; Halford, Chap. 22), and
viewed at this level, is used to achieve the cognitive neuroscience (see Goel, Chap. 20).
goal of calculating how much is owed for a The chapters of this handbook collectively
purchase. This task maps precisely onto the paint a picture of the state of the field at the
axioms of addition (e.g., the amount owed dawn of the new millennium.
should not vary with the order in which
items are presented to the sales clerk, a
constraint that precisely matches the com-
mutativity property of addition). It follows Overview of the Handbook
that, without knowing anything else about
the workings of a particular cash register, This volume brings together the contribu-
we can be sure (if it is working prop- tions of many of the leading researchers
erly) that it will be performing addition in thinking and reasoning to create the
(not division). most comprehensive overview of research
The level of representation and algo- on thinking and reasoning that has ever been
rithm, as the name implies, deals with the available. Each chapter includes a bit of his-
questions, “What is the representation of torical perspective on the topic and ends
the input and output?” and “What is the with some thoughts about where the field
algorithm for transforming the former into seems to be heading. The book is organized
the latter?” Within a cash register, addition into seven sections.
6 the cambridge handbook of thinking and reasoning

Part I: The Nature of Human Concepts Part III: Judgment and Decision Making
The three chapters in Part I address foun- We then turn to topics related to judgment
dational issues related to the representation and decision making. In Chapter 1 1 , LeBoeuf
of human concepts. Chapter 2 by Gold- and Shafir set the stage with a general re-
stone and Son reviews work on the core view of work on decision making. Then,
concept of similarity – how people assess in Chapter 1 2, Kahneman and Frederick
the degree to which objects or events are present an overarching model of heuristic
alike. Chapter 3 by Medin and Rips consid- judgment. In Chapter 1 3 , Molden and Hig-
ers research on categories and how concepts gins review research revealing the ways in
are organized in semantic memory. Think- which human motivation and emotion influ-
ing depends not only on representations of ence judgment.
individual concepts, such as dogs and cats,
but also on representations of the relation- Part IV: Problem Solving
ships among concepts, such as the fact that and Complex Learning
dogs often chase cats. In Chapter 4, Doumas
and Hummel evaluate different compu- The five chapters that comprise this section
tational approaches to the representation deal with problem solving and allied issues
of relations. concerning how people learn in problem-
solving situations. In Chapter 1 4, Novick
and Bassok provide a general overview of
Part II: Reasoning the field of human problem solving. Prob-
lem solving has close connections to the
Chapters 5 to 1 0 deal with varieties of topic of creativity, the focus of Chapter 1 5
the core topic of reasoning. In Chapter 5 , by Sternberg, Lubart, Kaufman, and Pretz.
Sloman and Lagnado set the stage by lay- Beyond relatively routine problem solving,
ing out the issues surrounding induction – there are occasions when people need to re-
using what is known to generate plausi- structure their knowledge in complex ways
ble, although uncertain, inferences. Then, to generate deeper understanding. How such
in Chapter 6, Holyoak reviews the liter- complex learning takes place is the topic of
ature on reasoning by analogy, an impor- Chapter 1 6 by Chi and Ohlsson. In Chap-
tant variety of inductive reasoning that is ter 1 7, Lovett and Anderson review work
critical for learning. The most classic as- on thinking that is based on a particular
pect of induction is the way in which hu- formal approach rooted in work on prob-
mans and other creatures acquire knowledge lem solving, namely, production systems.
about causal relations, which is critical for Finally, in Chapter 1 8, Litman and Reber
predicting the consequences of actions and consider research suggesting that some as-
events. In Chapter 7, Buehner and Cheng pects of thinking and learning depend on im-
discuss research and theory on causal learn- plicit mechanisms that operate largely out-
ing. Then, in Chapter 8, Evans reviews work side of awareness.
on the psychology of deductive reasoning,
the form of thinking with the closest ties
Part V: Cognitive and Neural Constraints
to logic. In Chapter 9, Johnson-Laird de-
on Human Thought
scribes the work that he and others have
performed using the framework of men- High-level human thinking cannot be fully
tal models to deal with various reasoning understood in isolation from fundamental
tasks, both deductive and inductive. Men- cognitive processes and their neural sub-
tal models have close connections to percep- strates. In Chapter 1 9, Morrison reviews the
tual representations that are visuospatial in wealth of evidence indicating that thinking
Chapter 1 0, Barbara Tversky reviews work and reasoning depend critically on what is
on the role of visuospatial representations known as “working memory,” that is, the sys-
in thinking. tem responsible for short-term maintenance
 thinking and reasoning: a reader’s guide 7

and manipulation of information. Current ferences in the nature and quality of hu-
work is making headway in linking thought man thinking. This section includes three
processes to specific brain structures such as chapters focusing on thinking in particu-
the prefrontal cortex; in Chapter 20, Goel lar practices and two chapters that deal
discusses the key topic of deductive reason- with variations in thinking ability. In Chap-
ing in relation to its neural substrate. Brain ter 28, Ellsworth reviews what is known
disorders, notably schizophrenia, produce about thinking in the field of law. In Chap-
striking disruptions of normal thought pro- ter 29, Dunbar and Fugelsang discuss think-
cesses, which can shed light on how thinking ing and reasoning as manifested in the prac-
takes place in normal brains. In Chapter 21 , tice of science. In Chapter 3 0, Patel, Arocha,
Bachman and Cannon review research and and Zhang discuss reasoning in a field –
theory concerning thought disorder. medicine – in which accurate diagnosis and
treatment are literally everyday matters of
Part VI: Ontogeny, Phylogeny, Language, life and death. Then, in Chapter 3 1 , Stern-
and Culture berg reviews work on the concept of intel-
ligence as a source of individual differences
Our understanding of thinking and reason-
in thinking and reasoning. Finally, Chapter
ing would be gravely limited if we restricted
3 2 by Ritchhart and Perkins concludes the
investigation to young adult English speak-
volume by reviewing one of the major chal-
ers. The six chapters in Part VI deal with the
lenges for education – finding ways to teach
multifaceted ways in which aspects of think-
people to think more effectively.
ing vary across the human lifespan, across
species, across speakers of different lan-
guages, and across cultures. In Chapter 22,
Halford provides an overview of the devel- Examples of Chapter Assignments
opment of thinking and reasoning over the for a Variety of Courses
course of childhood. In Chapter 23 , Gallistel
and Gelman discuss mathematical thinking, This volume offers a comprehensive treat-
a special form of thinking found in rudi- ment of higher cognition. As such, it serves
mentary form in nonhuman animals that un- as an excellent source for courses on think-
dergoes development in children. In Chap- ing and reasoning, both at the graduate
ter 24, Salthouse describes the changes in level and for upper-level undergraduates. Al-
thinking and reasoning brought on by the though instructors for semester-length grad-
aging process. The phylogeny of thinking – uate courses in thinking and reasoning may
thinking and reasoning as performed by apes opt to assign the entire volume as a text-
and monkeys – is discussed in Chapter 25 by book, there are a number of other possibili-
Call and Tomasello. One of the most contro- ties (including using chapters from this vol-
versial topics in the field is the relationship ume as introductions for various topics and
between thinking and the language spoken then supplementing with readings from the
by the thinker; in Chapter 26, Gleitman primary literature). Here are a few examples
and Papafragou review the hypotheses and of possible chapter groupings tailored to a
evidence concerning the connections be- variety of possible course offerings:
tween language and thought. In Chapter 27,
Greenfield considers the ways in which
Introduction to Thinking and Reasoning
modes of thinking may vary in the context
of different human cultures. Chapter 1 Thinking and Reasoning: A
Reader’s Guide
Part VII: Thinking in Practice Chapter 2 Similarity
In cultures ancient and modern, thinking Chapter 3 Concepts and Categories:
is put to particular use in special cultural Memory, Meaning, and
practices. Moreover, there are individual dif- Metaphysics
8 the cambridge handbook of thinking and reasoning

Chapter 5 The Problem of Induction Chapter 7 Causal Learning

Chapter 6 Analogy Chapter 9 Mental Models and Thought
Chapter 7 Causal Learning Chapter 2 2 Development of Thinking
Chapter 8 Deductive Reasoning Chapter 1 7 Thinking as a Production
Chapter 9 Mental Models and Thought System
Chapter 1 0 Visuospatial Reasoning
Chapter 1 1 Decision Making Applied Thought
Chapter 1 2 A Model of Heuristic
Judgment Chapter 1 4 Problem Solving
Chapter 1 4 Problem Solving Chapter 1 0 Visuospatial Reasoning
Chapter 1 5 Creativity Chapter 2 3 Mathematical Thinking
Chapter 1 6 Complex Declarative Chapter 2 6 Language and Thought
Learning Chapter 1 5 Creativity
Chapter 1 8 Implicit Cognition and Chapter 3 1 Intelligence
Thought Chapter 1 3 Motivated Thinking
Chapter 2 7 Paradigms of Cultural
Thought
Development of Thinking
Chapter 1 6 Complex Declarative
Chapter 2 Similarity Learning
Chapter 3 Concepts and Categories: Chapter 1 8 Implicit Cognition and
Memory, Meaning, and Thought
Metaphysics Chapter 2 8 Legal Reasoning
Chapter 2 2 Development of Thinking Chapter 2 9 Scientific Thinking and
Chapter 2 3 Mathematical Thinking Reasoning
Chapter 2 6 Language and Thought Chapter 3 0 Reasoning in Medicine
Chapter 2 4 Effects of Aging on Reasoning
Chapter 2 5 Reasoning and Thinking in Differences in Thought
Nonhuman Primates
Chapter 1 9 Thinking in Working Memory Chapter 3 1 Intelligence
Chapter 3 1 Intelligence Chapter 1 5 Creativity
Chapter 3 2 Learning to Think: The Chapter 1 9 Thinking in Working Memory
Challenges of Teaching Chapter 2 1 Cognitive and Neuroscience
Thinking Aspects of Thought Disorder
Chapter 2 2 Development of Thinking
Chapter 2 5 Reasoning and Thinking in
Modeling Human Thought Nonhuman Primates
Chapter 2 Similarity Chapter 2 4 Effects of Aging on Reasoning
Chapter 3 Concepts and Categories: Chapter 2 6 Language and Thought
Memory, Meaning, and Chapter 1 3 Motivated Thinking
Metaphysics Chapter 2 7 Paradigms of Cultural
Chapter 4 Approaches to Modeling Thought
Human Mental Chapter 2 9 Scientific Thinking and
Representations: Reasoning
What Works, What Doesn’t, Chapter 3 2 Learning to Think: The
and Why Challenges of Teaching
Chapter 6 Analogy Thinking
 thinking and reasoning: a reader’s guide 9

Acknowledgments tive reasoning.” In an old Western movie, a

hero in a tough spot might venture, “I reckon
Preparation of this chapter was supported we can hold out till sun-up,” illustrating how
calculation has crossed over to become a
by grants R3 05 H03 01 41 from the Institute
metaphor for mental judgment.
of Education Sciences and SES-00803 75
4. See Johnson-Laird, Chap. 9, for a current view
from the National Science Foundation to
of thinking and reasoning that owes much to
Holyoak, and by Xunesis (www.xunesis.org) Craik’s seminal ideas.
and a National Institute of Mental Health
5 . Indeed, Marr criticized Newell and Simon’s
National Service Research Award (MH- approach to problem solving for paying insuf-
064244) to Morrison. The authors thank ficient attention to the computational level in
Miriam Bassok and Patricia Cheng for his sense.
comments on an earlier draft of this chapter.

References
Notes
Bruner, J. S., Goodnow, J. J., & Austin, G. A.
(1 95 6). A study of thinking. New York: Wiley.
1 . Notice the linguistic connection between
“thinking” and “seeing,” and thought and per- Craik, K. (1 943 ). The nature of explanation. Cam-
ception, which was emphasized by the Gestalt bridge, UK: Cambridge University Press.
psychologists of the early twentieth century. Hobbes, T. (1 65 1 /1 968). Leviathan. London: Pen-
2. The distinction between deduction and in- guin Books.
duction blurs in the study of the psychol- Marr, D. (1 982). Vision. San Francisco: W. H.
ogy of thinking, as we see in Part II of Freeman.
this volume. Newell, A., & Simon, H. A. (1 972). Hu-
3 . There are echoes of the old meaning of man problem solving. Englewood Cliffs, NJ:
“reckon” in such phrases as “reckon the cost.” Prentice Hall.
As a further aside, the term “dead reckon- Rips, L. J., & Conrad, F. G. (1 989). Folk psychol-
ing,” a procedure for calculating the position ogy of mental activities. Psychological Review,
of a ship or aircraft, derives from “deduc- 96, 1 87–207.