Download Chapter 7 Swimming Against the Stream. Herbert Simon, Harvey

Chapter 7
Swimming Against the Stream. Herbert Simon,
Harvey Leibenstein, George Shackle, Friedrich von
Hayek
7.1 Introduction
This chapter focuses on four economists. It was initially designed to focus
on Simon and Leibenstein. As the book progressed I discovered that Shackle
and Hayek were relevant to the main thesis of the book. One of the things
which bind these four economists is that each were swimming against the
neo-classical stream. And, of course, they each have “two minds.”
7.2 Herbert Simon.
“Administration man recognizes that the world he perceives is a drastically
simplified model of the buzzing, blooming confusion that constitutes the real world”
(Herbert Simon, 1966, p. xxv).
Simon received a Ph.D. in Political Science from the University of
Chicago, the same University where Frank Knight taught and William James
had a large influence. This may explain Simon’s use of a phrase which uses
the words buzzing and blooming, the reverse order used by the psychologistphilosopher James. After Chicago, Simon was instrumental in the
development of Artificial Intelligence, developed the Graduate School of
Industrial Administration at Carnegie Tech, and was a Professor of Computer
Science and Psychology at Carnegie Mellon. In 1978 he won the Nobel Prize
in Economics, “for his pioneering research into the decision-making process
within economic organizations.” Some economists were outraged by his
winning the Nobel Prize. Yet his research (agenda) was prescient, because
among other things it was ‘first generation’ behavioral economics (Camerer
and Lowenstein, 2004).
Simon’s work was motivated by the belief that neither the human mind,
human thinking and decision making, nor human creativity need be
mysterious. His life work was devoted to proving this point. His motto was
"Wonderful, but not incomprehensible" (Simon, 1969, p. 4). Where he carried
out his motto was at the intersection of economics, psychology, cognitive
science, and organization theory. A major part of this intersection was
creating computer programs which allow machines to “think” and make
choices. It was after he helped create “thinking” machines that Simon came to
understand human intuition as subconscious pattern recognition. In doing so
he showed that intuition need not be associated with magic and mysticism. He
114
Two Minds
also showed that "intuition is not a process that operates independently of
analysis; rather the two processes are essential complementary components of
effective decision-making systems" (Simon and Gilmartin, 1973, p. 33).
7.2.1 Intuition. The "Problem" Illustrated
Simon’s philosophy of intuition may be said to begin with the publication
of Chester Barnard’s The Functions of the Executive (Barnard, 1942),
specifically an Appendix titled "Mind in Everyday Affairs" in which Barnard
discusses intuition. Herbert Simon, in his lecture given in Stockholm upon
receiving the Nobel Prize in Economics, referred to Barnard as an
"intellectually curious business executive who distilled from his experience as
president of New Jersey Bell Telephone Company…a profound book on
decision making…” (Simon, 1965, p. 25). Yet, when it came to Barnard's
philosophy of intuition, Simon says that Barnard, "presents an interesting, but
perhaps too optimistic view of the 'intuitive' elements in administrative
decisions…" (ibid, p. 60). What makes Barnard’s presentation too optimistic
for Simon?
Barnard's philosophy about intuition was stated matter-of-factly. First,
intuition may seem abstract because it arises from the subconscious, but it's
not abstract. Intuition is a non-logical process, defined by Barnard as a
process which takes place in the subconscious or is so rapid as to seem
subconscious, and hence also seems to be instantaneous and devoid of
reasoning. Examples of intuition cited by Barnard include studying a complex
balance sheet for only a few minutes or seconds before being able to derive a
coherent picture of the company.
Second, intuition is as much an expression of intelligence as is logic.
Third, intuition is useful and so people should use it. Fourth, many people use
intuition at work but are in the ‘intuitive closet.’ The reason is that many
people feel the need to rationalize their beliefs and have them appear
plausible, and hence do not want to admit using something that is
unexplainable. At the same time, many people insist on using logic but are not
skilled at applying logic in actual decisions. Fifth, intuition is most
appropriate when working with short time horizons and data which is either of
poor quality and/or very limited. This covers according to Barnard the
majority of situations used in every day affairs in both business and
government. In most situations, therefore, ‘women’s intuition’ is the only
mental process available. Understanding organizations also calls for intuition.
He says, "Our logical methods and our endless analysis of things has often
blinded us to an appreciation of structure and organization… You cannot get
organization by adding up the parts…To understand the society you live in,
you must feel organization -- which is exactly what you do with your nonlogical minds…" (Barnard, 1942, p. 317).
Swimming Against the Stream
115
Simon’s approach was very different. As a scientist he needed to
understand the phenomena we call intuition. In the early part of his career
there wasn’t a scientific – rational, logical -- theory of intuition, and so Simon
considered intuition to be a mystery. In time and with advances in cognitive
science and AI as a framework, Simon concluded that intuition is
subconscious pattern recognition. Simon did not consider intuition to be
irrational, he considered it to be non-rational, meaning it is not a conscious
analytical method of decision making (Simon, 1987). Simon had his own
intuitive experiences. His preliminary exam in statistics for the University of
Chicago in 1940) required that he produce a derivation of the chi-square
statistic: he provided two derivations. He says, “While taking my shower on
the morning of that exam, it came to me with blinding and unaccountable
certainty that there would be a question on chi-square, and I boned up on it
before setting out for the exam room. On no other occasion have I had such
loving attention from my guardian angel” (Simon, 1996, p. 84). As a second
example, Simon had been contemplating programming computers to simulate
chess playing since the early 1950’s. In 1955 during a walk on the campus of
Columbia University he says, “Suddenly, I had a clear conviction that we
could program a machine to solve such problems (p. 203). Third, speaking
about government employees specializing in security he said, “Intuitively,
they know that intellectuals seek to be loyal to abstractions like ‘truth,’
‘virtue,’ or ‘freedom,’ rather than to a national state or its flag” (Simon,
1989a, p.133).
7.2.2 Rationality
For Simon, problem solving was a “search through a vast maze of
possibilities, a maze that describes the environment” (Simon, 1982h, p. 66).
Rationality is bounded BR or limited by the vast maze of possibilities which
is our environment. The maze makes the procedures we use in decision
making, one of which is subconscious pattern recognition, more important
than traditionally given in economics. And it means that the decisions we
make are more satisficing than maximizing. Thus, Simon challenged the
economic orthodoxy on the definition of rationality by proposing the concepts
of bounded and procedural rationality, and satisficing.
The orthodox definition is represented by Economic Person (EP) who is a
substantively rational maximizer of subjective expected utility. Substantive
rationality occurs when behavior is appropriate to attain a given goal, under
given conditions (constraints). In other words, substantive rationality is about
outcomes. EP is substantively-rational because she is assumed to have at least
a sufficient amount of information about all relevant aspects of their
environment, the ability to compute benefits and costs of available alternative
courses of action, information about the probability of each outcome of each
116
Two Minds
chosen behavior, and a willingness and ability to understand and consider
simultaneously all current and future available alternatives. Assuming that
individuals are substantially rational, and that they have a definite goal,
economics can be “done” with calculus. And, it can be done without
psychology. But, Simon says, "…there is a complete lack of evidence that, in
actual human choice situations of any complexity, these computations can be,
or are in fact, performed" (Simon, 1982a, p. 244).
Simon’s ideas on bounded-rationality (BR) were initially contained in the
first (1947) and subsequent editions of Administrative Behavior (Simon,
1965), and in more formal models published in 1955 in the Quarterly Journal
of Economics (Simon, 1982a), and in 1956 in the Psychological Review
(Simon, 1982f). In 1956 Simon wrote a short story as an attempt at a
“transmigration of the soul” of his model of bounded rationality. The story,
“The Apple: A Story of the Maze” (Simon, 1989a), is about a young man
named Hugo who lived alone in a large castle. Hugo’s problem was that he
had to find food, which, mysteriously, was being left on the tables in some of
the rooms. Some of the food left was of a variety he had never seen, so he had
to discover his tastes and preferences for food. He also needed to save time at
finding the right room so he began looking for clues to tell him which room
had which variety of food. For example the rooms containing his favorite
foods had various paintings on the wall. His preference for certain paintings
had developed un-consciously as an association with certain preferred foods.
With experience, finding his preferred foods became easier. Hugo’s
search for food did not continue until he found his favorite food. Hugo was
often hungry and he didn’t know when he would find food again, so his
searches ended when he found food which was satisfactory. His experience
showed him that finding the right food depended upon the number of turns or
choice points in the maze, the number of available paths at each choice point,
the number of moves or steps between choice points, the number of moves
and choice points the individual can see ahead, and the ability of the
individual to find clues (patterns) in order to avoid walking around in circles.
The hungrier he was, the greater the number of food types and groups fell
under the category of satisfactory. Hugo was a satisficer whose rationality
was bounded by the shape of his world and the circumstances of his life.
Associations or patterns were stored in memory and retrieved automatically
from memory when it served to satisfy his goal. Hugo engaged in
subconscious pattern recognition, that is, he engaged his intuition. Substantive
rationality and global maximization in a maze is possible, but only when the
maze is very small.
Swimming Against the Stream
117
7.2.3 Substantive vs Bounded and Procedural Rationality
In contrast to EP, "Bounded Rational Person" (BRP) lives in a world
which offers a set of objectively available behavior alternatives, but a more
limited set of "perceived" behavior alternatives. BRP lacks both the
information and the computational capacities to be globally rational. RWP
lives in a world with too much uncertainty -- unintended con-sequences, and
computational limits. Given our computational limits the environment that we
are aware of is only a fraction of the "real" environment within which
decisions are made. The number of possible alternatives in so immense that
they can't be examined. The best and only feasible solution is to find a
satisfactory solution. In psychology it is aspiration levels which perform this
function. And "problem solving and decision making that sets an aspiration
level, searches until an alternative is found that is satisfactory by the
aspiration level criterion, and selects that alternative" (Simon, 1982j, p. 415)
is called satisficing. The concept of satisficing is common within psychology
(Simon, 1982c). It is part of a model of behavior in which the motivation to
act comes from "drives," and the termination of action occurs when the drive
is satisfied. The definition of drive-satisfaction varies upon aspirations and
experience. Satisficing in economics is assumed to be less important because
standard economics assumes that individuals are objectively or globally
substantively rational expected utility maximizers. Of course these
assumptions amount to nothing more than assuming away the importance of
satisficing.
However, in the real world inhabited by BRP perception and cognition do
not merely passively filter only a small part of the entire environment into our
consciousness. Of the entire amount of new information generated by our
entire environment, our senses filter out 99%+ before it reaches our
consciousness. Given these facts, human behavior is in most cases restricted
to satisficing behavior. Rationality is thus bounded by the complexity of the
world we live in relative to our cognitive abilities. It seems intuitively obvious
to the casual observer that BR is more descriptive of the way people with
modest computational abilities make decisions, stay alive, and even thrive
(Simon, 1983). Intuitive rationality (IR) is a subset of bounded rationality, but
more about this later.
Because rationality is bounded the process we use in making the best
decisions we can becomes more important. Simon thus considers behavior to
be rational when it is the outcome of an appropriate deliberation process.
Behavior is procedurally rational when it is the outcome of an appropriate
deliberation process. Behavior is procedurally irrational when it is the
outcome of impulsive behavior (Simon, 1982d, p. 426).
118
Two Minds
7.2.4 Rationality and Chess
There are about 30 legal moves in a chess game. Each move and its
response creates an average of about 1000 contingencies. In a 40 move chess
game there are about 10120 contingencies. Chess masters are believed to look
at no more than 100 contingencies, only ten percent of the possibilities
existing for one move and a response (Simon, 1982j). Beginning with an
inordinately large number of possibilities, chess masters, and humans in
general, humans search for outcomes whose utility values are at least
satisfactory. Once found, the search stops. In other words, chess masters are
satisficers, and their rationality is bounded by their limited cognitive capacity
relative to their environment.
Chess grandmasters take so little time to decide on a move that Simon
says that it is not possible for their moves to be the product of "careful
analysis" (Simon, 1996, p.133). A grandmaster takes five or ten seconds
before making a strong move, which 80% to 90% of the time proves to be
correct and one that is "objectively best in the position" (Simon, 1983, p. 25).
Their skill barely diminishes when they play 50 opponents at once rather than
one opponent. How do they do it? When grandmasters are asked how they
play they respond with the words intuition and professional judgment. Simon
says that intuition is a “label for a process, not an explanation of it” (Simon,
1982h, p. 105; emphasis added). The process is subconscious pattern
recognition based on experiences stored in memory and retrieved when
needed. While short term memory (STM) can store only a relatively small
amount of information, long term memory (LTM) is, metaphorically
speaking, a large encyclopedia with an elaborate index in which information
is cross referenced. Cross referencing means that information is associative
with one piece of information linked or associated with other associated
thoughts. Cross referencing and chunking makes subconscious pattern
recognition or intuition easier.
Studies on recognition among chess masters have used eye movements to
assess recognition abilities. Chess masters examining a previously unknown
board position taken from an actual game immediately – within two seconds - shift their eyes to the most relevant part of the board. This means that they
immediately grasp or “see” the most important relationships on the board.
Simon concludes that it is sufficient to state that a chess master’s performance
is based on a knowledge of chess and an act of (subconscious) pattern
recognition. In fact, Simon helped develop a computer program with the
ability to mimic the eye movements of a chess master. His computer program
and human chess masters make the same mistakes, and both recover in a
similar way. For example, in one program a queen who was in trouble did
exactly what a human would do to not only save their queen but to get their
opponent in checkmate. Simon ads that the ultimate nature of human
Swimming Against the Stream
119
intellectual activity is best known through a chess playing machine. Human or
machine experts at chess or in any field of activity experts are expert (in part)
because of their ability for subconscious pattern recognition.
7.2.5 Artificial Intelligence
Writing in 1966, Simon believed that the word mysterious was the
adjective most often used to describe thinking, but that mysterious no longer
applied. The reason was increases in knowledge about the process of thinking
based on AI, that is, computer programs which mimic human problem solving
(Simon, 1966b). Simon's view of thinking affected by AI is that thinking is a
form of information-processing. Both human thinking, and informationprocessing programs perform three similar operations: they scan data for
patterns, they store the patterns in memory, and they apply the patterns to
make inferences or extrapolations. In fact, some programs reproduce and even
outperform human experts at problem solving. Simon concluded that there is
sufficient reason to believe that some kinds of human thinking closely parallel
the operations of an information-processing computer program. AI also led
him to conclude that intuition is a subset of thinking. The fact that the mind is
a serial information processor, it performs one (or only a very few) operations
at a time, places severe limits on human attention, binds our rationality, and
limits our capabilities for problem solving to a set of satisficing rather than
maximizing solutions.
Problem-solving thus involves two generalizations. First, a selective trialand-error search is made which by necessity can only consider a relatively
few possible solutions. The solution is thus a satisficing solution and the
search is based on rules-of-thumb or heuristics. Second, one of the basic
heuristics is means-end analysis. Means-end analysis involves three steps.
First, the current situation is compared to a goal, and differences between
them are noted. Second, a memory search is performed to identify an operator
which can bring the current situation more in line with the goal. Third, the
operator is applied in the hope of getting closer to the goal.
Since computers solve problems as humans do using heuristics and
means-end analysis, Simon concluded that computers display intelligence,
defined as behavior which is appropriate to the goal and adaptive to the
environment. Intelligence allows the limited processing capacity of the
organism, be it man or machine, to use efficient search procedures to generate
possible solutions, with the most likely solutions being generated early in the
search process (Newell and Simon, 1990). In order to test whether machines
display intelligence, Simon (and his colleagues) identified tasks requiring
intelligence and then built computer programs which carried out these tasks.
These tasks include playing chess, solving math and physics problems,
diagnosing disease, making discoveries in science, and even formulating
120
Two Minds
hypotheses and testing them empirically. In doing these things, Simon showed
that computers “think,” and that they possess (artificial or man made)
intelligence.
7.2.6 Machine’s That “Think”
Simon’s machine think in that they recognize patterns and apply "if –
then" rules in making decisions. Margaret Boden (1990), in her book The
Creative Mind. Myths and Mechanisms, uses the example of soybean diseases
to show that a set of “if – then” rules incorporated into a computer program
allows it to find patterns in a maze of data on soybean symptoms and then
correctly diagnose soybean diseases. Programs such as the ID3 algorithm not
only diagnose soybean diseases with an accuracy which would make any
psychic jealous, but it does so with maximum efficiency. That is, it asks the
right questions in the right order so as to make the right diagnoses in the
minimum amount of time. ID3 has discovered patterns in data which humans
have not, and it has discovered strategies in chess previously unknown by
chess masters. Computer programs have been developed which input and
output words, formula, images, and musical notations. These computer
programs have been said to display creativity.
The first AI program, developed by Simon, is the Logic Theorist (LT).
Written in 1956, LT discovers proofs for theorems contained in Alfred North
Whitehead and Bertrand Russell’s Principia Mathematica (Whitehead and
Russell, 1962). In order to do this LT mimics expert human decision-makers
by working "backward." Both LT and expert human decision-makers use
information about the goal to eliminate many paths without having to try
them. Human novices, on the other hand, solve problems in a more time
consuming inefficient "forward" manner. Work on LT demonstrated to Simon
that trial-and-error (a procedure of science) and insight (an apparent
discontinuity, or mystery) are complementary with each other. LT also lead
Simon to conclude that the human brain is analogous to a digital computer.
The value of the brain- as-computer metaphor is that it takes the mystery out
of concepts such as intuition and insight (Simon, Newell, and Shaw, 1989, p.
7).
The General Problem Solver (GPS), an early program developed in 1957
engaged in means-end analysis, a basic heuristic in problem-solving. The
EPAM program (Feigenbaum and Simon, 1989) simulates human recognition
or learning, while the MAPP program (Simon, 1989b) which simulates the
ability to recognize patterns in a manner similar to chess (grand) masters. The
APS (“adaptive production system”) program engages in learning-by-doing,
and learning-by-example. The APS program learns-by-example to solve
algebraic problems by inspecting each step in an algebraic problem placed in
its memory. When faced with any algebraic problem it goes through the steps
Swimming Against the Stream
121
and arrives at a solution to a particular problem. APS programs learns-bydoing an algebraic problem, and then uses that example to learn how to solve
other algebra problems.
Simon and his colleagues also developed several programs which make
discoveries in science including BACON, BLACK, GLAUBER, and STAHL.
BACON analyzes data sets and derives quantitative relationships among data
sets. BACON “discovered” many well known scientific laws including
Galileo’s law of uniform acceleration, Kepler’s third law, Boyle’s law, and
Ohm’s law (Langley et.al., 1987). And it does by considering the simplest
explanation (pattern) first before moving on to more complex explanations.
BLACK, named after Joseph Black, works on situations in which two
substances are additive. If analyzing the data show that the two substances are
not additive then BLACK finds one or more unobservable properties of the
substances to explain non-additivity.
GLAUBER, named after the chemist Johann Glauber, divides substances
into groups according to their observable properties. Similar to Glauber,
GLAUBER uses a sample of acids and alkalis to infer correctly that every
acid reacts with alkalis to form salt. STAHL, named after chemist Georg
Stahl, is given a set of heuristics used by chemists and a list of experimental
results on the nature of combustion in historical sequence. Similar to human
chemists, STAHL’s hypotheses about combustion are sometimes incorrect,
but similar to human chemists STAHL reviews previous experimental results
and corrects its mistakes. The result is that STAHL correctly reproduced the
approximately 80 year development of the oxygen theory of combustion from
the phlogiston theory.
7.2.7 Intuition. Simon's Early and Later Views
In the second edition of Administrative Behavior (Simon, 1965) Simon
recognized the value of experience and habit in decision-making. Experience
becomes human capital; habit becomes internalized as unconscious and
automatic reflex actions. Decision-making using experience and habit relies
on "clues." Whether clues are recognized consciously or known only to the
subconscious, they enhance our understanding of particular situations, and
improve decision-making. Experience and habit become part of effective
procedures in decision making. Simon comments that “human rationality
relies heavily upon the psychological and artificial associational and indexing
devices that make the store of memory accessible when it is needed for the
making of decisions” (Simon, 1965, p. 87). In the fourth edition of
Administrative Behavior, 1997, and having AI as his framework, Simon
(1996) would refer to the associational and indexing devices of memory as
intuition.
122
Two Minds
Another value of experience and habit in performing purposive or rational
behavior is that it “permits conservation of mental effort by withdrawing from
the area of conscious thought those aspects of the situation that are repetitive”
(Simon, 1965, p. 88). And it permits similar stimuli or situations to be met
with similar responses or reactions, without the need for a conscious
rethinking of the decision to bring about the proper action (p. 88). In the
fourth edition, Simon (Simon, 1997) would also refer to this as intuition. In
the earlier additions of Administration Behavior, Simon did not discuss
intuition because he was uncertain about the nature of subconscious thinking
processes. In the fourth edition he introduced material about intuition
because, he says, "…we have acquired a solid understanding of what the
judgmental and intuitive processes are" (Simon, 1997, p. 31).
These processes are subconscious and/or rapid, and based on experience
which by-passes a conscious "orderly sequential analysis" of a situation.
Simon went so far as to say that intuition is actually analytical thinking
"frozen into habit and into the capacity for rapid response through recognition
of familiar kinds of situations" (Simon, 1997, p. 139). Intuition and analysis
are complementary with each other and almost always present in all human
decisions, including those of scientists. Thinking about the use of intuition
among scientists in general and physicists in particular, Simon spoke about
“physical intuition,” that is, intuition used by physicists or scientists in
general. In “New Age,” and spiritual literature, physical intuition refers to
physical sensations in your body. For example, when you meet someone for
the first time, feeling an ache in your gut (gut feeling) is your intuition
“telling” you that this person is to be avoided. The difference in the use of the
term physical intuition between Simon and New Age/spiritual literature is
both startling and revealing of differences in approach to the topic of intuition.
The combination of intuition and analysis is present in chess grandmasters because chess “is usually believed to require a high level of intellect”
(Simon, 1987, p. 28), and grandmasters use the word intuition when
describing how they do what they do in chess. Chess grandmasters take only a
very few seconds to decide on their next move and then a longer period of
time verifying that their “educated hunch” is correct. One test for the use of
unobservable physical intuition was done with the use of protocol analysis in
which a novice and an expert were given a physics problem to solve, and each
person verbalized what they were thinking. The results showed that the more
experienced person solved the problem in less time, required fewer steps to
solve the problem, spent less time per step, did not write down as many
relevant facts or equations to solve, and expressed more confidence in
themselves. In essence, the skilled person took a series of appropriate short
cuts and avoided conscious calculation of how to solve the problem. This is
possible because an expert’s knowledge is similar to an encyclopedia with a
large index in which entries are cross referenced. That is, not only does the
Swimming Against the Stream
123
expert have more knowledge than the novice, but the expert can more rapidly
elicit relevant facts from memory. The expert exhibits "the usual appearance
of intuition," while the novice uses more "conscious and explicit analysis"
(Simon, 1996, p. 136). The conclusion Simon reaches is that experience
allows people to make decisions intuitively, or judgments "without careful
analysis and calculation" (ibid, p.136).
Simon says that intuition "…is no deeper than the explanation of your
ability, in a matter of seconds, to recognize one of your friends whom you
meet on the path tomorrow as you are going to class" (Simon, 1983, p. 26).
Experience and knowledge is the key to intuition because paraphrasing
Poincare, "inspiration comes only to the prepared mind" (ibid, p. 27).
7.2.9 Herbert Simon on Intuition Revisited
Simon’s studies of MBA students and experienced business executives
asked to analyze a situation show that the two groups come to similar
conclusions. However, the experienced executives came to their conclusion in
much less time, "with the usual appearance of intuition” (Simon, 1997, 136).
The work of the MBA students, on the other hand, “was done slowly, with
much conscious and explicit analysis" (ibid., 136). The conclusion Simon
reaches is that experience allows people to make decisions intuitively, or
judgments "without careful analysis and calculation" (ibid., 136). In a study of
decision making among physicists, Simon and Simon studied the protocols of
two persons solving a physics problem by recording them verbalizing what
they were thinking while solving the problem. One of the persons was a
“novice” at this type of problem solving, the other an “expert.” The expert
solved the problem in less time, did not follow the reason-only steps, required
fewer steps to solve the problem, spent less time per step, did not write down
as many relevant facts or equations to solve, and expressed more confidence
in themselves. In essence, the skilled person took a series of appropriate short
cuts. These short cuts imply that they used, what Simon calls physical
intuition – intuition used by physicists. That is, they read the problem, created
a mental representation, and created a set of equations based on that mental
representation to solve the problem. While the expert’s approach is more
“physical” or “primitive” (Simon and Simon, 1989, 224), the novice’s
approach is more algebraic. The conclusion that physical intuition “accounts
for the superior ability of physicists to solve physics problem should occasion
no surprise. Physicists and teachers of physics have been saying that for
years” (ibid., 230).
124
Two Minds
7.2.10 Intuitive Rationality
Intuition is useful, and it is a subset of bounded rationality. IR and BR are
consistent with each other, and all "serious thinking" uses both. Both use
search-like processes both lead to sudden recognition of underlying patterns,
and the focus of one's attention plays a major role in the choices one makes.
Intuition is said to be what is responsible for people finding solutions to
problems "suddenly" and having the "aha" experience. Having AI as a
framework, Simon understood these to be genuine experiences, and to lead to
judgments which "frequently are correct" (Simon, 1983, p. 25). Simon helped
create computer programs which mimic expert human decision makers who
use the word intuition as a label for how they make decisions. Creating
machines that think led Simon to his “explanation” for that very human
phenomena called intuition.
7.3. Harvey Leibenstein
“Great economists, it could be argued, achieve (much of) their status…by
creating penetrating images in the minds of their peers which govern how
these others view the world” (Szostak, 1999, p. 11).
During the 1940’s, before going to Princeton and receiving a Ph.D.,
Leibenstein taught at Illinois Tech (IT) While at Princeton he had a
conversation with Mark Perlman in which he told Perlman that IT had some
brilliant faculty members who were on par with the faculty at Princeton.
When Perlman asked who, Leibenstein said, “Some guys you never heard of,
Franco Modigliani and Herb Simon." Leibenstein and Simon, along with
Scitovsky, and George Katona were first generation behavioral economists.
Leibenstein and Simon’s work have both similarities and differences. Both
emphasize procedures by which decisions are made, the internal operation of
the firm, motivation, peer pressure, non-fully rational behavior, habit,
resistance to change except under duress, internal inefficiency, and
multiplicity of goals. There are differences between the works of Simon and
Leibenstein. First, satisficing is similar to maximizing in that both are goaloriented, directed behavior. On the other hand, X-inefficient behavior stems
not from satisficing behavior, but from sloppiness, laziness, and the like.
Second, organizational slack is treated as having the positive effect of helping
to stabilize the organization during times of crises by providing it with a
cushion. Leibenstein’s general approach was to call organizational slack a
form of inefficiency. Third, the Carnegie School takes more of an operations
research and organizational behavior approach, while XE theory orients itself
within an economic theory framework. In fact, Leibenstein insisted that he
Swimming Against the Stream
125
was a neoclassical economist, but one who did not believe that human
behavior was always maximizing.
Leibenstein’s X-Efficiency theory began with his own intuition, aha
experience. While he was at Berkeley in the early 1960’s, one of Harvey’s
graduate assistants displayed an effort level which varied considerably from
day to day. He knew about studies done using data collected in various
countries showing that given amounts of labor and capital produced output
levels which varied widely. Interacting with his graduate student one day it
suddenly seemed intuitively obvious that for humans in general, effort and
efficiency vary: it was a moment of discovery. To capture his intuition in
terms of its implications for economic theory Leibenstein developed Xefficiency theory and illustrated it with three visual images, or intuitions: a
torn net, a split personality, and an army on the battlefield.
7.3.1 The Army on the Battlefield
In Keynes, Bloomsbury and the General Theory, Piero Mini says that
Keynes had “a tendency to think in terms of the spiritual, the psychological,
the non-material and the imponderable” (Mini, 1991, p. 158). Mini says that
had Keynes been a general he would have stressed willpower and solidarity
above the number of soldiers. This is also true about Leibenstein. In the first
page of the Preface of Beyond Economic Man, Leibenstein quotes Tolstoy for
a military analogy of the nature of x-efficiency. In War and Peace, Tolstoy
writes that “Military science assumes the strength of an army to be identical
with its numbers. Military science says that the more troops the greater the
strength. Les gros battaillons ont toujours raison (Large battalions are always
victorious) . . . In military affairs the strength of an army is the product of its
mass and some unknown x . . . That unknown quantity is the spirit of the
army, . . . The spirit of an army is the factor which multiplied by the mass
gives the resulting force. To define and express the significance of this
unknown factor—the spirit of an army—is a problem for science. This
problem is only solvable if we cease arbitrarily to substitute for the unknown
x itself the conditions under which that force becomes apparent—such as the
commands of the general, the equipment employed, and so on—mistaking
these for the real significance of the factor, and if we recognize this unknown
quantity in its entirety as being the greater or lesser desire to fight and to face
danger” (Leibenstein, 1976, p vii). Leibenstein goes on to say that
understanding the level of economic output requires more than knowing the
“observable inputs” including technology. What Tolstoy calls “spirit”
Leibenstein calls effort, both physical and mental. It should be intuitively
obvious that one thinks intuitively, not logically about the spiritual, the nonmaterial, and the imponderable. At the same time, to express one’s intuitions
rationally requires combining intuition and analysis.
126
Two Minds
7.3.2 The Torn Net
Frank Knight’s explanation of the role of an entrepreneur included the
ability to utilize intuition in the face of uncertainty. Leibenstein’s explanation
about why entrepreneurs are important made use of his own imagery or
intuition. Harvey questioned the assumption that individuals are always fullyrational – maximizers – and thought it would be helpful if we could observe
the degree of a person’s rationality. In his book, General X-Efficiency Theory
and Economic Development, Leibenstein asks how a psychological variable
such as effort might affect economic development. One way that answers this
question is by distinguishing labor time from labor effort. This distinction is
between inputs purchased by a firm, and the effectiveness with which they are
used by the firm, respectively. A traditional economic model (of
development) emphasizes time while Leibenstein emphasizes effort. In
addition, traditional economic models assume that effort is a given, while
Leibenstein emphasizes that effort varies according to the pressures felt by an
individual. In a traditional model of economic development, once you assume
that prices are well known, production and cost functions are well defined,
markets are well functioning, then there is little if any use for entrepreneurs.
Leibenstein’s view of the economy is that these assumptions often do not
hold. In order to express this, he uses the image of a net. The economy
described by a traditional economic model of development is a well construed
net, all the strands in place, and all of them in proper condition.
The strands of the net are the pathways which carry inputs to the firms
and final goods and services to the households. The intersection of strands are
the firms and industries which receive inputs, and the households which
receive final goods and services. In a well-functioning market inhabited by
cost-minimizing firms the net is complete without any tears. In many cases
the net contains gaps from being worn or torn. Tears in the net come from
some inputs not being available, some goods or services not being available to
all buyers, some opportunities for mutual gain not being exploited because of
inertia. The function of the entrepreneur is a gap-filler, mending or
completing the net (Leibenstein, 1978).
Leibenstein acknowledges a debt to David McClelland’s work in The
Achieving Society. According to McClelland, economists’ models of
economic development are incomplete because they emphasize rational
behavior, capital, technology, division of labor, and entrepreneurship at the
expense of sociological and/or psychological explanations including “need for
achievement.” Understanding development, he says, would be helped if we
possessed a “psychic X-ray” (McClelland, 1967, p. 39) allowing us to observe
a person’s need for achievement rather than relying on the a priori or selfevident assumptions about economic-man. We can reasonably interpret a
Swimming Against the Stream
127
psychic x-ray as intuitive abilities. One source of need for achievement is a
person’s core religious value, specifically “positive mysticism” (ibid, p. 367)
or the “mystical strains” of Christianity. A mystical strain emphasizes a
mystical relevance for life rather than renouncing the world in order to
achieve this link. The rise of the business class in 19th century Japan is
associated with Zen and its emphasis on positive mysticism, individuality, and
rejecting of many traditional religious rituals.
7.3.3 The Split Personality
The concept that an individual can contribute to X (in)efficiency arises
from an individual’s split personality. Of course, a human personality is not
literally split, but the image of a split personality conveys the intuition that
individuals are torn -- not literally – in their preferences. On the one hand, we
want to adhere to standards, to strive for the maximum, and to strive by being
calculating and attentive to details. In other words, this aspect of the
personality is our rational self, the part of us that uses "tight," focused, or
rational decision-making procedures. Leibenstein refers to this aspect of our
personality as the "superego function." The other tendency is for each of us to
"kick back," to use "loose" decision-making procedures, to follow our "animal
spirits." Leibenstein refers to as the "id function." It is the id function that
leads us to be "unconstrained," that is, unwilling but not necessarily unable to
be calculating, attentive, rational.
XE theory assumes that, on the average, each individual is influenced by
both functions in a way that leads to a compromise between the two. That is,
each of us forges a compromise between the way we feel we must behave and
the way we would like to behave were it not for a sense of obligation to duty
or to a set of standards. In other words, each individual strikes a compromise
that provides them with a sense of (psychological) "comfort." Individuals are
thus selectively rational. X-E.
Leibenstein assumes that, on average, each individual is pushed by their
superego to maximize, and pulled by their id to kick back and watch the world
flow by. That is, each of us forges a compromise between these two forces
that provides them with a sense of (psychological) "comfort." The human
personality is thus seen as having two end points. On the one end is an intense
concern for constrained behavior – economic man -- while on the other end is
a complete lack of constrained behavior. The former would be a fully rational
person, with others exhibiting "selective rationality." Rationality is thus a
continuous variable with an "economic person" exhibiting complete constrain
concern. The economic person is thus a limiting case: a characteristic of the
decision-making procedures used by some people at some times but not
necessarily characteristic of all people at all times.
128
Two Minds
Leibenstein says that his superego function is similar to Harry Levinson’s
“ego-ideal” in his book, The Exceptional Executive (Levinson, 1968). The
ego-ideal leads to creativity, itself expressed as a “sudden inspiration or
illumination” (ibid, p. 179) – the intuitive aha experience. The ego-ideal is
part of the functioning of our conscience, is an “internal self-governing agent,
and is our “internalized image of oneself at his future best” (ibid., 13).
People’s needs are divided into 3 categories: ministration, maturation, and
mastery. Maturation needs includes the use of “female intuition.” Mastery
needs include affection or cooperation with others through “empathy”
(intuition).
7.4 George Shackle
Fore George Shackle reason and imagination are the “two faculties which
make us human” (Shackle, 1972, xii). Reason or rational thinking is the same
for all people, and hence is “sure, safe, even in a sense simple” (ibid., xii).
Reason means the use of knowledge to evaluate the relative costs and benefits
of various options about the future. The underlying assumption is that human
behavior is self-interested, and that reason is applied to whatever
circumstances a person find themselves. However, in reality we can not have
sufficient knowledge about the future, and hence we can not either apply
reason or be rational. In addition, we do not have sufficient knowledge to
understand our current circumstances. On the other hand says Shackle,
economists assume that we have the knowledge of both the future and our
current circumstances which leaves no purpose to ask what should be the
central question in economics – what is the source of our knowledge?
What do we substitute for knowledge in the absence of knowledge? In the
absence of knowledge, people use imagination, or intuition. Imagination is
used when thinking about practical matters - not theorizing - “in which it is
only useful to imagine what is deemed to be possible” (ibid., xii). Reason may
be sure, safe and simple, with one set of rules for all people, but imagination
has neither of these characteristics. “The future is imagined by each man for
himself and this process of the imagination is a vital part of the process of
decision” (ibid., 3). If the goal of rational choice is maximizing (expected)
utility, then how do we evaluate our imagination? The “test of success in an
expectational system is the maximum attainment of a good state of mind, a
good state, that is, of imagination, of the conception of states to be attained
which cast the warmest glow of anticipation on the present (ibid., xvii-xviii).
The source of a lack of knowledge and an inability to be rational is the
existence of time. The past no longer exists, and present can not be fully
known, and the future is unknown. Thus there is a basic conflict between
assuming rationality, and including time as a variable in economic analysis.
The conflict is that time coexists with ignorance, a lack of knowledge, and the
Swimming Against the Stream
129
impossibility of rationality. The “insufficiency of knowledge is permanent
and part of the nature of things, for consciousness consists precisely in the
continuous gaining of knowledge…What does not exist cannot now be
known” (ibid., p. 3). We have knowledge of the present moment, but make
choices about future moments. The idea of a present moment implies a
“successor moment,” which also has a successor, which has a successor.
Shackle names this idea of a continuous stream of present and successor
(present) moments as “calendar intuition” (ibid., 278). Shackle also says that
our “spatial intuition of time” (ibid., 284) creates a sense that time has room
for both earlier and later. The concept of time lies within the mind, but the
mind’s own activity leads to the conclusion that time is external and
independent of itself (ibid., 264). Our notions of time, the meaning of the
future, of our anticipated experience “arise from our intuitions of the nature or
meaning of experience” (ibid., 245). So must be our notions of uncertainty,
risk, expectations of self and others, motivations of self and others. In
business it means using imagination or intuition to understand the mind and
intentions of its rivals, innovators, union leaders, and consumers. Rational
choice requires full information, including what others are choosing or are
going to choose. Let’s go out on a limb here and say that rational choice
requires mind reading, or what Smith called sympathy.
Accepting time but rejecting rationality is a rejection of economics,
whereas rejecting time is called partial equilibrium analysis or (subjective)
marginalism. Marginalism developed the idea that differences -- ∆U/∆Q, and
∆C/∆Q -- is the key behavioral component in understanding market behavior,
and is “natural and intuitive in men’s management of their affairs” (ibid.,
247). This intuitively appealing behavior is also “logically correct and
justified conduct, it was mathematically correct, it was conduct which led to
making the most of given resources…it was conduct which achieved
economy” (ibid., 247). Hence, neoclassical economics from which
“economics was suddenly given (so it seemed) the full standing of science”
(ibid., 248), is based on a notion which is intuitively obvious to the casual
observer, and which is consistent with mathematics and the principle of
economy.
The mind is “a scheme of taxonomy for filing impressions in a way which
relates them to each other and this scheme ... is essentially mathematical”
(ibid., 32). In other words, the mind looks for and creates patterns, and these
patterns can be expressed mathematically. The patterns are expressed both
mathematically and in terms of quantities. Focusing on quantities allows some
insight into the world of economics which would otherwise appear as “a mere
chaos of proliferating and unintelligible detail, reasonable and orderly only in
the small, in space and time, and otherwise lacking any sense of architecture”
(ibid., 48-49). As Frank Knight said, the world is a big blooming buzzing
confusion.
130
Two Minds
According to Shackle, the neoclassical assumption that humans pursue
their self-interests by applying reason to their environment is itself an
intuition. But since this is not possible, the only way to analyze the future is to
“suppose conduct to transgress the bounds of demonstrable reason” (ibid.,
91). That is, if we want to assume rational behavior then we must assume that
behavior is being guided by something other than reason, something such as
imagination or intuition. Shackle says that we must “examine the psychic
processes, transcending and supplanting reason, which serves for guidance
where data are lacking and demonstration at a loss: (ibid., 272). Including
time and assuming rational behavior can only be done by assuming the
existence of a “psychic process which is ultimately extra-logical” (ibid., 255).
Reason has been so attractive to humans because it is within us. But rational
choice means having a “psychic weighing machine” (ibid., 82) in order to
have preferences and balance costs and benefits. Economics tried to avoid the
psychic or intuitive realm by calling this machine, “tastes.” But preferences,
tastes, and choices are about the future, and hence beyond the field of
rationality.
The assumption of rationality is also paradoxical because it assumes that
human conduct in part of the “determinant order and process of Nature, and at
the same time, humans choose for themselves what is best for particular
circumstances. That is, we are free to choose and our choices are predictable
at the same time. We are free to choose, but only one choice will be choosen!
This is similar to Leibenstein’s complaint about NCT. Natural science
includes time in its analysis, so why can’t economics? Science which studies
inanimate objects is different from economics which is or should be
“concerned with the thoughts and deeds, not with the ultimate chemistry of
man” (ibid., 4). Assuming that humans are inanimate objects allows the
economist to more easily assume that humans have perfect knowledge, and to
assume that human behavior is predictable. These assumptions have lead,
according to Shackle, to the development of both partial and general
equilibrium theory. In both theories, alternative options are compared, but the
central question, what does the individual know about the options, is never
asked. This is because of the assumption that the knowledge of the options
exists in the mind. But in a world of uncertainty and novelty where there are
“fundamentally undeducible things, waiting to be encountered for the first
time” (ibid., 26) -- complete knowledge, reason, and mathematical models are
severely limited.
And what of economists trying to imitate natural science? “Economics
has gravely and greatly misled itself by a tacit belief that rational self interest
is as simple a basis of prediction as the laws of physical motion” (ibid., 37).
There are fundamental differences between natural science such as physics
and economics. First, the former does not depend on “conscious knowledge,
memory, past experience, reason or imagination with their more than infinite
Swimming Against the Stream
131
possibilities of variation in character and combination” (ibid., 28). Second,
natural laws such as gravity can be expressed in a single formula, whereas
human behavior cannot. Third, natural science looks “inward at the Origin” -the search for the unity of life, ultimate reality, the “single Secret of
Nature,”or “one big thing” -- while economics looks “outward at the
Manifestation”(ibid., 29). Fourth, rational self-interest will not explain the
behavior of inanimate objects, so it’s not part of the theory of, for example,
atoms. (It can’t even explain the behavior of humans.) Taking up the
axiomatic method and assuming rational self-interest, economics neglected
the equivalent of “case law” (ibid., 38). Hence economics became “an attempt
to study in outline, by means of an imposed simplicity and precision, some
aspects of a subject-matter which in the fullness of its unabstracted nature
involves a vast richness of intricate and yet essential detail (ibid., 38). Fifth,
natural science is based on what we see, economics is built on what we think,
the imagination (ibid., 66), “thoughts about thoughts” (ibid., 71). Sixth,
mathematics may demonstrate superiority of one choice over another, math
and physics may show the unity of life, but a “simple, single principle” of
determining human behavior is not possible. Human behavior is about human
thoughts and the analogy between thoughts, atoms and mathematical symbols
breaks down.
“The prime duty of a theory is to tidy the mind, leaving it clear for
practical activity” (ibid., 71). Theory imposes order on the disorderly. Smith
couldn’t say it better. Economists talk about desire, filling in our gaps of
knowledge, and uncertainty by inventing ideas. Leibenstein saw the
limitations and created X-E from an intuition. (Entre as gap filler; xe theory
created surprise but no wonder to among others; critics attacked; HL
wondered about neoclassical theory and created xe theory.)
7.5 Frederick von Hayek.
Hayek discusses our “two types of mind” (von Hayek, 1978) as two types
of scientific thinking. One is “the perfect master of the subject” (ibid., 50), a
person who has at his fingertips an apparently inexhaustible amount of facts
and theories. The second consists of “puzzlers,” or “muddlers” (ibid., 51).
Puzzlers process “wordless thought,” they “see” connections between and
among things they cannot put into words, and they make discoveries after a
period of reflection which comes as a “surprise,” what we refer to as the aha
experience (ibid., 53-54). Hayek describes himself as a puzzler. He describes
Jacob Viner as a perfect master of the subject, but Frank Knight as “a puzzler
if there ever was one” (ibid., 51). About Keynes he says, “He was, by gift and
temperament, more an artist and politician than a scholar or student. Though
endowed with supreme mental powers, his thinking was as much influenced
by aesthetic and intuitive as by purely rational factors. Knowledge came
132
Two Minds
easily to him and he possessed a remarkable memory. But the intuition which
made him sure of the results before he had demonstrated them, and led him to
justify the same policies in turn by very different theoretical arguments, made
him rather impatient of the slow, painstaking intellectual work by which
knowledge is normally advanced” (ibid., 287; emphasis added).
In The Sensory Order Hayek (1976) distinguishes “two orders,” the
“subjective, sensory, sensible, perceptual, familiar, behavioural, or
phenomenal” order, and the “objective, scientific, ‘geographical’, physical, or
sometimes ‘constructional’” order (Hayek, p. 4). He refers to these as the
phenomenal or macrocosm, and the physical or microcosm orders,
respectively. The sensory order is our perceptions of the world, and is
different from the objective reality which is the world. The sensory order is,
our sense of order. It is created by physical processes occurring in our nervous
system, specifically, the continuous stream of electrical impulses of a
seemingly endless number of neurons connected with each other in seemingly
endless numbers of combinations or patterns. These physical processes or
patterns are spontaneous, changing and evolving, and are occurring without us
being conscious of them. Without our sense of order, the world would appear
as an “indistinguishable homogeneity” (Horwitz, 2000, p. 32). The patterns of
electrical impulses or excitation creates our sense of order about the world
and hence is our model of the world. Our model of the world is forwardlooking allowing us to anticipate the future. More than being our model of the
world, the patterns of excitation is our mind. Our mind is the spontaneously
created and recreated sensory order. As such, there are serious limits to our
ability to “know, predict, and control” (Horowitz, p. 24) our own mind. And,
there are serious limits for us to know the world through reason, as usually
defined. There are three important implications from the limits to reason.
First, human reason contains both intuitive and logical or analytical elements.
Second, there are severe limits to rational choice. Third, not all knowledge is
the explicit or static form which is used to express scientific theories. Nonexplicit form of knowledge mean that we cannot articulate everything we
know or all the rules which govern our behavior. Non-explicit knowledge is
tacit, or intuitive-based knowledge (or patterns). Our mind is the coordinator
of impulses which creates a sense of order. What we know from experience
are past patterns (of impulses) and expectations of future patterns. Being
unconsciously aware of patterns is thus significant in the theories of both
Hayek and Herbert Simon.
There are similarities between Hayek’s view of the mind and his view of
the market. First, what the mind is for the individual, prices -- the market
system – is for the economic sphere of life - the coordinator and
communicator of knowledge (Hayek, 1945). Second, just as the mind is a
spontaneous creation which we cannot predict or control, so too is the market
system an institution spontaneously created and beyond our control (without
Swimming Against the Stream
133
destroying it). Third, knowledge possessed by the mind and the market are
conjectural. Fourth, the mind and the market are forward-looking. Fifth,
knowledge is generated spontaneously in both the minds and the market.
Sixth, just as the mind generates tacit or intuitively-based knowledge, so does
the market rely on tacit knowledge. Seventh, he rejects the rationalist view of
both knowledge, and choice. One could say that Hayek was an advocate of
behavioral economics/economic psychology.
134
Two Minds