The word
“personality” comes from the Latin word persona meaning the
mask worn by an actor in a drama to indicate his character. It is probably
useful to consider a person’s personality as the totality of the ways in which
he behaves in various situations. Historically psychology has primarily
conceptualized personality in terms of specific theories such as those of
Freud, Jung, Adler, Rank, and Horney. Many of the constructs of the different
theories are now part of common language (e.g., “anal character,” “extrovert,”
“birth trauma”). Such theories have been useful for suggesting ideas to be
experimentally tested and developed. They have also been useful in
conceptualizing possible relationships between different behaviors. However,
there are many problems with these classical theories.
The main problem
is that most of the constructs of the different theories cannot be tested
and measured independently. A person’s behavior might easily be explained
in terms of the interactions between different constructs of a specific theory,
but such post hoc explanations have limited value if we cannot
get at the individual constructs. For example, Freud’s classic theory, including
his constructs of ego, id, and superego, can explain almost any behavior after
it happens. But how do we get an independent
measure of the nature and functions of the id? The Freudian analyst might
try to do this by techniques such as dream interpretation or word association.
However, such procedures are confounded with other dynamic personality variables,
do not meet the type of scientific rigor necessary to give them widespread
usefulness, and are generally more a matter of artistic interpretation than
logical measurement. It is sometimes argued that the human personality cannot
be broken down into constituent parts, but must always be viewed as a whole.
The experimental psychologist, however, generally assumes that it can
be broken down. This in no way devalues the
beauty of man or the complexity and richness of his experiences.
If we could measure
and test the various constructs and assumptions of the different personality
theories, we could begin a much needed evaluation and synthesis of the different
ideas. But since many of the constructs are not readily measured and tested,
many theories continue on with little change. A disadvantage of any particular
theory is that much information may be distorted or lost in an attempt to
fit the person into the theory. The classic personality theories are probably
most useful in understanding the personality of the person who made up the
theory and secondarily the personalities of a few of the people he counseled.
In this chapter we
will look at personality from the viewpoint of learning. This approach basically
involves two categories of phenomena:
(1) those genetic
and physiological variables that affect a person’s behavior directly and/or
that predispose him for certain types of learning, and (2) specific types
of behavioral abnormalities that result from certain learning experiences.
The vast majority of human behavior is learned by principles such as those
discussed in the preceding chapters. Thus the best understanding, and probably
also the best approach to change or treatment, of these behaviors is from
the viewpoint of learning rather than from a personality theory. But there
are individual differences resulting from such factors as genetic and physiological
variables that affect what a person learns and how fast he learns it. A personality
classification based on such variables would then be very useful in understanding
and dealing with behavior. There are also a number of abnormalities, such
as experimental neurosis and learned helplessness, whose genesis and treatment
are being systematically researched. It may be that in a discussion or description
of personality, reference to one of these phenomena would have heuristic and
explanatory value.
PSYCHOPATHOLOGY
Psychologists tend
to perceive behavior more in terms of learning than of instinct, while the
opposite is often true of zoologists, particularly ethologists. Thus both
groups, with different biases, are concerned with the nature-nurture controversy. That is, to what extent is
behavior determined by hereditary influences (nature) and to what extent by
environmental influences (nurture), such as learning? With simple organisms,
where breeding and environment can be well controlled, it is often relatively easy to separate genetic variables from
environmental variables. But as the possibility of such control decreases and
as the organism becomes more complex, it is harder to isolate the different
variables. Human behavior is a highly complex interaction between nature and
nurture in which a person’s genetics affect his behavior and physical characteristics,
which in turn affect the types of interactions he has with his environment and
how other people respond to him, which then affects his behavior and aspects of his physical appearance (e.g., clothes style),
and so forth. Such complex interactions are difficult to break down into nature
and nurture components.
Dobzhansky (1972)
has also argued that the fundamental peculiarity of human evolution, as opposed
to the evolution of most other species, is that man has been selected for
“trainability, educability, and consequent plasticity of behavior.” Whereas
for most species of animals it is biologically
adaptive for all members
of the species to have certain uniform behaviors under genetic control, man’s
evolutionary success is his ability to adapt to cultural variables. For man, adaptation by cultural changes is more effective than
adaptation by genetic changes. But this adaptability in man makes it even
harder to identify the role of genetic variables.
Animals such as mice
(e.g., Bovet et al., 1969) have been selectively bred for a variety of “behavioral”
characteristics. Mice have been bred for different degrees of vigor, aggressiveness,
ability to win fights, and performance in mazes, while rats have been bred
for different degrees of activity and fearfulness, and for maze performance.
In the Tryon strains of rats, those rats that learned mazes well (maze bright)
were mated with similarly bright rats, while slow maze learners (maze dull)
were mated with similarly dull rats. After seven generations of such selective
breeding there was little overlap in maze performance between the maze bright
strain and the maze dull strain.
If characteristics
such as fearfulness have a genetic component in rats (see Gray, 1971), we
might expect something similar in humans, particularly in the case of psychopathy.
Unfortunately in the area of mental health the criteria and diagnostic definitions
of “illnesses” such as schizophrenia are so poorly defined that it is difficult
to distinguish nature from nurture. But, drawing from Rosenthal (1971), we
can suggest some possible relationships between genetics and psychopathology
in man.
Schizophrenia
Of all the
traditional clinical diagnostic categories, “schizophrenia” is one of the
broadest and most poorly defined, so that it is very difficult to make any
generalizations about people classified as schizophrenic. However, we do know
that symptoms of schizophrenia usually include things such as unusual thought
processes and associations, inappropriate or deteriorated emotions, ambivalence
of feelings, detachment and inadequate adaptation to “reality,” and lack of
interest or will.
One way of investigating
the possible role of genetics in schizophrenia is to study twins. Twins are
either monozygotic (MZ), which means that they are from a single
fertilized ovum (identical twins), or dizygotic (DZ), meaning
that they come from two different fertilized eggs fraternal twins. If the
two twins of each pair are raised in approximately the same way and in the
same environment, then a comparison of the incidence of schizophrenia between
MZ twins and DZ twins should give us some general idea of the role of genetics
in schizophrenia. This is done by looking at the concordance rate for the two twins of each pair, i.e., the probability
that they both display the same trait, in this case schizophrenia. In summarizing
a number of such studies, Rosenthal (1971, p. 74) suggests that the concordance
rate for MZ twins is usually 3 to 6 times as high as the rate for DZ twins,
which offers “strong but not conclusive evidence of a genetic contribution
to schizophrenia.” However, the concordance rate for MZ twins is always less
than 100 per cent, sometimes much less. Thus nongenetic factors play a large
role in determining who develops schizophrenia. Also, it appears that the more severely schizophrenic one twin is, the more probable
it is that the other twin will be schizophrenic.
Another way of trying
to separate nature and nurture variables in schizophrenia is to compare the
incidence of schizophrenia between parents and their children when the children
were separated from the parents early in life and raised in other homes. Again,
in summarizing such studies, Rosenthal (1971, p. 84) concludes that “the evidence
has turned up so consistently and strongly in favor of the genetic hypothesis
that this issue must now be considered closed.”
An important question
is whether the genetic factor actually produces the schizophrenic behavior
per se, or if it produces a predisposition to learn those behaviors called schizophrenic.
Manic-Depressive Psychosis
When a person has an
“attack” of mania he feels elated, more energetic and lively,
or perhaps irritable and impatient. In more extreme cases he may be incoherent,
very irritable, and extremely active. With attacks of depression the person is sluggish, has a low energy level, poor appetite, and
disturbed sleeping habits; he may also have feelings of worthlessness or of
being evil. A person classified as manic-depressive drifts
from the normal state into the manic or depressive state. Some go primarily
into the manic state, some into primarily the depressive state, and some people
cycle between the manic and depressive states. Studies of manic-depressive
twins have not generated data as clear as that from studies on schizophrenia,
but it appears that the concordance rate for MZ twins is higher than that for
DZ twins (Rosenthal, (1971, p. 119).
Overall, Rosenthal
(1971, p. 153) concludes that “the genetic studies—with their sundry faults—provide
much evidence for the view that schizophrenia and manic-depressive psychosis
are valid, distinctive, genetic disorders, and that unless future studies
provide evidence to the contrary, it would be foolhardy to dismiss them out
of hand.”
Criminality
Many characteristics
such as physical appearance, I.Q., and psychological abnormalities, which
probably have genetic components, often affect a person’s interactions in the
social environment and thus may help to determine whether he will become a
criminal. One major area of investigation has been relating chromosome
constitution to aggressiveness and crime. Females generally have two X
chromosomes, while males usually have an X and a Y chromosome. However, it is
estimated that somewhere between 0.05 and 0.4 per cent of all males have an
extra Y chromosome; that is, their chromosome constitution is XYY. There are
reports (Hook, 1973; Montagu, 1968) that this extra Y chromosome may make the
person more aggressive, and that males may be more aggressive than females
because the females have no Y chromosome. There are also reports that there is
a higher frequency of XYY males among criminals than in the population at
large. However, the absolute numbers being compared are small, and there are
some procedural problems with the studies. Montagu points out that XYY males
are usually taller, and argues that as children they might learn to be more
aggressive because of the manner in which other children respond to their
height.
The evidence relating
an extra Y chromosome to aggression and crime is suggestive, but still very
speculative. In a review of XYY males, Owen (1972) concluded that “no consistent
personality or behavioral constellation has been successfully predicted from
the XYY complement,” while Hook (1973) concludes that “There is a definite
association between the XYY genotype and presence in mental-penal settings,
but both the nature and extent of this association are yet to be determined.”
“Mental-penal settings” include hospitals for the criminally insane and security
wings in hospitals for the mentally retarded.
Other Psychopathologies
The data on genetic
influences on neurosis is sparse and confusing, but “the overall evidence
points to the likelihood that heredity plays a role in the development of
psychoneurotic symptoms” (Rosenthal, 1971, p. 144). However, the genetic
component in neurosis seems small, particularly in comparison to environmental
factors.
A person’s physiological
response to alcohol may have a genetic component. Wolff (1972) has reported
ethnic differences in reactivity to alcohol, independent of where the people
lived. He found that Japanese, Taiwanese, and Koreans, after drinking amounts
of alcohol that have no detectable effect on Caucasoids, showed mild to moderate
symptoms of intoxication. Rosenthal (1971, p. 149) has reported studies that
suggest that some aspects of drinking behavior may be heritable.
Intelligence
Although not a
psychopathology, it seems relevant to mention the intelligence and race issue.
Are some races of humans genetically more intelligent than others? The answer
is that for many reasons no-one knows. First, it is not clear exactly what
intelligence is, nor are we sure of the best ways to measure it. I.Q. tests are
not as reliable or valid as desired and have strong cultural biases. Many of
the items and terminology on traditional I.Q. tests favor one cultural group
over another. A second problem is that defining and identifying members of a
“pure” race is often a difficult task. The major problem, however, is that the
postbirth experiences of members of different races are so different that it is
almost impossible to factor out environmental influences. There is no way that
a black and a white in the United States can have the same social learning
experiences. Cultural differences, prejudices, different social environments,
varying expectancies of success or failure, and different diets are just a few
of the many confounding variables. Thus the race and intelligence issue is
unresolvable at the present time, and perhaps this is socially and politically
desirable (see Bodmer & Cavalli-Sforza, 1970).
PREPAREDNESS OF LEARNING
The behavior of
simple organisms is controlled largely by instinct. As organisms become more
complex, learning plays a larger and larger role until, in the case of human
behavior, learning far surpasses instinct in importance. Yet man is still a
biological animal, and the instinctual components of the behavior of the human
species must still be accounted for.
Unfortunately in
the last few years the pendulum has swung to the side of instincts. Several
popular books have tried to account for large parts of human behavior in terms
of just a few instincts such as territoriality and bonding. But such oversimplifications
are merely crude analogies that have not given appropriate weight to the role
of learning. It may be fun to draw parallels between human behavior and the
behavior of apes, but one must be careful about pushing this too far.
Seligman (Seligman,
1970; Seligman & Hager, 1972) has pointed out how animals can learn some
things more easily than others. For example, Thorndike had no trouble training
cats to pull strings to get out of puzzle boxes, but he had considerable trouble
trying to train them to scratch or lick themselves to get out of the boxes.
Seligman argues that we must consider how evolutionarily prepared the animal
is to learn different responses. In the case of Thorndike’s cats it would
be evolutionarily advantageous for cats to be able to learn to associate manipulating
objects with escape. But there seems to be little evolutionary pressure for
the cats to quickly associate licking with escape. For how often in the natural
world is licking related to escape? Similarly, Chapter 5 discussed how rats
can be conditioned to associate gastric upset to taste stimuli more easily
than to audio-visual stimuli, while avoidance reactions produced by electric
shock were more easily conditioned to audiovisual stimuli than to taste stimuli.
Seligman suggests
a preparedness dimension, a continuum which specifies for different types
of learning the evolutionary constraints on the animal’s ability to acquire
the particular learning. At one end of the continuum are those behaviors for
which the animal is prepared. This means
that the biology and genetics of the animal facilitate this particular type
of learning. At the other end of the continuum are those behaviors for which
the animal is contraprepared
—i.e., the animal’s biology and genetics impede
the learning. Different types of learning lie at different points on the continuum.
Near the middle of the continuum is the neutral point at which the animal
is unprepared — his
natural history has no bearing on the learning. Instinct, according to this
formulation, is an extreme form of preparedness. Similarly, rats are prepared
to associate taste with illness, but perhaps contraprepared to associate taste
with electric shock.
Seligman suggests
that prepared associations are relatively inflexible and resistant to extinction,
to such a degree that the behavior may acquire some autonomy in itself. Also,
the learning does not involve cognitive processes, whereas unprepared associations
are more flexible in nature, extinguish more readily, and often are mediated
by consciousness, attention, and expectations.
It is not clear exactly
what kinds of things humans are prepared to learn, although Seligman suggests
that language acquisition might be one example. Seligman (1971) has also argued
that man is prepared to learn certain phobias, such as fear of snakes, spiders,
or heights. This would explain why such phobias are so common, are acquired
so rapidly, are so resistant to extinction, and why they are probably outside
conscious control. (The point about conscious control will be elaborated on
in the next chapter.)
Gray (1971, p. 15),
on the other hand, argues that man has an innate fear of snakes that does
not develop until the child is several years old, although Gray allows for
a possible learning component. Similarly fear of the dark and of some animals,
such as dogs, may have an innate basis, according to Gray.
PHYSIOLOGICAL BASES OF SCHIZOPHRENIA
What is
physiologically different about people classified as schizophrenic? Which of
these differences might have been a cause or a partial cause of the person
becoming schizophrenic? How can you tell whether a variable was really a
partial cause or was actually a result of schizophrenia? That is, if there is a
brain chemistry difference between schizophrenics and normals, did the
difference in brain chemistry produce schizophrenia or did schizophrenia
produce the difference in brain chemistry?
Questions such as
these have led researchers and theorists in a host of different directions
looking for physiological bases of schizophrenia. There has been considerable
research into the biochemical differences in the blood and brain of schizophrenics
and possible enzyme deficits or abundances that might result in abnormal experiences
and behaviors (e.g., Mandell et al., 1972). We will consider here just a sample
of the research on the bases of schizophrenia, remembering that “schizophrenia”
is a very broad, poorly defined category.
Mednick (1971) studied
over 200 Danish children whose mothers were schizophrenic. Of the children
in this group who later became schizophrenic it was observed that for 70 per
cent of them their mothers had experienced complications during pregnancy
or birth (e.g., oxygen deficiency, prolonged labor). Also the galvanic skin
response (GSR) of these children did not habituate as fast as that of normals
to irritating noise, and they took longer to extinguish a conditioned GSR.
This suggested to Mednick that the birth complications damaged the body’s
ability to regulate stress-response mechanisms (perhaps through damage to
the hippocampus); this deficit then may lead to schizophrenia. For example,
these children may be more sensitive to threats and stress, and may learn
avoidance responses, such as the thinking of bizarre thoughts. However, birth
complications did not produce these results in children whose mothers were
not schizophrenic, suggesting that the effect requires an interaction between
birth complications and a genetic variable related to schizophrenia.
In the previous chapter
we discussed the fact that areas of the brain when electrically stimulated
produce a reinforcing effect, and that these areas may underlie reinforcement
in general. Stein and Wise (1971) have offered a model of schizophrenia that
is based on deficiencies in these reward systems. They noted that two primary
symptoms of schizophrenia are a deficit in goal-directed thinking and a deficit
in the capacity to experience pleasure. They suggest that both of these symptoms
could result from impairment of one of the physiological reward systems. Stein
and Wise believe that the aberrant metabolite that produces these deficits
is 6-hydroxydopamine (6-H). 6-hydroxydopamine injected into rats will decrease
the rate of self-stimulation in the relevant brain area (medial forebrain
bundle). It may be that chlorpromazine, a drug often used in the treatment
of schizophrenia, produces its effects by preventing the reward area from
taking in 6-H. Chlorpromazine, then, would not be as useful after the reward
area had suffered irreversible damage. Finally, a particularly odorous substance
(trans-3-methyl-2- hexenonic acid) often found in the sweat of schizophrenics
is a possible biochemical result from 6-H. A question about the Stein and
Wise model is how many people classified as schizophrenic would have this
particular type of deficit in experiencing pleasure (e.g., Watson, 1972).
Another possibility
is that some schizophrenics are in a continuous state of overarousal (Depue
& Fowles, 1973; Lang & Buss, 1965). Since performance, and probably
also learning, is best at some optimal level of arousal, people who are continually
overaroused would be deficient in simply learning to behave in the most productive
manner. According to this theory, the effect of drugs such as chiorpromazine
is to reduce the arousal level to some optimal point. Too much of the drug,
however, would result in passing the optimal point and making the person’s
arousal level too low.
There is no single
physiological explanation for the wide range of behaviors that are called
schizophrenic: For any one individual, the explanation of his schizophrenic
behavior may be a very complex interaction of many different variables including
genetic, current physiological, and social factors. In the next section we
will see a couple of the ways in which physiological and psychological variables
interact.
INTERACTIONS
When investigating
variables that affect behavior it would be nice if we could break them down
into two distinct lists: physiological
variables, related to the
physical construction of the organism, and environmental variables, related to the experiences of the organism. But, in fact, these two
sets of variables continuously interact, producing changes in each other in
quite complex ways. Here we will consider a few of these interactions and their
effects on personality.
Rosenzweig, Bennett,
and Diamond (1972) have investigated the effects of different types of living
environments on the brains of rats. They raised rats in three different environments:
(a) an impoverished environment of simple cages with one rat per cage, (b)
a standard environment of simple cages with about three rats per cage, and
(c) an enriched environment with 12 rats per large cage plus playthings in
the cage that were changed each day. The rats in the enriched environment
showed greater brain changes than rats in the other two groups: after 4 to
10 weeks in the enriched environment they showed greater weight of the cerebral
cortex, greater thickness of the cortex, greater total activity of acetyicholinesterase
(the enzyme described in Chapter 2 that breaks down the transmitter substance
acetylcholine) but less activity per unit of tissue weight, more glial cells,
no more nerve cells per unit of tissue but larger nuclei and cell bodies (indicating
higher metabolic activity), and increases in the ratio of RNA to DNA. The
greatest brain differences were found in the occipital cortex (the visual
cortex). The same effects of the enriched environment can be shown with adult
rats, but they generally require a longer exposure to the enriched environment
to get the maximum effect. Rats kept in an outdoor setting for a month show
even greater brain changes than rats from an enriched environment:
How do such brain
changes affect later behavior? Rosenzweig and his associates report that the
experience in the enriched environment facilitates later learning, but the
effects are often short-lived and depend on the measure of learning, the type
of task to be learned, and the age at which the enriched experience is provided.
The implications for humans are still quite speculative; it would be interesting
to investigate questions such as whether musicians, as a group, would show
an enhanced development of the auditory cortex.
Malnutrition has
been clearly shown to retard mental development in nonhumans. There is also
suggestive, although not conclusive, evidence (Warren, 1973) that malnutrition
is a contributing factor to mental deficiency in humans, preventing the person
from realizing his full genetic potential (see Kaplan, 1972). Malnutrition
may produce this effect in two ways. First, malnutrition in childhood, from
the prenatal period (via the intrauterine environment) through the first years
of life, may impair physiological development and produce mental deficiency.
Many of these effects are irreversible, although some may be reversible. The
I.Q.’s of mentally retarded children have been raised 10 points by changing
what they eat. Low income families are, of course, more affected than others
because they often cannot provide adequate nutrition (especially adequate
amounts of protein) for their children. The second way in which malnutrition
impairs mental development is by indirectly impairing learning. For example,
poorly nourished children may miss more days of school because of illness
or may be distracted from learning in school as the result of being hungry.
The eating habits of parents may thus lead to malnutrition and mental deficiency
in their children, affecting a whole range of the children’s behaviors, particularly
learning potential. Kaplan pointed out that 75 per cent of pre-school children
in South America, Asia, and Africa are underweight for their age; in 1968
more than 10 million Americans were affected by hunger and malnutrition.
Holmes and Masuda
(1972) report on a series of investigations that relate psychological events
to physical illness. Although many illnesses clearly have a physical cause,
such as a virus, when a person acquires the illness, or perhaps even
whether he acquires it, may be influenced by psychological
variables. Holmes and Masuda suggest that events in life trigger illness because
the effort required to cope with these events weakens resistance. For example,
they found that colds were triggered by events such as a visit from a mother-in-law,
a change in job, or the birth of a child. After the subjects recovered from
their colds, simply talking about the particular event often renewed the cold
symptoms. They were also able to relate life events to other illnesses, including
tuberculosis, heart disease, skin disease, and hernias. The common theme to
illness-triggering events is that they are important changes in the life pattern,
either positive or negative. For American subjects (there are cultural differences)
the top 10 events in terms of triggering illness are as follows, in order
of decreasing importance: death of spouse, divorce, marital separation, jail
term, death of close family member, personal injury or illness, marriage,
being fired, marital reconciliation, and retirement. Thus it appears that
coping with life changes reduces a person’s resistance to illness, particularly
if the person has not learned effective ways of coping with these changes.
INDIVIDUAL DIFFERENCES IN CONDITIONING
Since most behaviors
are learned, the ease with which an individual can be conditioned is an important
personality characteristic, particularly since it appears that there are
individual differences in conditionability (Franks, 1964; Nebylitsyn &
Gray, 1972). (Personality theories based on such differences will be discussed
later in this section.) Assuming that there are individual
differences in conditionability, the rate at which an individual is conditioned
would be a combination of his conditionability and situational factors. A
general issue, as Franks points out, is how useful the concept of a general
factor of conditionability is,
for if the situational
factors can account for most of the differences in learning rate, then
conditionability may be an insignificant factor. Much research is still
required before this issue can be resolved. In Chapter 3 we discussed how
organisms seek out activities that involve a certain amount of complexity.
Events that provide such complexity thus function as incentives and
reinforcements and underlie a considerable amount of learning. Humans have a
need for a certain amount of complexity. We need variation and seek it out
through activities such as playing, reading, daydreaming, and the use of drugs.
Maddi (1961) suggested that the need for variation is an aspect of personality
and that “individuals show reliable differences in the intensity and quality of
their variation-seeking.” A moderate amount of this need for variety is
probably desirable; people without such a need may acquire undesirable
characteristics, such as rigid defensiveness. However, Maddi suggests that too
much of this need may produce an undesirable degree of behavioral instability.
Maddi also points out that there are great differences between individuals in
the ways in which their need for variety is expressed. Some individuals seek
out variation in a relatively passive manner, such as through reading, whereas
others are more active.
From our previous
discussion of complexity it is clear that a person’s experiences affect the
level of complexity that he seeks. Since we know that learning affects the
level of variation-seeking in humans, and since we assume that a moderately
high level of this need is desirable, it might be useful for parents and teachers
to build a moderate level of this need into children. Maddi suggests that
this might be accomplished by eliciting and rewarding unusual responses in
the child and by introducing the child to a wide range of experiences.
Toward the end
of his life Pavlov put considerable thought into how his research
on conditioning and the nervous system might extend to personality and psychiatry
(see Franks, 1970). The system in which animals respond to environmental stimuli
such as sights and sounds and learn associations between these stimuli was
called by Pavlov the first signaling system. Pavlov
assumed that man, but no other animal, has a second signaling system in
which words stand for the stimuli of the first signaling system. This second
signaling system is responsible for language and speech and underlies attributes
Pavlov considered uniquely human, such as the human forms of communication.
Pavlov believed that
many psychological problems were due to an imbalance between the ideal amounts
of excitation and inhibition in the nervous system. Thus treatment might consist
of sleep therapy to protect a person from overstimulation or use of stimulant
drugs to increase excitation in a chronically tired or apathetic person. Pavlov
suggested that excessive or prolonged stimulation of the nervous system would
produce protective inhibition,
a form of inhibition which
serves to protect the nervous system from too much strain. People with weak
nervous systems — systems
particularly sensitive to stimulation are assumed to generate more protective
inhibition than others. Such people with weak nervous systems are considered
subject to schizophrenia, which, accordingto Pavlov, results from excessive
protective inhibition in the cerebral cortex. Pavlov suggested that as a result
of this protective inhibition, schizophrenics condition slower than normals.
Pavlov’s psychological
theorizing drew on neurophysiological concepts such as the irradiation of
excitatory and inhibitory waves through the cortex. The neurophysiological
aspects of the theories have not held up as more has been learned about cortical
functioning, but many of Pavlov’s observations and suggested relationships
have proved valuable and may still generate useful ideas (see Nebylitsyn &
Gray, 1972).
Eysenck (see Eysenck,
1967; Eysenck & Beech, 1971) has suggested a personality model with two
basic dimensions: neuroticism
and introversion-extroversion. People low on the neuroticism scale are usually
calm, not easily aroused, and have fairly stable emotions, whereas people
high on this scale are easily aroused, moody, restless and have labile emotions
According to Eysenck, the neuroticism scale is a measure of an innate reactivity
and lability of the autonomic nervous system. People high on the neuroticism
scale presumably have a more labile, easily aroused autonomic system that
is more susceptible to the conditioning of fear and neurotic disorders. These
people react more strongly to emotion-arousing stimuli and experience an aversive
reaction to more stimuli than people low on the neuroticism scale.
People on the introvert
end of the introversion-extroversion scale are usually quiet, introspective,
and reserved; they like a well-ordered life and keep their emotions under
close control. People at the extrovert end of he scale are usually sociable,
impulsive, and easygoing; they
crave excitement and like change. It is assumed that the introvert has a higher
state of cortical arousal, often a high excitatory state and a low inhibitory
state, and that this cortical excitation inhibits lower brain centers and
many behaviors, thus making the person more introverted. The high level of
cortical arousal also presumably makes many forms of conditioning easier for
the introvert than for the extrovert, again bringing more behavior under learned
control. The introvert has a low sensory threshold which produces stimulus
avoidance and makes more stimuli painful than is the case for the extrovert.
.
The extrovert, on
the other hand, has a lower level of conditionability and a lower state of
cortical arousal. This low excitatory and high inhibitory cortical state thus
frees lower brain centers and inhibits fewer behaviors associated with these
lower centers. (A parallel is that alcohol may depress cortical activity but
disinhibit many behaviors.) The extrovert has a higher sensory threshold than
the introvert, which results in stimulus hunger and a relative disregard of
painful stimuli. Since the conscience is assumed to be a function of learning
and the extrovert conditions more slowly than the introvert, the extrovert
generally has less conscience.
Eysenck thus can
classify a person’s personality according to where its components fall along
his two dimensions, and his research has revealed a number of generalities.
Eysenck has found psychopaths and some criminals to be high on the neuroticism
scale and highly extrovert, while hysterics are also high on neuroticism but
are intermediate on the introversion-extroversion scale. There are considerable
data supporting Eysenck’s theory, but also considerable data that tend to
refute it. Research on the relationship between conditionability and introversion-extroversion
is somewhat ambiguous and is based on only a couple of different types of
conditioning tasks. Thus Eysenck’s theory should be considered tentative,
and of course oversimplified, until we can factor out the host of other relevant,
but generally confounding, variables relating personality and learning.
Janet Taylor Spence
offered a theory relating anxiety, as a personality variable, to learning
(J. T. Spence, 1963). She composed a questionnaire called the Manifest Anxiety
Scale (MAS) consisting of about 50 items drawn from the personality questionnaire
of the Minnesota Multiphasic Personality Inventory (MMPI). These items were
judged by clinicians to be indicative of manifest anxiety. (Eysenck would
say that MAS measures a combination of neuroticism and introversion).
People who differ
in their scores on the MAS are assumed to differ in anxiety. This anxiety,
according to the Hull and K. W. Spence theory, feeds into a general non-specific
drive which energizes ongoing responses. It is then assumed that in simple
learning situations in which there is a single or highly dominant response
tendency, such as respondent conditioning or paired-associate learning with
little intralist similarity of material, high MAS people will perform at a
level superior to that of low MAS people. This is because the anxiety increases
the drive, which then strengthens the correct responses. On the other hand,
in tasks where there are many response tendencies and the correct response
is relatively weak, high MAS people will show inferior performance to low
MAS people, at least in the initial stages of learning. For here the anxiety
increases the drive which energizes incorrect response tendencies as well
as the correct response. An increase in the incorrect responses, particularly
if they were somewhat dominant, will interfere with the correct response and
impair performance. For example, a high MAS student might do worse than a
low MAS student on a multiple choice test where the correct response is not
immediately obvious and incorrect responses are interfering with recall of
the correct response.
Psychological stress,
however, affects more than just the drive level. J. T. Spence suggests that stress might also
produce changes in effort, attention, and fear of failure, among other things.
Thus, as psychological stress increases, performance might first improve as
the result of increased effort and then decline as anxiety and irrelevant
responses are aroused.
Predictions relating
MAS and learning have been supported in a number of studies, but many other
studies have failed to confirm them. J.
T. Spence (1963, p. 13) responds that:
This inconsistency
in findings may in part be due to the fact that only a minor part of the performance
differences among subjects is attributable to variations in MAS scores, other
characteristics such as differences in learning ability among the subjects
playing a more important role. In addition, our theory is undoubtedly incomplete
in the variables it specifies, both with respect to task variables and to
properties other than drive level which may differentiate groups with extremely
high scores on the MAS from those with extremely low scores.
There are, of
course, other interpretations of the data generated by studies of the J. T. Spence theory. For example, Saltz (1970) suggests that people shown
to be highly anxious on the MAS scale show disruption of learning under
conditions of failure-induced stress, but not necessarily under pain-induced
stress, while people low on the MAS scale show disruption of learning under
pain-induced stress, but not necessarily under failure-induced stress.
According to Saltz, the MAS is not simply a measure of anxiety but is also a
measure of whether a person is disrupted more by failure or by pain-induced
stress. The MAS “represents an index to the types of situations that constitute
stress for different persons.”
EXPERIMENTAL NEUROSIS
In the remainder of
this chapter we will consider some learning analogues of abnormal behavior.
That is, we will discuss learning experiences that produce behaviors that have
marked similarities to some forms of psychological problems often dealt with in
clinical situations. The similarities suggest that the psychological problem
may have its roots in the corresponding learning experience. But it should be
kept in mind that this is not necessarily so; much of the learning research has
been done with animals and then extrapolated to humans, and there are always
problems in oversimplifying such extrapolations.
When an animal is
exposed to a strong conflict situation, he often ends up in a disturbed state
known as experimental neurosis
(Liddell, 1956; Masserman,
1967). This state is characterized by a wide range of behaviors that may include
some of the following: excessive anxiety, avoidance of the experimental situation,
trembling and tics, increase in blood pressure and heart rate, excessive vocalization,
diarrhea, drastic changes in the animal’s social interactions with other animals,
and responding to imaginary stimuli (e.g., the monkey that brushes off insects
that are not there). One type of conflict that often produces experimental
neurosis is known as approach-avoidance conflict, in which the animal is caught between tendencies
to make a particular response and tendencies not to make the response. For example, in an early experiment in Pavlov’s
laboratory a dog was trained to salivate to a circle but not to an ellipse,
a relatively easy discrimination for the dog. Then the experimenters gradually
decreased the longer axis of the ellipse so that it approached being a circle.
The dog had little trouble with the discrimination until the ratio of the
semi-axes was 9 to 8. At this point the discrimination was quite difficult
and the dog was in a conflict between salivating and not salivating to the
ellipse. After three weeks with this conflict, the dog developed experimental
neurosis. Now he no longer stood quietly in the test apparatus but struggled
and howled. At this point he could no longer perform even the simplest of
the circle-ellipse discriminations. Another approach-avoidance conflict that
has produced experimental neurosis involved training a cat to approach a food
dish for food and then shooting a puff of air in its face. The cat is caught
in a conflict between approaching the dish for food and avoiding the dish
because of the puff of air (Masserman, 1967). Similarly monkeys may develop
experimental neurosis if toy snakes suddenly appear in the food box, as many monkeys appear to have
an innate fear of snakes. Many of the behaviors of these experimental neuroses
occur only in the presence of the test situation, whereas other behaviors
carry over to different situations.
Masserman has also
produced experimental neurosis with approach- approach conflicts, a situation where the animal has response tendencies to make approach
responses toward two different and mutually exclusive goals. For example,
a hungry female cat in heat might be forced to choose between food and a male cat, or a monkey might be made to choose between two favorite foods.
Although experimental
neurosis is generally produced by a conflict situation, researchers have reported
it following a range of other treatment procedures. Pavlov suggested five
ways of producing this type of breakdown in his dogs (see Franks, 1970): (1)
use of intense stimuli such as loud explosions or swinging the dog’s platform;
(2) increasing the interstimulus interval (ISI) — the
time interval between the CS and UCS; (3) use of difficult discriminations
such as the circle-ellipse described above; (4) continually changing which
stimuli the dog should respond to and which he should not respond to; and
(5) subjecting the dog to physical stresses such as disease, accident, or
surgery.
Because the concept of neurosis has not been well specified at either the animal or human level, it is difficult to decide what types of breakdowns in animals should be called experimental neurosis or how similar animal neuroses are to human neuroses. However, the parallels between experimental neuroses in animals and neuroses in humans are quite striking and probably involve many common elements.
Masserman has investigated
a number of ways of treating experimental neurosis in animals. One procedure
consists in resolving the conflict by satisfying one of the needs, such
as feeding a cat caught in an approach-avoidance conflict relative to its
food dish. A second way is to force the animal to resolve the conflict,
such as mechanically forcing the cat to approach the food dish after the
air puff is gone. Exposing the neurotic animal to a normal animal modeling
the desired behavior may facilitate breaking down the neurosis. On the negative
side, Masserman did not find electroshock therapy to be a useful treatment
procedure.
SUPERSTITION AND LEARNED HELPLESSNESS
In Chapter 6 we discussed the effects of contingent events such as reinforcement and punishment. We defined a dependent event to mean that the event occurs only if the animal makes a specified response or set of responses, but the environment also provides many non-dependent events — events that occur independent of what the person or animal is doing. If the non-dependent event is pleasant (a potential positive reinforcer), the occurrence of the event may result in superstitious behavior;
whereas if the
non-dependent event is aversive (a potential positive punishment), the event
may result in learned helplessness.
These two effects are
discussed below.
Consider a hungry
pigeon in a test apparatus in which at random intervals a grain of food
drops into the food well. The food is presented aperiodically, independent
of what the pigeon is doing; i.e., the food is not dependent on the
pigeon’s behavior. But the pigeon will be doing something when the food
appears, so that some behavior will be reinforced. Through such chance
reinforcements a number of behaviors may be accidentally reinforced. Eventually
the response strength of one behavior will come to dominate and the pigeon
will keep repeating this behavior, which will occasionally be reinforced
by having the food presented non-dependently. Skinner (1948), who studied
pigeons in such situations, found that the pigeons soon came to emit very
stereotyped behaviors which he named superstitions. For example,
one of Skinner’s pigeons “learned” to turn counterclockwise in its effort
to receive food, while another kept thrusting its head into one of the upper
corners of the cage.
Since the animal
in such a situation will not be reinforced each time he performs his superstitious
behavior, the accidentally reinforced behaviors might extinguish before
being reinforced. Thus, to develop superstitions the potential reinforcement
must be presented at intervals shorter than the duration of complete extinction
of the various superstitious behaviors. If this is done, the superstitious
behavior will be reinforced on an intermittent schedule of reinforcement,
which is likely to result in a long time to extinction.
It is easy to see
human parallels of this type of behavior. A person playing a slot machine
may alter the way he puts money in the machine and the way he pulls the
handle if he thinks that doing these things a certain way will bring him
luck. Independent of these behaviors the machine will occasionally pay off
(reinforcement). Such a situation allows the person to develop a superstitious
behavior, such as not looking at the machine while he pulls the handle.
Observation of a gambling casino will reveal a large number of people displaying
their superstitious behaviors at the slot machines. Each person’s superstition
may be unique to him, as each of Skinner’s pigeons had a unique superstition.
Human superstitions
are quite abundant. A college student in an elevator may keep pushing the
button of his floor as if this would cause the elevator to move faster.
A card player may pick up his cards one at a time as if to improve the hand
he was dealt. A businessman may wear a “special” tie when going to an important
meeting.
There are, however,
a number of differences between human and animal superstitions. First, humans,
as opposed to animals, often spend considerable time justifying why they
are not reinforced each time they do their superstitious behavior. (“I have
some questions about that so- called virgin we sacrificed to the volcano
god.” “I lost the golf match today because my lucky hat doesn’t seem to
work two days in a row.”) Second, humans spend more time than animals trying
to convince others to adopt their superstitious behaviors. Children often
carry on many of the superstitions of their parents. Finally, as Herrnstein
(1966) points out, “Human superstition, unlike that of animals, arises in
a social context.” The acquired superstitions in humans are not as arbitrary
as those of animals. Rather they are molded by the person’s culture. Thus,
although it is possible to develop a superstition about Wednesday the 11th,
it is more probable in our culture to be superstitious about Friday the
13th.
A group of investigators
(Overmier & Seligman, 1967; Seligman, Maier, & Geer, 1968) studied
the effects of non-dependent aversive events (electric shock) on dogs’ behavior.
First they periodically gave harnessed dogs electric shock that the dogs
had no control over, as it was non-dependent. No response that a dog could
make would enable it to terminate or avoid the shock. After this experience
the dogs were put in an avoidance task where on cue the dogs could learn
to cross a barrier to escape, and later learn to avoid, foot-shock. Normal
dogs learn this avoidance task quite readily. However, a majority of the
dogs that experienced the non-dependent shock made few escape responses
and basically no avoidance responses. They would often just lie on the grid
floor receiving shock rather than escaping across the barrier. To anthropomorphize,
it is as if the dogs during the non-dependent shocks learned that there
was nothing they could do that affected the occurrence of the shocks, so
that when they were later in the avoidance task they didn’t even try to
do anything about the shocks, although in this situation they could have.
The investigators labeled this phenomenon learned
helplessness, a passive
state resulting from the learning of independence between behavior and the
presentation and/or withdrawal of aversive events. It should also be noted
that other types of pretraining can also impair later learning of avoidance
responses, such as pretraining on an escape procedure that reinforces long
response latencies and interresponse times (Cohen, 1970).
Dogs showing learned
helplessness in the avoidance task can be cured through a guidance procedure
of physically forcing the animal to make the avoidance response. To do this
the dog is put on a leash and literally dragged through the test apparatus
when the cue for the avoidance response occurs. It seems that only by such
physical force can the learned helplessness be readily overcome. Future
research might find other cures.
Interestingly,
if before the dogs are exposed to the non-dependent shocks they are put
in a situation where they learn to press panels to turn off shock, then
it is significantly less probable that they will later develop learned helplessness.
Once the dogs have learned in the panel-press situation that shocks can
be controlled, the experiences with non-dependent shocks do not affect their
later learning of the avoidance task.
To date there have
been only a few controlled studies with humans that have demonstrated a
phenomenon similar to learned helplessness (Dweck & Reppucci, 1973;
Thornton & Jacobs, 1971), but there are a number of human situations
highly suggestive of learned helplessness (see Seligman et al., 1968). For
example, prisoners in Nazi concentration camps often had no hope or control
over what happened to them, and many became passively resigned to everything.
Similarly some mental patients believe that they have no control over their
environment, and their behavior is like that in learned helplessness. Seligman
(1973) suggests that some forms of depression, particularly those that are
set off by external events rather than those that are hormonally or genetically
based, appear very similar to states of learned helplessness. It should
be noted, however, that learned helplessness does not have to be as dramatic
as these examples. Rather it seems that a more moderate form of learned
helplessness is a characteristic of quite a number of people.
One of the most
important principles for parents and teachers is to be consistent in their
dealings with children, for the responses that a parent or teacher makes
to a child should be a function of the child’s behavior rather than of the
adult’s mood. A key part of many, behavior modification programs, such as
contingency contracting, is to build in consistency. If a child in dealing
with a parent or teacher learns that there is little connection between
what he does and how the person will respond to him, the child may develop
some degree of learned helplessness.
By using the information
gained from the dog studies, we might be able to immunize humans against
developing learned helplessness (Seligman, 1969). This would consist of
providing children with many experiences in which they clearly had control
over parts of their environment; that is, they would learn a correlation
between their behaviors and different environmental events. With such pretraining,
and particularly if they have consistent parents, children might be less
affected by the non- dependent aversive events that happen in everyone’s
life.
FRUSTRATION-FIXATION
Similar to learned
helplessness is the type of fixated behavior observed in rats following
frustration (Feldman & Green, 1967; Maier, 1949). These experiments,
originated by N. R. F. Maier, used a Lashley jumping-stand. In this apparatus
a rat is placed on a small platform and jumps toward one of two windows.
If he chooses correctly the window opens and he goes through to an area
where he receives a reinforcement, such as food. If the rat chooses the
incorrect window he hits himself on the closed window and falls four feet
into a net. Rats that are not interested in jumping off the platform at
the windows are goaded into jumping with electric shock or a blast of air.
This apparatus has been used for a variety of learning tasks, such as discrimination
learning where the rat learns to jump to the darker of the two windows regardless
of whether it is on the left or right side on any one trial. Usually there
are two sets of cues to which the animal may respond: position (left vs.
right) and brightness (dark vs. light).
Maier exposed some
rats to an insolvable discrimination task on the Lashley jumping stand.
There was no “correct” response; half of the rats’ responses to any cue
were randomly reinforced and the other half were randomly punished. The
usual result of this frustrating experience was that the rats would adopt
a stereotyped response to a position, such as always jumping to the left
window. Maier called this stereotyped way of responding fixation.
Maier then exposed the
fixated rats to a solvable discrimination problem in which, for example,
the dark door, regardless of position, was correct. Although 15 to 20 per
cent of the fixated rats solved this problem, a task that normal rats learn
fairly readily, the majority
of the fixated rats continued their fixated
behavior throughout training. The fixated behavior is usually a positional
response. However, if rats first learn a discrimination, such as going to
the dark door, before the frustration, then they may later fixate
the earlier discrimination and always jump to the dark door even though
this is no longer correct. The fact that a rat fixates a response (e.g.,
always jumping left) in a solvable discrimination (e.g., dark is correct)
does not mean that on the perceptual side he cannot learn the discrimination.
In fact the fixated rats often do seem to have learned the discrimination.
This is evidenced by the fact that the response latencies are usually shorter
when the fixated response is the correct response (dark door on left) than
when incorrect (dark door on right). Also, when making an incorrect fixated
response the rat will often turn his body during the jump in such a way
as to minimize hurting himself on the window, but in a way so that he couldn’t
possibly go through the window if it were correct. A rat fixated to the
left can even be shown the right window wide open for him to jump through
and he will still jump left. Again, to anthropomorphize, it is as if the
fixated rat ”knows” what the correct response is but can't break his fixation.
The rats’ fixated
behavior is not alterable by simple reinforcement procedures, and punishment
seems only to increase the fixation. Rather, breaking the fixation, like
breaking learned helplessness, involves guidance. The rat must be physically
forced to make the correct response by having his jump physically guided
from the platform.
There are many
explanations for the frustration-fixation effect. Feldman and Green (1967)
suggest that the rat during the insolvable discrimination is in a double
“go—no go” conflict, an approach-avoidance conflict to each of the two responses.
The approach, or “go,” consists of the food reward on half the trials plus
the tendency to escape or avoid the goad shock. The avoidance, or “no go,”
is based on the punishment that the rat receives on half the trials. This
double conflict holds true for both the spatial dimension and the brightness
dimension. There is also conflict about which of these two dimensions to
respond to. According to Feldman and Green the stereotyped response that
develops is due in large part to a general avoidance of one of the stimuli,
which then “pushes” the animal toward the other.
Extrapolating from
rat fixations to parallels in human behavior is difficult for a number of
reasons. First, Maier’s ideas on frustration-fixation have not been adequately
followed up by research on humans despite the wealth of relevant ideas that
Maier has suggested. Second, fixation is just one of a number of different
possible results of frustration, other classic results being aggression
and anxiety. Finally, human behavior is so complex that it is difficult
to factor out fixated behaviors from other high probability behaviors.
However, it does seem likely that a number of human behaviors, such as some forms
of compulsions, obsessions, and ritualistic acts, may be examples of fixations
that follow a particularly bad frustrating situation. Also, successful treatment
of human compulsions by implosive therapy often involves physically helping
the person to make a response that he is avoiding, which parallels the “guidance”
that is used to break fixations. Perhaps some compulsions are a combination
of conditioned anxiety and response fixations. It also may be that frustration
is the common element to a number of phenomena such as fixation, learned
helplessness, and experimental neuroses.
APPROACH-AVOIDANCE CONFLICTS
Approach-avoidance
conflicts have been discussed in the preceding sections in relation to experimental
neurosis and frustration-fixations. In this section we will discuss some
of the experimental analyses of such conflicts, primarily those by Neal
Miller and his associates (Dollard & Miller, 1950; N. E. Miller, 1959).
The following is a fairly common experimental procedure for studying approach-avoidance
conflicts: Rats are first trained to run down an alley for food. Then they
receive shock in the goal box where they previously received the food. The
rats are now in an approach-avoidance conflict relative to the goal box,
a conflict between the response tendency to approach the goal for food and
the response tendency to avoid the goal because of the shock.
The general findings
are that both the tendency to approach the goal and the tendency to avoid
it increase as the goal becomes closer. However, the strength of the avoidance
tendency increases more rapidly as the goal is neared than does the approach
tendency. The avoidance gradient is thus said to have a steeper slope. These
relationships are shown in Figure 7—1. Although the rest of our discussion
will center on this figure, two qualifications should be made. First, the
gradients usually are not linear, but for the types of conclusions that
we will draw, the deviations from linearity do not seem critical. Second,
the gradients need not intersect as they do in the figure; one gradient
could be completely higher than the other and never intersect it. But the
non-intersection case is not interesting for our purposes.
Note that the point
of intersection of the two gradients occurs at some distance X from the
goal. Now assume that the figure is for one of the rats in the food-shock
conflict. When this rat is farther than X from the goal, the approach tendency
is stronger than the avoidance tendency. Hence the rat moves toward the
goal. However, when the rat is nearer than X to the goal, the avoidance
tendency is stronger than the approach tendency. In this case the rat moves
away from the goal. At X, of course, the tendencies balance out. All of
this results in the rat’s tending to fluctuate around a distance about X
from the goal, but in a state of conflict.
A human parallel,
would be the young child at the beach who is both fascinated by the ocean
(approach tendency) and afraid of the waves (avoidance tendency). The child
may run back and forth from the edge of the water, vacillating around a
point where his gradients cross.
One of Freud’s
important contributions to psychology was his emphasis on behavior as often
being the result of a conflict, an example being the conflict between approach
tendencies of the id and avoidance tendencies from the superego. Using our
approach-avoidance conflict model we may also look at Freud’s idea of displacement, the shift of energy from one outlet to another.
Consider a small boy who is punished by his father. The boy may wish to
strike the father, but past experiences have built in inhibitions against
this, so the boy displaces his aggression and hits the family dog. From
the conflict model we can see the boy in a conflict between the approach
tendency to hit his father and the avoidance tendency of fear of the father.
Again, we would expect vacillation at distance X from the father. But this
instance involves psychological distance rather than physical distance.
That is, in Figure 7—1 the bottom axis would now be a dimension of stimulus
similarity, rather than simple physical distance. Moving along this dimension
from the goal of father, we may pass points corresponding to mother, brother,
sister, dog, cat, and teddy bear. In our case of the boy hitting the dog,
the intersection of the gradients occurs at an X-psychological distance
from the father at a point corresponding to the dog.
There are basically
two ways of resolving an approach-avoidance conflict (i.e., getting the
subject to the goal). First, we can increase the approach tendency, such
as by increasing the incentive to reach the goal and/or increasing the relevant
drive (e.g., making the rat more hungry). This generally raises the approach
gradient without altering its slope significantly. As the approach gradient
is raised, the point of intersection of the two gradients moves nearer the
goal. (This can be seen by laying a pencil along the approach gradient in
Figure 7—1 and slowly moving the pencil up the page, keeping it parallel
to the original approach gradient.) The second way of getting the subject
nearer the goal is by decreasing the avoidance gradient, such as by extinguishing
or counterconditioning fear. This lowers the entire avoidance gradient and
moves the intersection point closer to the goal. (The pencil demonstration
may be used again, moving the pencil down the page parallel with the avoidance
gradient.)
The height on the
vertical axis of Figure 7—1 that corresponds to the point of intersection
of the gradients is a rough measure of the amount of anxiety the conflict
produces. The higher the point, the greater the anxiety. Moving the subject
nearer the goal, by either raising the approach gradient or lowering the
avoidance gradient, will raise the point of intersection and thus the anxiety.
However, raising the approach gradient will raise the point of intersection
more than lowering the avoidance gradient will. Thus in many practical situations
it may be preferable to lower the avoidance gradient through procedures
such as desensitization.
MASOCHISM
In Chapter 5 we
discussed how a stimulus may become a conditioned reinforcer if it is paired
with a reinforcer. It may also be possible for a stimulus that is naturally
aversive to become a conditioned reinforcer, if the reinforcing effect is
dominant to the aversive effect. If this should be true, we could have an
animal working for a stimulus because of its conditioned reinforcing properties
even though the stimulus is aversive. This is relatively easy to do, such
as producing a dog that will bar-press to receive electric shock because
the shock had previously been paired with receiving food. In this situation,
the conditioned reinforcing properties of the shock, if no longer paired
with food, should soon extinguish. This type of situation may be an experimental
analogue of masochism, although there are other analogues and theories (see
Dreyer & Renner, 1971).
Ayllon and Azrin
(1966) demonstrated this effect in an experiment with three female schizophrenics.
The subjects were first trained to pull . either
of two levers for tokens that could later be exchanged for back-up reinforcements.
After a number of such sessions an annoying buzzer was made contingent on
pulling one of the levers. This caused the subjects to emphasize pulling
the other lever, and the buzzer was thus demonstrated to be punishing. Then
the buzzer and tokens were paired by having one lever produce the buzzer
and tokens and the other lever produce nothing. After this pairing the subjects
were given two buttons to push. One button produced the buzzer, while neither
button yielded any tokens. In this situation the subjects preferred to press
the button that produced the aversive buzzer, even though they received
no tokens. The buzzer here seems to have become a conditioned reinforcer.
Following are some
naturalistic examples of this phenomenon. A mother spanks her child, but
afterward, so that the child will not feel a loss of love, the mother hugs
and cuddles the child. This relatively common practice may produce masochistic
tendencies by pairing spanking and caressing, particularly if the mother
is more loving after a spanking than she usually is. Or a student’s antics
in the classroom may result in punishment from the teacher but social rewards
from his peers. If this pairing occurs often enough and if the social rewards
are powerful enough, the student may soon work simply for the conditioned
reinforcement of the punishment from the teacher.
REPRESSION
Often, unwanted
thoughts that create anxiety are kept out of consciousness. This effect,
known as repression, can be thought of as the inhibiting of the response
of thinking particular thoughts that make the person uncomfortable or unhappy
(Dollard & Miller, 1950). If the thought produces anxiety, then repression
of the response will be anxiety-reducing, and thus repression is reinforced.
Repression is simply an avoidance response.
In this chapter
we have discussed a few of the learning-related variables and phenomena
that probably should be part of a conceptualization of personality. Much
more could be included, and this chapter is just a sample. What will be
relevant in the future will depend in part on the directions taken by learning-oriented
personality theorists.
SUMMARY
In this chapter
personality is discussed from the viewpoint of learning, under two main
divisions: (1) those genetic and physiological variables that affect a person’s
behavior directly or predispose him for certain types of learning; (2) specific
types of behavioral abnormalities that result from certain learning experiences.
Human behavior
is a highly complex interaction between hereditary influences (nature) and environmental influences (nurture) including
learning. Therefore it is difficult to separate these two classes of variables,
especially when trying to relate them to poorly defined classifications
of behavior, such as schizophrenia. There are environmental effects on the
physiology of an organism which then affect the organism’s learning and
behavior. Rats that are kept in more enriched environments have brains that
are, among other things, larger and heavier. These brain differences, although
reversible, may affect learning. Malnutrition clearly retards mental development
in nonhumans and probably contributes to mental deficiency in humans as
well. On the other hand, there is research suggesting that the severity
of a person’s physical illness, or even whether he acquires the illness
at all, may be due to psychological variables, primarily major changes in
the person’s life.
There is considerable
evidence that there is a genetic component in some forms of mental illness,
such as schizophrenia and manic-depressive psychosis. There is also a possible
genetic factor in some forms of criminality, in psychoneurosis, and in alcoholism.
There is no valid evidence that any particular race is genetically more
intelligent than any other race. Research into possible physiological bases
of schizophrenia has suggested biochemical imbalances in blood or brain,
enzyme imbalances, inability to regulate stress-response mechanisms, impairment
of a physiological reward system, and a continual state of overarousal.
There are probably several different physiological deficits that may lead
to behaviors we call schizophrenic.
Man as a species
has apparently been selected in the evolutionary scheme to adapt culturally
rather than genetically. Adaptation by cultural changes is more effective
for humans than adaptation by genetic changes. Man may also be evolutionarily
prepared to learn certain things, such as language, and perhaps some phobias,
such as fear of snakes.
An account of personality
should include individual differences as they relate to learning, for example,
individual differences in the ease with which a person can be conditioned.
Several theorists have suggested personality variables that relate to what
and how the person will learn:
Maddi suggested
that people differ in their need for variation and in the types of stimulus
complexity that they seek. Pavlov argued that many psychological problems
are due to an imbalance between the ideal amounts of excitation and inhibition
in the nervous system, which then affects learning. Eysenck proposed a personality
classification of basically two dimensions — neuroticism and introversion-extroversion
— and he related these dimensions to learning.
Spence related anxiety, as measured by the Manifest Anxiety Scale (MAS),
to learning by suggesting that anxiety feeds into a general non-specific
drive.
The chapter concludes
with a number of learning paradigms and behavioral abnormalities suggestive
of some common human psychological problems. The topics include experimental
neurosis, superstition, learned helplessness, frustration-fixation, approach-avoidance
conflicts, masochism, and repression.
Dollard, J., &
Miller, N. E. Personality and Psychotherapy. New York: McGraw-Hill,
1950. .
Kimble, G. A. Hilgard
and Marquis’ Conditioning and Learning. New York: Appleton Century-Crofts,
1961, Chapter 14.
Lundin, A. W. Personality:
A Behavioral Analysis. New York: Macmillan, 1969.
Maher, B. A. Principles
of Psychopathology: An Experimental Approach. New York: McGraw-Hill,
1966.
Rosenthal, D. Genetics of Psychopathology. New York: McGraw-Hill, 1971.