142 A Psychological Behaviorism Theory of Personality
We Need Theory Constructed in Certain Ways
and With Certain Qualities and Data
We need theorists to work the field of personality. And they
need to address certain tasks, as exemplified earlier and later.
This is only a sample; other characteristics of theory also
need to be considered in this large task.
Commonalities Among Theories
In the field of personality theory there is much commonality,
overlap, and redundancy among theories. This goes unrecog-
nized, however, because theorists are free to concoct their own
idiosyncratic theory language. The same or related phenom-
ena can be given different names—such as ego, self, self-
concept, and self-efficacy—and left alone as different. Just in
terms of parsimony (an important goal of science), each case
of multiple concepts and unrecognized full or partial redun-
dancy means that the science is unnecessarily complex and
difficult, making it more difficult to learn and use. Unrecog-
nized commonality also artificially divides up the science,
separating efforts that are really relevant. Personality theo-
rists, who are in a disunified science in which novelty is the
only recognized value, make their works as different as possi-
ble from those of others. The result is a divided field, lacking
methods of unification.
We need theorists who work to remove unnecessary the-
ory elements from our body of knowledge, to work for sim-
plicity and standardization in theory language. We need to
develop concepts and principles that everyone recognizes in
order to build consistency and consensus. It is essential also
for profundity; when basic terms no longer need to be argued,
work can progress to deeper levels.
Data of Theories and Type of Knowledge Yielded
A fundamental characteristic of the various theories in per-
sonality is that despite overlap they address different sets of
phenomena and their methods of data collection are differ-
ent. For example, Freud’s theory was drawn to a large extent
from personal experience and from the stated experiences of
his patients. Carl Rogers’s data was also drawn from per-
sonal experience and clinical practice. Gordon Allport em-
ployed thelexical approach,which involved selecting all
the words from a dictionary that descriptively labeled differ-
ent types of human behavior. The list of descriptive words
was whittled down by using certain criteria and then was or-
ganized into categories, taken to describe traits of personal-
ity. This methodology rests on large numbers of people, with
lay knowledge, having discriminated and labeled different
characteristic behaviors of humans. Raymond Cattell used
three sources of data. One consisted of life records, as in
school or work. Another source was self-report in an inter-
view. And a third could come from objective tests on which
the individual’s responses could be compared to the re-
sponses of others. These data could be subjected to factor
analytic methods to yield groupings of items to measure per-
sonality traits.
What is not considered systematically to inform us about
the field is that the different types of data used in theories
give those theories different characteristics and qualities. To
illustrate, a theory built only on the evanescent and imprecise
data of personal and psychotherapy experience—limited by
the observer’s own concepts and flavored by them—is un-
likely to involve precisely stated principles and concepts and
findings. Moreover, any attempt by the client to explain her
behavior on the basis of her life experience is limited by
her own knowledge of behavior and learning and perhaps by
the therapist’s interpretations. The naturalistic data of self-
description, however, can address complex events (e.g.,
childhood experiences) not considered in the same way in an
experimental setup. Test-item data, as another type, can stip-
ulate behaviors while not including a therapist’s interpreta-
tions. However, such items concern how individuals are, not
how they got that way (as through learning).
Let us take as an example an intelligence test. It can predict
children’s performance in school. The test was constructed to
do this. But test data do not tell us how “intelligence” comes
about or what to do to increase the child’s intelligence. For
in constructing the test there has been no study of the causes
of intelligence or of how to manipulate those causes to change
intelligence. The theory of intelligence, then, is limited by
the data used. Generally, because of the data on which they
rest, tests provide predictive variables but not explanatory,
causal variables. Not understanding this leads to various
errors.
The data employed in some theories can be of a causal
nature, but not in other theories. Although data on animal
conditioning may lack other qualities, it does deal with cause-
effect principles. Another important aspect of data used in-
volves breadth. How many different types of data does a
theory draw on or stimulate? From how many different fields
of psychology does the theory draw its data? We should as-
sess and compare theories on the types of data on which they
are based. Through an analysis of types of data we will have
deeper knowledge of our theories, how they differ, how they
are complementary, the extent to which they can be devel-
oped to be explanatory as well as predictive, and also how
they can or cannot be combined in organizing and unifying
our knowledge.