the psychologist does not, therefore, have to believe that all human
characteristics are explainable by deterministic, scientific methods.

John Watson's ( 1913) shift in emphasis from the study of the con-
tents of consciousness to the study of behavior, championed direct
observation over the previous dependence on introspection practiced
by 19th-century psychologists. Watson, against the study of con-
sciousness per se, developed his argument against the validity of
introspection as the principal method of psychology. The introspective
analysis of consciousness had yielded a great deal of controversy. The
laboratories of different investigators produced different conclusions
about the same phenomenon depending on the theoretical bias of the
psychologist. Discrepancies were essentially uncheckable because they
depended on the introspection of the trained observers of each of the
laboratories. There was no agreed upon basis by which hypotheses
could be tested.

Watson's idea that behavior had to be the data of psychology re-
quired the additional assumption that the sense data of the observer of
behavior was the key element in the process. What the observer could
see, hear, smell, and so forth, could be confirmed by another observer
and consensus could easily be reached regarding the nature of the data
under examination. This, in turn, meant that the controlled experiment
could become the principal method of the behaviorist. Watson was
careful to indicate that this approach did not eliminate the fact of either
the presence or absence of consciousness, which was, by definition,
unobservable.

Karl Lashley ( 1923), working 10 years after Watson's position was
initially published, held that science, defined as the examination of
sense data via an objective method such as the experiment, is not
appropriate for an examination of felt experience and, therefore, of
consciousness. Later, Lashley, Tolman ( 1927), and Hull ( 1943), al-
though not denying its existence, were also to exclude consciousness
from the possibility of fruitful examination by the methods of science.

Despite those reservations concerning the use of experiment to
answer questions about complicated human functions by some of the
leading psychologists of the early part of the century, Carl Iver Hov-
land and his associates ( Hovland, Janis, & Kelley, 1953; Hovland,
Lumsdaine, & Sheffield, 1949) launched postwar experimental social
psychology with his persuasion experiments while gathering about
him many disciples who were to become the country's leading expo-
nents of experiment applied to social problems. From 1945 until the
mid-1960s, we worked diligently on discovering the details of how
people formed and changed their attitudes, beliefs, and opinions, and
how, or if, behavior followed, and how it could be changed in a good

-2-