The Psychological Assessment of Political Leadersconflicts. For example, Tetlock (1979) found significantly lower lev-
els of integrative complexity in speeches and public statements of
U.S. leaders during three crises showing Janis's (1972) "groupthink"
pattern that escalated to armed conflict (the Bay of Pigs, crossing the
38th parallel in 1950, and escalating the Vietnam War), as com-
pared to two of Janis's "non-groupthink" crises, where conflict was
controlled (the Marshall Plan and the Cuban Missile Crisis).
Suedfeld and Tetlock (1977) compared communications and
statements from two crises that ended in war (1914 and the 1950
outbreak of the Korean War) and three peacefully resolved crises (the
1911 Morocco crisis, the 1948 Berlin airlift crisis, and the 1962
Cuban Missile Crisis). As expected, they found higher levels of inte-
grative complexity when war was avoided. Suedfeld, Tetlock, and
Ramirez (1977) studied United Nations (UN) speeches on the Mid-
dle East over thirty years and found significant decreases in integra-
tive complexity during periods just before the outbreak of wars in
1948, 1956, 1967, and 1973. (On the other hand, integrative com-
plexity also dropped during 1976, when no war followed.) While the
post—World War II Berlin situation never led to a shooting war,
Raphael (1982) found a similar negative relationship between inte-
grative complexity in speeches and statements about the Berlin issue
and levels of tension over the status of Berlin.
One study, however, does not support the presumed relationship
between integrative complexity and war. Scoring both Japanese
intragovernmental documents and formal diplomatic communica-
tions from Japan to the United States, Levi and Tetlock (1980) found
no tendency for levels of integrative complexity to decrease during
the last months of 1941, just prior to the Pearl Harbor attack.
In a further study of communications from Pearl Harbor and eight
other "surprise attacks," however, Suedfeld and Bluck (1988) found
that the "attackers" showed a drop in integrative complexity
between three months and several weeks before the attack, while the
"to-be-attacked" nations showed increases during the month before
the attack (with a dramatic drop just after the attack). Suedfeld and
Bluck suggest that, as the attacking nation hardens its negotiating
position (low integrative complexity), the to-be-attacked nation tries
even harder to be flexible and to understand the other side (high