was some weak support for the existence of greater positive affect
reciprocity (a marginal effect, and only for reciprocity from wife to
husband, W → H).

During the events-of-the-day task, conflict engagers differed from
avoiders in a greater likelihood of negative start up (only by husbands)
for conflict engagers compared with avoiders. This means that engager
husbands were more likely to engage in sequences that brought them
from neutral to negative affect.

On the positive conversation, the two groups did not differ on the
sequential variables.

Thus, on two of the tasks, the sequential structure of the interaction
demonstrated that conflict engagers were more likely to start up and
continue in negative affect cycles, and some weak evidence that enga-
gers were more likely to reciprocate positive affect.

Conflict engagers expressed more emotion, both positive and negative,
and not just during the conflict discussion. They expressed more
emotion on their faces. They had a more lively exchange and took more
turns at speech. Conflict engagers started negative affect exchanges
more than avoiders, and they continued them more than avoiders, even
during the events-of-the-day discussion.

Perhaps the greater physiological linkage of avoiders is a clue to
understanding their avoidance of emotional expression (both positive
and negative) and their reluctance to conflict. In a previous study
( Levenson & Gottman, 1983), Levenson and I found that physiological
linkage was related strongly to marital unhappiness. We also found that
linkage was positively related to the amount of negative affect ex-
pressed. We speculated that linkage may be the physiological basis of
negative affect becoming (at times) an absorbing state, which is typical
of unhappy marriages. Perhaps avoiders are avoiding this negative
affective absorbing state, which would be their lot because of their

-156-