relying on the reinstatement of familiar (previously learned) conditions. Since
such reinstatement involves highly automated networks, a total pattern can
be activatedpars pro totoeven from fragments of the input to which the net-
work was trained (Tucker, 1992; see also Clore & Ortony, 2000; Epstein &
Pacini, 1999). In contrast, affect differentiation involves explicit appraisal as
individuals differentiate schemas into their components that are processed in
an individuating fashion—weighing and comparing different goals and out-
comes in terms of whether they promote or thwart one’s goals and desires;
thus, attention is consciously directed to a systematic treatment of different
dimensions of emotion-relevant information. As different goals are weighted,
individuals’ emotions themselves can change (Clore & Ortony, 2000). Hence
this mode typically involves the unfolding of compact emotional information
into blends of differentiated emotions, including different valences (e.g.,
Harter, 1999; Labouvie-Vief, Hakim-Larson, DeVoe, & Schoeberlein, 1989;
Lane & Schwartz, 1987).
Another aspect that differentiates the modes is the way in which they cre-
ate meaning (Bruner, 1986; Piaget, 1955; Werner, 1955). Affect optimization
is inherently tied to personal experience or inner states of the individual. It
appears to be especially closely related to the activation of emotions in set-
tings that involve subjective and intersubjective experience that is relatively
implicit—such as affective communication through facial gestures or pro-
sodic features (Gianotti, 1999; Tucker, 1992; Vingerhoets, Berckmoes, &
Stroobant, 2003). Hence meanings are inherently ineffable—experiential and
nondeclarative, personal, and connotative. Affect differentiation, in contrast,
is based on denotative, precise meanings. These meanings, moreover, are rel-
atively formal and decontextualized through semantic structures and propo-
sitions. They refer less to intimate personal or interpersonal experience but
rather represent experience in terms of relatively impersonal and conven-
tional structures of meaning.
Related to the relatively personal–interpersonal versus relatively formal
nature of meaning is a further difference of how the two modes are affected
by learning and experience. Although optimization often applies to amyg-
dala-based biological triggers (Metcalfe & Jacobs, 1998; Metcalfe & Mischel,
1999), these triggers can be conditioned to new ones through implicit learning
processes that bypass consciousness. LeDoux (1996; LeDoux & Phelps, 2000)
suggested that this implies a low road route, exemplified in the conditioning
of fear reactions to tone. In this case, direct pathways from sensory cortex to
amygdala provide subcortical circuits for learning, providing a quick and
dirty processing road of high survival value in certain emergency situation.
Such implicit learning also permits priming of the amygdala to evaluate sub-
sequent information received along cortical pathways.
This low road contrasts with the high road cortical pathway in which
prefrontal cortical mechanisms participate in behavior change through proc-
246 LABOUVIE-VIEF AND GONZÁLEZ