self has varied and broad definitions. While there is not universal agreement
among affect researchers regarding the way in which affect is defined, it is
necessary to provide a working definition for any analysis regarding the links
between affect and cognitive processing. Accordingly, in this chapter we fol-
low Rosenberg’s (1998) definition of affect in which affect is defined in terms
of affective traits and states. This definition is somewhat narrow in that it
does not include general preferences or sentiments and thus leaves out moti-
vational aspects of affect such as interest.
Affective traits refer to stable ways or predispositions to emotional re-
sponding (Rosenberg, 1998). However, in this chapter we focus on affective
states, namely on moods and emotions, as this is more closely linked to re-
search from social psychology on affect and cognitive processing (e.g., Bless,
2000; Forgas, 2000b) and recent work on academic emotions and cognitive
processing (Pekrun et al., 2002). Moods and emotions are distinct in terms of
intensity and duration (Rosenberg, 1998; Schwarz, 1990; Schwarz & Clore,
1996). Moods tend to be longer lasting than emotions, which are character-
ized by short, intense episodes. However, while emotions tend to be intense or
rather short-lived, they may also fade into general mood states over time. In
addition to intensity and duration, Schwarz and Clore (1996) also note that
mood states do not have a particular referent; the source of the mood is un-
clear. In contrast, emotions tend to be a reaction or response to a particular
event or person. This distinction between moods and emotions is not used by
all (for alternative perspectives, see Batson, Shaw, & Oleson, 1992; Morris,
1992); however, for the purpose of this discussion, we define moods as longer
lasting general affective states without a particular referent and emotions as
short, intense affective episodes with a specific referent.
Typically, research on affect and cognitive processing fails to consider the
arousal or activation dimension of affect and instead focuses almost exclu-
sively on the valence dimension, positive versus negative (Revelle & Loftus,
1990). This is due, at least in part, to the focus of social psychological re-
search on the relations between moods and cognitive processing. Since
moods are typically less intense, they usually do not differ in terms of arousal
whereas emotions often tend to vary in terms of arousal. However, it is im-
portant to note that a number of researchers and theorists (e.g., Cacioppo,
Gardner, & Berntson, 1999; Russell & Feldman Barrett, 1999; Tellegen, Wat-
son, & Clark, 1999; Thayer, 1986; Watson, Wiese, Vaidya, & Tellegen, 1999),
who focus on the nature or structure of affect, have developed models distin-
guishing between valence and arousal–activation.
Although the proposed models differ somewhat, they share the same basic
dimensions including high–low engagement, pleasantness–unpleasantness,
high–low positive affect, and high–low negative affect. The latter dimensions
reflect a mixture of arousal (or engagement) with valence (pleasant–unpleas-
ant). In this way, engagement and pleasantness may be distinct but they may
58 LINNENBRINK AND PINTRICH