iScale: Studying Long-Term Experiences through Memory - Modeling Users' Experiences with Interactive Systems - page 87

5.2 Reconstructing Experiences from Memory

Memory was for long understood as a faithful account of past events, which can be
reproduced, when trying to remember details of the past. This idea was first chal-
lenged by Bartlett (1932). He described remembering as an act of reconstruction,
which never produces the exact past event, but instead alters representation of the
event with every attempt to recall. Bartlett (1932) asked participants to recall an un-
familiar story told 20 hours earlier. The recalled stories differed from the original
in detail, order and importance of single events. In addition, participants augmented
their memories by applying rationalizations and interpretations to the original story.
Each further reconstruction distorted the stories even further.
At the heart of the notion of reconstruction lies the distinction between episodic
and semantic memory (Tulving, 2002). While episodic memory “is specific to a
particular event from the past, semantic memory is not tied to any particular event
but rather consists of certain generalizations (i.e. beliefs) that are rarely updated”
(Robinson and Clore, 2002, p. 935). These two types of memory serve different
needs, such as learning new information quickly - a capacity of episodic memory -
or developing relatively stable expectations about the world - a capacity of semantic
memory (Robinson and Clore, 2002). Reconstruction happens through the retrieval
of cues stored in episodic memory. In the absence of such cues, beliefs stored in
semantic memory may be used to reconstruct the past. This results in distortions,
where details that actually happened are replaced by generalizations based on what
we know about the world. Thus, the accuracy of remembered events depends on the
degree to which contextual cues are present and active in episodic memory.
Experiences do not only consist of contextual details, but also of value-charged
elements, such as emotions or overall evaluative judgments. One can distinguish
two approaches to the reconstruction of value-charged experiences: the construc-
tive and the value-account approach. The constructive approach assumes that felt
emotion cannot be stored in memory but is instead reconstructed from recalled con-
textual cues. In contrast, the value-account approach proposes the existence of a
memory structure, which stores the frequency and intensity of a person's responses
to a stimulus. This in turn can be used to cue the recall of contextual details of one's
experiences. In the following, we describe the two approaches in more detail.

5.2.1 The Constructive Approach

The constructive approach assumes that reconstruction happens in a forward tempo-
ral order (Anderson and Conway, 1993; Barsalou, 1988; Means et al., 1989). Barsa-
lou (1988) asked people to recall their experiences during the last summer. Most
participants started in the beginning and proceeded in a chronological order. Often,
recalling one episode cues the reconstruction of further episodes and more contex-
tual information surrounding the episode (Anderson and Conway, 1993) - just like
a string of pearls.
Robinson and Clore (2002) further argued that “emotional experience can nei-
ther be stored nor retrieved” (p. 935), but can only be reconstructed on the basis