environment and from the connections serving the various cognitive systems.
Recollection typically involves making plausible guesses about what probably
happened. Recollection is an active process, akin to fantasizing or speculating
about the future, whereby people recreate or infer their past rather than reex-
perience it. Another way to put it is that people learn reconstruction strategies
that enable them to deduce past events. Loftus (1982) provides a discussion of
some of the various types of reconstruction strategies.
To illustrate, suppose a person returns to the scene of a car accident and tries
to recall the details of the accident, which occurred several days earlier. Re-
turning to the intersection is likely to activate the same elements of the cogni-
tive system involved in originally perceiving the accident; consequently some
perceptual details necessary to reconstruct the accident will become available
(e.g., cars move quickly through the intersection). Thoughts about a car acci-
dent may also activate knowledge of how cars work (e.g., brakes often squeak
when a driver tries to stop a rapidly moving vehicle). Such knowledge may
then become a basis for reconstructing the accident. Information that was pro-
vided to the person after the accident occurred may also be activated and
inserted into the reconstruction of the accident (e.g., a friend at the scene of the
accident later claimed to have seen a van cut in front of the car). The confluence
of activated elements constitutes the memory of the accident (e.g., a van cut in
front of a fast-moving car, which tried to stop, causing its brakes to squeal). The
memory may appear to the person to be vivid and accurate, yet some details
maybeinerror(e.g.,perhapsthevannevercutinfrontofthecar).
Reconstructing the Past
An important implication of reconstruction is that when people try to recollect
a past event, what they will remember about that event will depend on what
they currently know or believe to be true about their lives. Errors in recollecting
events will not be haphazard, but will instead reflect knowledge and beliefs.
So researchers interested in demonstrating reconstruction often vary a person’s
currentknowledgeandshowthattheperson’srecollectionofsomepastevent
will be distorted as a consequence (Dooling & Christiaansen, 1977; Hanawalt &
Demarest, 1939; Snyder & Uranowitz, 1978; Spiro, 1977).
A nice demonstration of reconstruction is provided by Spiro (1977). In his
experiment, subjects read a passage about a couple. Bob and Margie, who were
engaged to be married. Bob was reluctant to tell Margie that he did not want to
have children, but, by the end of the story, finally confronted Margie with his
wishes. In one version of the story, Margie told Bob that she wanted children
very badly. Afterwards, the subjects were told either that Bob and Margie are
now happily married or that the engagement had been broken off. Days to
weeks later, subjects returned and tried to recall the details of the story. Sub-
jects who were told that the engagement had been broken off tended to recall
accurately that Bob and Margie disagreed sharply about having children. In
some cases they even exaggerated the disagreement. But the subjects told that
Bob and Margie were now happily married tended to recall that the disagree-
ment was much less severe than was actually depicted in the story. And the
longer the time between reading the story and recalling it, the more likely
these subjects distorted the story so as to resolve the inconsistency between the
Memory 333