vin waited in the lift line). The greater the number of memorized facts about a
character, the longer to verify whether any given fact about the character was
on the list. The opposite of the usual fan effect was observed if the test required
that subjects decide whether a fact was implied by other facts on the list (e.g.,
Marvin adjusted his skis). The greater the number of facts about a character, the
faster subjects could verify whether a fact was similar to one of the memorized
facts. The constructionist explanation is that the similarity judgment allowed
subjects to compare a fact to the pattern or theme extracted from the memo-
rized facts. The more facts there were to memorize, the more likely that such
patterns would be extracted.
Another demonstration that increasing the amount of information can im-
prove memory for patterns but undermine memory for details comes from
an experiment by Bower (1974). Bower required his subjects to learn a critical
passage describing the biography of a hypothetical character. The basic form of
the biography described the time and place of birth, the occupation of the
character’s father, the way the father died, and so on. Subjects then studied
some additional passages. For some subjects, the additional passages were
biographies similar in form but different in detail from the critical passage. For
other subjects, the additional passages were unrelated to the critical passage in
form and in detail. Later, all subjects had to recall the same critical passage.
Bower found that subjects who studied the related passages recalled fewer
details of the critical passage (e.g., the father was a servant) but more of the
general pattern of the passage (e.g., the passage described the father’s occupa-
tion) than did subjects who studied unrelated passages.
Students often feel overwhelmed by the amount of material they must learn
for an exam. Perhaps it would hearten them to learn that interference for newly
acquired information is not inevitable. If students can relate each piece of in-
formation to a common theme, then interference is not likely to occur. The stu-
dent should be able to remember large sets of information as well as small sets.
Similarly, if the examination tests for general principles rather than specific
details, then again interference is not inevitable. The more information one
must learn, the more likely the general principles can be extracted from the
information.
Summary and Conclusions
In the first section of this chapter, I introduced two competing types of theories
of memory. One is the record-keeping theory, which argues that memory is a
system for storing records of past events, that recollection is searching through
and reading the records, and that forgetting is caused primarily by the dis-
tracting presence of many memory records. The second is the constructionist
theory, which argues that memory reflects changes to the cognitive systems
used to interpret events, that recollection is reconstructing the past, and that
forgetting is caused primarily by the continuous changes each new experience
makes to the cognitive systems that interpret and act on stimuli. Few contem-
porary theories of memory embody all the features of record-keeping theories,
although some contemporary theories, especially those that use computers as
metaphors for memory, seem closer in spirit to the record-keeping than to the
Memory 349