Organization of Memory
Organization of MemoryA considerable amount of research has been
carried out on the nature and structure of
memory, its functions and processes but
recently, psychologists such as Schank
(1975) have started to conduct empirical
studies on the organisation of memory. This
essay will examine one of the more recent
developments in the organisation of memory,
the schema theory.
According to Gross (1999) the organisation of
memory involves two main concepts. These
are the format in which information is
arranged and how this information is
arranged. The earlier theories, such as the
matrix and hierarchical network models by
Collins and Quillians (1969, 1972) concerning
the organisation of memory adopted a logical,
sequential approach to the organisation of
memory. Later on, in the late 1970’s
psychologists such as, Schank, Abelson and
Rumelhart realised that the organisation of
memory is not as logical as the earlier
models proposed. The schema theories
make an attempt at addressing the more
complex aspects of the organisation of
memory.
Semantic memory plays a fundamental role in
the earlier theories but according to Baddley
(1990) it became increasingly obvious during
the 1970’s that the SM must contain structures
considerably larger than the simple concepts
discussed in the network models by Collins
and Loftus. Bartlett (1932) used the term
‘schema’ to refer to this larger unit of semantic
memory. According to Hayes (1999) a
schema is a ‘….form of representation which
we use to guide our actions’ Bartlett (1932)
described a schema as ‘ an active
organisation of past reactions, or of past
experiences which must always be supposed
to be operating in any well-adapted
organismic response’. Bartlett also realised
that our memory is often disorganised, vague
and incomplete and that any theoretical model
must attempt to explain such inaccuracies in
our representation of information or
knowledge. Bartlett also recognised that
memory is a reconstructive process in which
information already stored affects the
remembering of other events. According to
Cohen (1993), schema theory represents the
most influential approach to the fundamental
problem identified by Bartlett. At its core is the
belief that what we remember is influenced by
what we already know and that our use of past
experience to deal with new experience is a
fundamental feature of how the mind works.
Two major modern schema theories are
those of Rumelhart (1975) and Schank (1975)
and Schank and Abelson (1977). Their
theories greatly overlap and are, in many ways
inter-related. The main concepts and ideas in
these two theories were summarised by
Norman (183/85). First, that schema’s have
variables or slots. Schemas are packets of
information, which comprise a fixed,
compulsory value plus a variable or optional
value. Rumelhart and Norman (1983)
described schemas as ‘varying considerably
in the amount of information they
contain-some are quite simple while others
are extremely complex’. Here, there is a
subtle reference to the variable slots. Second,
schemas can be related together to form
systems. That is, they are not mutually
exclusive packets of information but can be
overlapping. For example, a schema for a
picnic may be part of a larger system of
schemas including, meals, outings and
parties. Third, Schemas represent knowledge
of all levels of abstraction. Schemas can
relate to abstract ideologies, abstract
concepts or concrete objects. For example,
they may be used to describe justice or the
appearance of a face. Fourth, Schemas
represent knowledge rather than definition.
Schemas incorporate knowledge and
experience of the world rather than abstract
rules. Fifth, schemas are active recognition
devices. This is very similar to Bartlett’s ‘ effort
after meaning’ whereby we try to make sense
of ambiguous and unfamiliar information in
terms of our existing knowledge and
understanding.
According to Schank (1975) and Schank and
Abelson (1977) we develop schemas and
scripts which represent commonly
experienced social events such as catching a
bus or going to a restaurant. Baddley (1990)
says, ” scripts can only be interpreted by
bringing in a great deal of additional
information” According to Schank et al, scripts
are essential ways of summarising common
cultural assumptions. Not only do they help us
to understand text and discourse but also
predict future events and behave appropriately
in given social situations. Furthermore,
scripts contain the sequences of actions one
goes through when carrying out stereotypical
events and would also include the sorts of
objects and actors we are likely to encounter.
Schank and Abelson (1977) built their scripts
into a computer program (SAM) which, they
claim is capable answers about restaurants
and accounts of restaurants. Their scripts
were supported by Bower et al (1979) where
he asked people to list 20 actions/events in
sequential order which commonly occur while
eating at a restaurant and found considerable
agreement. He found that 73% mentioned
sitting down, ordering, eating, paying the bill
and leaving. Also, at least 48% included
entering, giving the reservation name,
ordering drinks, eating dessert and leaving a
tip. Thus there were at least 15 events similar
to the majority of people. These findings
broadly agree with Schank and Abelson’s
restaurant script. Interestingly, when such
events were embodied in a story people
tended to falsely recall the facts, which were
not explicitly recorded but were consistent with
the script.
The schema theories have strengths and
weaknesses. They were able to provide a
fundamental basis and further development in
theorising the organisation of memory. Not
only did they address the more complex
situations of social and cultural aspects of
everyday situations but were also to provide a
basis for the explanations of real life
situations. However, one of the main
problems of the schema theory is that it is
often very difficult to define what a schema is.
Rumelhart (1980) identified four different
analogies of what a schema is but was
unable to actually define a schema. The
schema theory is also very limited as Alber
and Hasher (1987) argue that memory
representation is much richer than the
schema theory suggests. The schema theory
adopted a rather limited model of the schema,
which said that schema first encoded only
relevant information and then extracted only
relevant meaning, so that none of the original
form of the information would be needed.
When Alber and Hasher identified aspects of
memory, which were richer than that, and also
included, irrelevant details they concluded that
schema were not adequate to explain how
organisation and memory representation
works.
Cohen (1993) points out that ‘ the whole idea
of a schema is too vague to be useful’. Also
that there is an overemphasis on the
inaccuracies of memory and overlooks the fact
that complex events are sometimes
remembered in great detail. Cohen further
argues that the schema theory does not
suggest the origin of schema or how they are
created in the beginning. Nonetheless, he
admits that without schemas we are unable to
interpret new experiences and also these new
experiences build up schemas. It seems that
schemas definitely form part of the
organisation of memory although the schema
theory itself had many limitations and needs
to be developed further.
In conclusion, the schema theory has
provided us with a fundamental basis on
which theorists may develop further and use it
as a solid grounding and basis for their
theories. However, the theory is too limited
and vague and does not address all aspects
of our complex memory system. Yet, it is
evident that schema play an important and
essential role in our social and cultural
environments and do explain some aspects of
the way our memory is organised.