You are on page 1of 15

Scandinavian Journal of Educational Research, Vol. 45, No.

3, 2001

Critique of the Schema Concept


BO DAHLIN
Department of Educational Sciences, Karlstad University, S-651 88 Karlstad, Sweden

ABSTRACT A basic concept in most theories in cognitive psychology, as well as in many


versions of constructivism, is that of the schema. The questions behind this paper concern the
nature of schemata and what role the concept plays in educational theory and practice. It starts
with a brief look at the historical development of the concept, from Kant to its use in some
modern textbooks of psychology. One argument is that during its history the originally
hypothetical character of the concept has been forgotten. It is now more or less taken for granted
that schemata exist. But there are logical inconsistencies in applications of the concept and its
usefulness within educational theories can be questioned. The concept also lends itself to
manipulatory psychological practices, illustrated by the example of a US advertising company.
The paper ends with a suggestion that we return to Aristotle and a more phenomenological
approach to the questions of learning and knowledge formation.

Key words: cognitive schema; learning perception; phenomenology

INTRODUCTION
The schema concept is central to most theories in cognitive psychology, as well as
to many versions of constructivism. The basic question behind this paper concerns
the nature of schemata: what are they really and what role does the concept play in
educational theory and practice? I begin with a brief look at the historical develop-
ment of the schema concept, starting with Kant’s Kritik der reinen Vernunft. After
Kant some British psychologists took up the concept at the beginning of this
century. Here the notion acquired the status of an empirical psychological hypoth-
esis, considered useful for explaining certain perceptual phenomena. As such,
Piaget, for whom the schemata of intelligence were developments of innate sensori-
motor schemata, also took it up. Finally, the concept has been used in various ways
within present constructivist theories of learning.
One of my arguments is that during this historical process the hypothetical
character of the concept has been forgotten, or in any case neglected. It is now more
or less taken for granted, at least in educational circles, that schemata exist.
However, there are logical inconsistencies in applications of the concept and its
usefulness within educational theories can be questioned. Some of this critique has
been voiced earlier, for instance in Still & Costall (1991). However, its effect on the
discourse of cognitive psychology seems only marginal [cf. my discussion of Eysenck
ISSN 0031-3831 print; ISSN 1430-1170 online/01/030287-14 Ó 2001 Scandinavian Journal of Educational Research
DOI: 10.1080/00313830120074215
288 B. Dahlin

(1993) below]. Furthermore, the concept lends itself to manipulatory psychological


practices, as I will illustrate with the example of a US advertising company.
Finally, I suggest that we return to Aristotle and a more phenomenological
approach to the questions of learning and knowledge formation. Instead of talking
about students’ schemata and how they can be changed, we should focus on
students as persons involved in changing their understanding of things. The discourse
of schemata and inner mental representations tends to focus our attention on
invisible and reiŽ ed mental structures, whereas the true focus of education is the
living person and their struggle to understand the world.

SHIFTS AND SHADES OF MEANING IN THE SCHEMA CONCEPT


The Schema Concept in Kant’s Kritik der reinen Vernunft
My awareness of the history of the notion of schemata goes back to Kant. In his
Kritik der reinen Vernunft he introduces the term under the general heading ‘Tran-
scendental doctrine of judgement or Analytic of principles’ (A136/B175ff). As far as
I know, Kant scholars are still arguing about the real signiŽ cance of Kant’s schema
concept. However, it seems clear that the concept is an answer to the following
problem:
But pure concepts of the understanding, when compared with empirical
intuitions, or even with sensible intuitions in general, are quite heteroge-
neous, and never can be discovered in any intuition. How then is the
subsumption of the latter under the former, and consequently the appli-
cation of the categories to appearances, possible? (Kant, 1993, p. 142).
That is, pure concepts, on the one hand, and pure sense experience, on the other,
are so different in nature that it is difŽ cult to see how the former could ever be
‘applied’ to the latter. Some intermediary link between the two is needed:
… there must be some third thing, which on the one hand is homogeneous
with the category, and with the appearance on the other, and so makes the
application of the former to the latter possible. This mediating representa-
tion […] is the transcendental schemata. (Kant, 1993, p. 143).
Thus, Kant introduces schemata as the necessary links between the purely
conceptual realm of understanding and the purely sensational realm of appearances.
Why he qualiŽ es schemata as ‘transcendental’ I do not know. One could guess that
it is because he does not consider schemata as empirical, psychological entities or
functions, but as belonging to the general ‘transcendental conditions’ of knowledge.
For instance, a schema is not some sort of mental image:
The image is a product of the empirical faculty of the productive imagin-
ation—while the schema of sensible concepts (of Ž gures in space, for
example) is a product, and, as it were, a monogram of the pure imagination
a priori, whereby and according to which images Ž rst become possible… .
(Kant, 1993, pp. 144–145)
Critique of the Schema Concept 289

Here Kant seems to make a fundamental distinction between ‘the empirical faculty
of the productive imagination’ as a psychological function and ‘the pure imagination
a priori’ as a transcendental condition for speciŽ c mental images. However, there are
passages pointing in another direction:
This schematism of our understanding in regard to appearances and their
mere form, is an art, hidden in the depths of the human soul, whose true
modes of action we shall only with difŽ culty discover and unveil. (Kant,
1993, pp. 144)
If schemata are ‘modes of action’ in the depths of our souls, which, albeit with
difŽ culty, can be ‘discovered’, it seems that they must also be of the nature of
empirical psychological processes. A mode of action that can be discovered or
unveiled must also be in some sense observable, or at least perceptible, and hence
empirical. And is not the statement that only with difŽ culty shall we discover their
true modes of action almost a premonition of modern psychological research?
Especially that kind of research which Harré (1998) calls ‘realist psychology’. Realist
psychology is based on scientiŽ c realism, which is the view that the theories of
natural science are true models of unseen physical entities or processes (for instance
the atomic model). When this kind of philosophical realism enters psychology it is
assumed that there are unobservable psychological processes and that models of
such processes can be achieved, which would play the same role in psychology as
models of atoms and molecules play in physics and chemistry.
Finally, let us note that for Kant schemata were not identical with concepts and
categories of understanding. In this he differs markedly from present day cognitive
psychology, in which concepts and categories are virtually the same as schemata.

The Schema Concept in British Psychology at the Beginning of the 20th Century
OldŽ eld & Zangwill (1942) attempted to ‘trace the development and application of
the concept of schema’ (p. 267) in British psychology. According to them the schema
concept was introduced into psychology at the beginning of the 20th century, as a
‘retreat’ from the failure of 19th century theories to come up with fully explanatory
and deductive psychological systems. Originally the schema concept was taken as a
descriptive term, ‘to permit the accurate characterisation of phenomena within the
limits imposed by available methods of investigation’ (OldŽ eld & Zangwill, 1942,
p. 268; my emphasis). Thus it was not seen as having any deductive or explanatory
value, although it was hoped that out of the ‘somewhat  uid conceptual systems’ in
which it appeared there would ‘eventually crystallise the means of attempting once
again the construction of large bodies of explanation internally integrated by rela-
tions of a deductive character’ (OldŽ eld & Zangwill, 1942, p. 268). However, it
seems that this crystallisation occurred automatically, i.e. the descriptive nature of
the concept was forgotten and schemata were taken for granted and reiŽ ed simply
by being referred to more and more often.
Latour & Woolgar (1979) point out a tendency to reify scientiŽ c concepts in
natural science research. Scientists have to model their new, hypothetical concepts
290 B. Dahlin

of invisible physical entities and manipulate them in their minds in order to Ž nd


possible explanations for observational data. In time this constant mind manipu-
lation of concepts and models make them more and more ‘real’ to the scientists and
their originally tentative character is forgotten. This tendency to reify tentative
concepts surely does not occur only in research in natural science. It can also be
found in the social sciences, not least in psychology. Like nuclear physics and
molecular chemistry, psychology deals with essentially invisible phenomena, which
have to be modelled and explained with equally invisible concepts. The same mind
manipulation of tentative ideas goes on in psychological research as in the chemist’s
laboratory. In this process the schema concept seems to have undergone a gradual
reiŽ cation and the hypothetical and purely descriptive character it once had has been
forgotten. Schemata become ‘real things’, existing in the mind, intelligence or brain,
depending on the perspective within which the concept is used.

Bartlett’s Schema Concept


Bartlett’s (1932) book about memory is part of OldŽ eld & Zangwill’s (1942) review
of the schema concept. However, Bartlett’s use of the term deserves a separate
section, since he is often quoted or referred to when the schema concept is
introduced [1]. According to Caramelli (1987), and more recently Beals (1998),
most modern cognitive psychologists have fundamentally misunderstood Bartlett’s
schema concept. For Bartlett schemata were primarily functional and holistic in
nature, having to do with the interaction between the ‘organism’ and its environ-
ment, whereas cognitive psychologists more often see schemata as ‘mental struc-
tures’ or ‘representations’, i.e. as mental entities of some sort. The functional
character of Bartlett’s schemata shows in the following quote:

What precisely does the schema do? Together with the preceding incoming
impulse it renders a speciŽ c reaction possible. It is, therefore, producing an
orientation of the organism toward whatever it is directed to at the
moment. But that orientation must be dominated by the immediately
preceding reactions or experience. To break away from that the schema
must become not merely something that works in the organism, but
something with which the organism can work. The organism discovers how
to turn around on its own schemata. In other words, it becomes conscious.
(Bartlett, 1932, p. 207)

It seems that for Bartlett the schema was an expression of the organism as a
whole, by being the result of complex and dynamic processes, both within the
organism itself and between the organism and its world. It is important to note the
implication of the quote above: that schemata can change by the conscious work of
the organism. Also, Bartlett did not overlook the signiŽ cant role that culture plays
in these processes, which has inspired Beals (1998) to construct an interesting
alignment between Bartlett and Bakhtin.
Critique of the Schema Concept 291

Caramelli (1987) suggests that by returning to Bartlett’s original idea of schema


we can develop a functional, holistic and ‘ecological’ clariŽ cation of the concept.
The schema can then be seen as
… a higher order behavioural unit whose functioning aims at establishing
the attunement between the organism and its environment according to the
boundary constraints set up by the organism’s attitudes, interests, and
needs. … Its [the schema’s] task is the maintenance of the continuous link
between the organism and its environment established by its very life.
(Caramelli, 1987, pp. 14–15)
Still, I cannot help wondering if this view of schemata makes us any wiser. What can
we do with such a concept? Can we explain human learning in a better way than we
can without it? Or do we just make a new description of what we already know: that
organisms are in constant interaction with their environment; that these interactions
often follow patterns established by habits, norms or culture; that these patterns
nevertheless can be changed by the organism itself (as well as by external factors).
And why pay so much attention to schemata and so little to the ‘organism’ or, better,
the person who ‘discovers’, ‘turns around on’ and (above all) ‘works with’ their
schemata?

The Schema Concept in Present Cognitive Psychology


Eysenck’s (1993) book about the principles of cognitive psychology seems to sum up
some central aspects of more recent schema theories. When introducing the schema
concept he says:
it seems probable that we possess much larger chunks of organised knowl-
edge, and the term schema (plural schemata) to describe this was proposed
by Sir Frederic Bartlett in 1932. He used the term in a rather vague way,
but there has been subsequent agreement that a schema should be
regarded as a mental representation … . (p. 86)
This passage seems to illustrate the misunderstanding of Bartlett’s original
intention, described in the previous section. Schemata has come to be regarded as
‘mental representations’, rather than holistic expressions of processes of interaction
between man and world. In this passage the hypothetical and descriptive nature of
the schema concept is also pointed out. It is said that we probably have schemata
and that the term is used to describe ‘larger chunks of organised knowledge’ or
‘mental representations’ [2].
However, the schema concept very quickly turns into an explanatory tool when
it is stated that ‘it is one of the prime uses of scripts or schemata to allow inferences
to be drawn’ (Eysenck, 1993, p. 87; my emphasis). And further:
schemata can in uence the way in which we understand and make sense of
information that is presented to us. According to Sir Frederic Bartlett
(1932), memory is also affected by schemata. More speciŽ cally, he claimed
that memory is determined in a complex fashion by the presented infor-
292 B. Dahlin

mation and by relevant prior knowledge in the form of schemata. (Eysenck,


1993, p. 87; my emphases)
In this quote schemata are invested with the power of explaining how we make
sense of new information, as well as how the form and content of our memory is
constituted. However, what happens if we substitute schemata by the Ž rst deŽ nition
of the term? Then we learn that it is ‘our chunks of organised knowledge’ which
allow us to draw inferences, i.e. what we know in uences the implications we see in
new information. And the Ž rst sentence of the quote above would say that the same
chunks of knowledge in uence the way we understand and make sense of new
information. Could anything be more obvious? Is it not ‘commonsense’ that what
we understand from new information depends upon what knowledge we already
have? As for the determination of memory, would not a few moments of re ective
thinking come to the same conclusion: that what we remember is also partly
determined by our knowledge? So what have we actually gained by the introduction
of the schema concept?
In an interesting critique of the concepts of adaptation and Ž tness within
biological theories of evolution Brady (1979) considered the consequences of using
analytical deŽ nitions as explanations:
Analytical statements are quite useful for purposes of deŽ nition—‘a deaf-
ness is an impairment of the hearing’—but only a synthetic construction
can serve as a causal explanation. When the deŽ nition strategy is used with
causal intent, language breaks down. (p. 601)
It seems to me that we have something of this sort in the above examples from
Eysenck. Admittedly, the deŽ nition of schemata as ‘large chunks of organised
knowledge’ is not analytical in the strict sense. But what Brady says about analytical
statements applies also to nominalistic and purely descriptive deŽ nitions: if they are
used with a ‘causal intent’, i.e. as explanations referring to empirical causes, strange
things may happen. We may create the illusion of having explained something,
whereas we are really just moving around in circles.
However, schemata seem to be something more mysterious than a merely
nominalistic device, because, according to Eysenck and other cognitive psycholo-
gists, we need ‘evidence’ for their existence and such evidence is hard to Ž nd:
… it has been argued that Bartlett’s work is of little general interest
because he had to resort to stories from an unfamiliar culture in order to
obtain evidence for schemata. However, studies carried out by John Brans-
ford and his colleagues … indicate that Bartlett-type effects can be found
with very ordinary material. (Eysenck, 1993, p. 88; my emphasis)
The search for ‘evidence’ would seem rather unnecessary if the schema concept
was only another name for ‘organised knowledge’ or ‘structured knowledge’. Who
would deny that knowledge could be organised and structured, at least partly or in
‘chunks’? And who would deny that such knowledge exists, in the ordinary sense of
the word? Therefore, the search for ‘evidence’ must implicate something more
serious. The term schema turns into something more than a redescription of
Critique of the Schema Concept 293

something we already know. It is accepted as a real causal factor in mental processes.


In science such factors must be experimentally ‘proved’ to exist. Is this need for
‘proof’ the result of an attempt to move away from an everyday language discourse
(in terms of ‘chunks of organised knowledge’ and the like) to a scientiŽ c one of
reiŽ ed objects? But if the hypothetical reiŽ ed objects are from the very beginning
deŽ ned in everyday language and constituted by everyday experience, can such
attempts ever succeed? Would it not be better to stay within the realm of everyday
experience and discourse and explore it deeper?
However, Eysenck deserves some merit for pointing out that the schema theory
has difŽ culties with incorporating facts of accurate remembering, because of ‘its
emphasis on the ways in which schemata alter what is presented in systematic ways’
(Eysenck, 1993, p. 89). Many social scientists today, in the heyday of constructivism
and post-modernism, seem to deny that it is even possible to remember something
as it actually was.
Another important theoretical issue pointed out by Eysenck is ‘ … whether
schematic knowledge is always used at the time of comprehension and storage or
whether it is sometimes used at the time of retrieval’ (Eysenck, 1993, p. 116; my
emphases). It seems that this is a question that is often forgotten or neglected in
schema theories. Most forms of constructivism, whether information processing,
social or radical, seem to take for granted that schemata are always used in all types
of cognitive processes, including perception and comprehension.

LOGICAL PROBLEMS WITH THE SCHEMA CONCEPT


In most constructivist theories of learning the concepts of cognitive structure and
schemata are fundamental. As an example, let us look at Head & Sutton (1985).
One section of their text is headed ‘Cognitive Structures as Mosaics’. The authors
write:
Structures are Built Up from Discrete Parts. We have many different experi-
ences that we make sense of gradually, a bit here and a bit there as we meet
them. On some occasions, links can be made between these bits so that a
wider sense is made, incorporating previously separated understandings
(superordinate learning, in Ausubel’s terms). (p. 92)
Cognitive structures are built up from ‘discrete parts’. These parts are the ‘bits’
of meaning that we make out of our (sense) experiences. By making ‘links’ between
these bits, so that more extensive meanings are produced, a cognitive structure
arises. Thereby previously separated ‘understandings’ are united. The reasons why
Head and Sutton want to liken these structures to mosaics are:
1. each person idiosyncratically builds his own mosaic, and
2. the tiles employed by the individual are limited in range both by the
constraints of language and of personal experience. (p. 93)
In the mosaic metaphor the ‘bits of sense’ which the individual has produced
from previous experiences correspond to the tiles which the individual puts together
294 B. Dahlin

according to their own mind (although presumably also in communication with


others). However, all analogies have a  aw. Head and Sutton point out that the
mosaic metaphor is lacking in two respects: it does not consider ‘the  uidity of
thought’ and the ‘multi-dimensional connections’ between cognitive structures. But
there is another weakness of this metaphor, which they do not mention: where do
the tiles come from? In other words, given that we create structures out of ‘bits of
sense’ from earlier experiences, how are these bits of sense produced? In the quote
above, it is only said that we have many experiences ‘that we gradually make sense
of’, but how?
Schema theory appears not to say very much about this basic level of meaning
creation. The ‘bits of sense’ are portrayed as the result of a mental process, but how
does this mental process actually occur? A few pages later Head and Sutton write
that ‘Once the new concept has been successfully integrated … into the individual’s
existing, personal cognitive structure, it becomes part of that person’s repertoire of
tools used to make sense of the world’ (p. 95).
Thus, the cognitive structures which have been built out of ‘bits of sense’ are in
their turn used ‘to make sense’, to create meaning. This reasoning would be
circular if one did not assume that there are primitive, original cognitive structures
that we use in our very Ž rst sense-making acts. This is where the schema concept
comes in. For instance, Rumelhart (1980) assumes that schemata may be broken
down into sub-schemata, but somewhere this process of breaking down must
stop: ‘ … there must be a set of schemata that are elementary in the sense that
they do not consist of a further breakdown in terms of sub-schemata’
(p. 40). Rumelhart calls these elementary schemata ‘primitives’. However, the
supposition that such primitives exist is only a logical consequence of the schema
theory as a whole. It is an ad hoc hypothesis. Rumelhart gives no empirical evidence
that such primitives exist [3], he only says that ‘there must be’ such elementary
schemata [4].
Another aspect of the schema theory concerns the ‘Ž ltering’ and ‘sorting’
functions of schemata (cf. Howard, 1987, p. 34 ff). Regarding their Ž ltering func-
tion, Howard (1987) writes that ‘Schemata Ž lter out data. We can only absorb a
limited amount of information and need some way to extract what is most import-
ant for our purposes …’ (p. 38). In this quote a different but equally interesting
aspect of the schema is revealed: a schema deals with information. This information
comes through our senses: ‘The mass of data coming through our senses has to be
Ž ltered, analysed and interpreted, for which a person needs schemata’ (Howard,
1987, p. 37).
Schemata choose the information that is important for our purposes, analyse it
and interpret it. But what is information if not a form of meaning? Can one imagine
information that does not carry any inherent meaning? However, it is precisely our
schemata that are supposed to ‘make sense’ of, i.e. create meaning out of, our sense
impressions. That which is ‘conveyed’ through our senses and ‘treated’ by our
schemata is described as ‘information’, or ‘data’. Sometimes it is also called
‘stimuli’. Howard (1987) seems to treat these three terms as synonyms. However,
the concept of information can hardly be deŽ ned without (at least an implicit)
Critique of the Schema Concept 295

reference to meaning. (This applies even for purely mathematical deŽ nitions of the
concept of some information theories. Such deŽ nitions view information in terms of
order, but order is an elementary form of meaning.) As for ‘stimuli’, Howard writes:
‘There may be many different stimuli present and we may miss some key ones
suggesting which schema should be selected. We may misinterpret certain stimuli
and pick an inappropriate schema’ (1987, p. 36). This quote evokes the following
question: how can stimuli ‘suggest’ which schema to activate if they do not already
have an inherent sense?
In summary, whether what is coming ‘into’ our mind ‘through’ our senses is
called ‘information’, ‘data’ or ‘stimuli’, it must be regarded as already carrying
meaning or ‘making sense’ to us. However, this basic level of meaning creation is
neglected and left unexplained within schema theories of learning.
Thus, there is a deeper, more fundamental level of experience which tends to
be glossed over by schema theories. This level concerns the actual change of the
structures of schemata. Following Piaget, such a change is often described in terms
of adaptation: the phenomenon is approached by a set of schemata which constitute
an anticipated experience. If the actual experience does not conform to this antici-
pation, the schemata which constituted the anticipation are accordingly restruc-
tured. One has then learnt to approach the phenomenon with a new set of schemata,
i.e. with other anticipations.
Piaget and his followers explain this restructuring by referring to the self-organ-
ising character of the cognitive system and by drawing a parallel with the changes
described by chaos theory. However, in spite of the niceties and subtleties of the
concepts of chaos theory, when I read such explanations I always feel that at the
bottom of it all lies a purely spontanistic conception: changes happen because they
happen. Look for instance at the following passage from Fosnot (1996, p. 16; my
emphasis):

Just as the genome when disturbed generates new possibilities, cognitive


structures when disturbed generate new possibilities—possibilities of new
actions or explanations of surprising results. These possibilities are ex-
plored and correspondences and/or patterns are constructed because of the
human’s self-organizing tendency.

Do such accounts of cognitive change tell us anything that we did not already
know? Do they make us any wiser as teachers? It may be argued that one of the
things schema theories have contributed to educational practice is the idea that
teachers need to challenge (‘disturb’) their students’ schemata, i.e. their understand-
ing. But would not any thoughtful teacher come to the same conclusion? All one
needs is the insight that we tend to stick to the understanding we already have as
long as we experience no necessity to change it. As far as I remember this has
already been said by Plato, and he had no schema concept. Anyone doing some
serious observation of themselves and other people can come to the same con-
clusion.
296 B. Dahlin

SOCIAL AND POLITICAL ASPECTS: THE EXAMPLE OF BURSON-


MARSTELLER
In a culture which in many or most areas is dominated by an instrumental rationality
it is not surprising to Ž nd technological applications of all kinds of research-based
theories, including psychological ones. Frohmann (1992) presents an interesting
example of this tendency. It is an analysis of the discourse of what he calls ‘the
cognitive viewpoint’. This point of view is based on information processing con-
structivism and is applied within library and information science. The analysis
identiŽ es a number of discourse strategies ‘which constitute information as a
commodity, and persons as surveyable information consumers, within market econ-
omy conditions’ (p. 365, abstract). He even concludes that by constituting the
‘information scientist as an expert in image interpretation and delivery, the cognitive
viewpoint performs ideological labour for modern capitalist image markets’ (p. 365,
abstract). Thus, Frohmann’s analysis indicates the actual and potential ideological
uses and effects of information processing theories, and in all such theories the
schema concept is of central signiŽ cance.
The US advertising company Burson-Marsteller seems to serve well as a
concrete illustration of ‘experts in image delivery’ serving ‘capitalist image markets’.
Burson-Marsteller is a globally active public relations agency. One of its clients is the
European genetic engineering industry, who engaged the company in order to bring
about public acceptance of genetically modiŽ ed foods and support for genetic
engineering in Europe [5]. On Burson-Marsteller’s web site (http://www.bm.com)
we can read the following (quoted verbatim): ‘Perceptions are real. They colour
what we see … what we believe … how we behave. They can be managed … to
motivate behaviour … to create positive business results’. Isn’t this the basic con-
structivist assumption that we construct the world according to the schemata that
make up our cognitive system? Change the schemata and you change the
signiŽ cance of the corresponding phenomena. This seems to be the gospel of
Burson-Marsteller, and what could be better news for a business company (or
anyone else) having problems with their public image?
Via Greenpeace, a Burson-Marsteller internal strategy document has reached
the public (Woitsch, 1998). Naturally, among the strategies used for changing
people’s perceptions of a particular business or industry is changing the key words
of the language used to describe it. For instance, ‘genetic engineering’ and ‘genetic
manipulation’ have now been generally changed to ‘biotechnology’. Another strat-
egy is avoiding talk about sensitive issues and focusing on positive aspects, i.e.
creating a perceptual ‘Ž lter’ among people which is more in accord with the clients’
interests.

RETURN TO ARISTOTLE
In Analytica Posteriora (II:19) Aristotle (1993) puts forward a descriptive analysis of
knowledge formation in four stages. The Ž rst is (1) sense perception, out of which
arises (2) memory. The recurrence of many memories concerning a particular thing
Critique of the Schema Concept 297

or phenomenon gives rise to (3) a general representation or conception of that thing.


Finally, out of such general representations arises (4) pure concepts or principles.
These four stages, Aristotle points out, are not to be taken as distinct and external
to each other. Instead, they are all internally related to perception as their primary
source.
The advantage of abandoning schema theories and returning to Aristotle, or
similar phenomenological approaches, is that one simultaneously moves away from
scientiŽ c realism and the reifying tendencies inherent in that perspective. Instead, we
gain closeness to practice and everyday experience, i.e. to our socially and culturally
constituted life-world. Consider for instance the following experiential description of
knowledge formation:
What, precisely, is ‘thinking’? When, at the reception of sense-impressions,
memory-pictures emerge, this is not yet ‘thinking’. And when such pictures
form series, each member of which calls forth another, this too is not yet
‘thinking’. When, however, a certain picture turns up in many such series,
then—precisely through such return—it becomes an ordering element for
such series, in that it connects series which in themselves are unconnected.
Such an element becomes an instrument, a concept. (Albert Einstein,
quoted in Holton, 1996, p. 197)
Even though this description comes from someone who is commonly considered a
genius of creative thinking, I believe most people can recognise the experience
described, as well as the similarity between this description and Aristotle’s stages of
knowledge formation. (Of course, it cannot be ruled out that Einstein had read
Aristotle and was in uenced by his analysis. But that would only prove the relevance
of Aristotle’s thinking even for modern people struggling with epistemological
questions.)
Furthermore, Aristotelian and kindred approaches to learning and knowledge
are free from mechanistic ideas and concepts. They are open to the speciŽ cally
human and spiritual dimensions of learning and knowledge building. It is possible
to distinguish a Ž fth stage in Aristotle’s theory of knowledge formation, having to do
with the spiritual and creative aspect, what Aristotle called nous poietikos (cf.
Aristotle, 1988, p. 430a). This is the level where Aristotle and Plato overlap and
agree. For Aristotle the nous poietikos, or active intelligence, was the immortal part
of the soul [6]. It can be seen as the basis of Plato’s theory of learning. In other
terms, these Ž ve stages of the development of knowledge could perhaps be summed
up as follows:
1. sense perception;
2. memories in the form of mental images;
3. general conceptions of discursive thought and verbal accounts;
4. analytical understanding of concepts and general principles;
5. holistic or synthetic understanding of relations and structures.
Even though everything grows out of perception, the higher stages presuppose more
and more of a creative, spiritual activity of the knowing subject, i.e. the learning
298 B. Dahlin

human being, the person. And it is precisely the person that seems to be forgotten,
or pushed out to the periphery, when schemata enter psychological theorising [7]. I
believe we would be far better off as educators if we talked about how we can
interact with our students in a way that stimulates and encourages them to change
and develop their understanding of things, rather than speculating about how we can
affect permanent changes in their invisible schemata.
From his phenomenological studies of the variations of learning experience
Giorgi (1999), among other things, concludes that ‘ … the establishment or holding
of a cognitive schema as a simulated intuition or as a guide for an adequate intuition
is obstructing rather than facilitating. It is usually maintained to avoid the risk of
error and that slows down learning’ (p. 85). This conclusion is in accord with the
above critique of the schema concept. In teaching there is always the risk that
students do not learn what the teacher is trying to teach them, i.e. they make
mistakes. However, according to Giorgi, making mistakes is an unavoidable part of
a genuine learning process. Exclusive attention to the ‘mental structure’ that
supposedly is to be established in the students’ mind may actually be dysfunctional,
especially from a wider, holistic point of view. In other words, let’s talk more about
the persons we meet in education, their doings and sufferings in trying to learn, and
less about the schemata they may or may not carry around in their heads.

CONCLUSION
No doubt the schema concept(s) can successfully describe various educational
phenomena, such as learning and knowledge formation. It has given rise to much
research activity and has thereby produced a lot of interesting information. It has led
to a large consensus among educational researchers. But none of these things is
proof of its explanatory value, let alone its truth. Furthermore, such a proof seems
impossible to achieve. By all means let us go on with researching people’s way of
understanding, perceiving or experiencing things, but, as educationists, let us turn
more attention to the persons, the subjects, who are re ecting on and working with
these ways of understanding and experiencing.

NOTES
[1] With reference to Bartlett’s own introduction of the term, OldŽ eld & Zangwill (1942) said that it
‘has nothing in common with Kant’s use of the same word’. [On the other hand, they also noted
that what Kant called pure intuitions ‘have much the same function in articulating sense-impres-
sions and rendering perception possible’ (op. cit., p. 269, footnote 1).] But this seems to be an
overstatement, since there is at least one similarity between the functions Kant ascribed to schemata
and the functions ascribed to them by Bartlett and others, namely their ordering functions in
remembering, perceiving, etc. Whether one sees differences or similarities between terms and
notions used by various thinkers is often a matter of choice, depending on what aspects one wants
to emphasise.
[2] In the glossary to Eysenck’s book schemata are deŽ ned as ‘large chunks of organised knowledge
stored in long-term memory’ (1993, p. 179)
[3] At least not in the text quoted here.
Critique of the Schema Concept 299

[4] Bickhard (1995) delivers a critique of some basic notions of constructivism which is very relevant
to my argument. Among other things, he notes:
Usually, an atomic element version is postulated for representation […] in which the
presumed grounding encoding elements are taken as the atomic encodings (encodings
of basic features, or basic facts, perhaps)—the atomic encodings out of which all other
representations are constructed, and in terms of which all other representations are
deŽ ned. (p. 233)
[5] According to Woitsch (1998), other projects that Burson-Marsteller has been involved in are:
optimising the marketing of British meat during the BSE crisis; brushing up the public image of
former Indonesian President Suharto after the invasion of East Timor; improving the PR image of
Union Carbide after their 1984 accident in Bhopal, India, in which about 2000 people died.
[6] I know that not all scholars agree with this interpretation.
[7] Harré (1998) distinguishes between the ‘causal picture’ and the ‘agentive picture’ in cognitive
psychology. The causal picture arises out of mechanistic models, where the structures of schemata
are taken as causes of cognition and perceptions.

REFERENCES
ARISTOTLE (1988). De Anima [On the soul] (Trans. H Lawson-Tancred). London, UK: Penguin Books.
ARISTOTLE (1993). Analytica Posteriora. Berlin, Germany: Akademie Verlag.
BARTLETT , F. C. (1932). Remembering—a study in experimental and social psychology. London, UK:
Cambridge University Press.
BEALS, D.E. (1998). Reappropriating schema: conceptions of development from Bartlett to Bakhtin.
Mind, Culture, and Activity, 5, 3–24.
BICKHARD, M.H. (1995). World mirroring versus world making: there’s gotta be a better way. In L.P.
STEFFE & J. GALE (eds), Constructivism in Education, pp. 229–268. Hillsdale, NJ: Lawrence Erl-
baum.
BRADY, R.H. (1979). Natural selection and the criteria by which a theory is judged. Systematic Zoology,
28, 600–621.
CARAMELLI , N. (1987). The ‘Schema’ Concept. Bartlett till now, Report no. 21. Lund, Sweden: Depart-
ment of Psychology, Lund University.
EYSENCK, M.W. (1993). Principles of Cognitive Psychology. Hillsdale, NJ: Lawrence Erlbaum.
FOSNOT, C.T. (1996). Constructivism: a psychological theory of learning. In C.T. FOSNOT (ed.),
Constructivism. Theory, perspectives, and practice, pp. 8–33. New York, NY: Teachers College Press.
FROHMANN, B. (1992). The power of images: a discourse analysis of the cognitive viewpoint. Journal of
Documentation, 48, 365–386.
GIORGI, A. (1999). A phenomenological perspective on some phenomenographic results on learning.
Journal of Phenomenological Psychology, 30, 68–93.
H ARRÉ , R. (1998). The rediscovery of the human mind. In U. KIM (ed.), Proceedings of the 50th
Anniversary Conference of the Korean Psychological Association. Seoul, South Korea: Department of
Psychology, Chungang University.
H EAD , J.O., & SUTTON , C.R. (1985). Language, understanding, and commitment. In L.H.T. WEST &
A.L. PINES (eds), Cognitive Structure and Conceptual Change, pp. 91–100. Orlando, FL: Academic
Press.
H OLTON, G. (1996). Einstein, History, and Other Passions. The rebellion against science at the end of the
twentieth century. Reading, MA: Addison-Wesley.
H OWARD, R.W. (1987). Concepts and Schemata. An introduction. London, UK: Cassell.
KANT, I. (1993). Critique of Pure Reason. A revised and expanded translation based on Meiklejohn. London,
UK: Everyman’s Library.
LATOUR , B., & WOOLGAR , S. (1979). Laboratory Life: the social construction of scientiŽ c facts. Beverly Hills,
CA: Sage.
300 B. Dahlin

OLDFIELD , R.C., & ZANGWILL, O.L. (1942) Head’s concept of the schema and its application in
contemporary British psychology. Part 1. Head’s concept of the schema. British Journal of Psy-
chology, XXXII, 267–286.
RUMELHART , D.E. (1980). Schemata: the building blocks of cognition. In R.J. SPIRO , B.C. BRUCE &
W.F. BREWER (eds), Theoretical Issues in Reading Comprehension. Perspectives from cognitive psychology,
linguistics, artiŽ cial intelligence, and education, pp. 33–58. Hillsdale, NJ: Lawrence Erlbaum.
STILL , A., & COSTALL, A. (eds) (1991). Against Cognitivism. Alternative foundations for cognitive psy-
chology. New York, NY: Harvester Wheatsheaf.
WOITSCH, I. (1998). Manipulating consciousness with advertising strategies, e.g. ‘biotech’ instead of
‘genetic engineering’. Das Goetheanum Wochenschrift für Anthroposophie, 30, 441–443.