Journal of Curriculum Studies

ISSN: 0022-0272 (Print) 1366-5839 (Online) Journal homepage: https://www.tandfonline.com/loi/tcus20

The role of beliefs in the practice of teaching

Jan Nespor

To cite this article: Jan Nespor (1987) The�role�of�beliefs�in�the�practice�of�teaching , Journal of

Curriculum Studies, 19:4, 317-328, DOI: 10.1080/0022027870190403

To link to this article: https://doi.org/10.1080/0022027870190403

Published online: 09 Jul 2006.

Submit your article to this journal

Article views: 2959

Citing articles: 680 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=tcus20
J. CURRICULUM STUDIES, 1987, VOL. 19, NO. 4 , 317-328

The role of beliefs in the practice of teaching

Jan Nespor
University of Texas at Austin

It has become an accepted idea that teachers' ways of thinking and understanding are
vital components of their practice. This has spurred a considerable amount of
research on teacher decision-making and information processing.1 However, in spite
of arguments that people's 'beliefs' are important influences on the ways they
conceptualize tasks and learn from experience,2 relatively little attention has been
accorded to the structures and functions of teachers' beliefs about their roles, their
students, the subject matter areas they teach, and the schools they work in.
Moreover, the research that does exist on these issues has relied either on very broad
and inclusive concepts (e.g., the 'traditional/progressive' beliefs dichotomy)3 or on
descriptive frameworks closely bound to specific cases and of limited use for
generalization or comparison.4
We need a theoretically-grounded model of 'belief systems' that can serve as a
framework for systematic and comparative investigations. The present paper is a
preliminary (and admittedly incomplete) attempt to provide such a model. Rather
than enumerating or describing beliefs and attempting to correlate them to specific
behaviour patterns, the paper develops a conceptualization of beliefs grounded in
current research in cognitive psychology and cognitive science, and suggests several
key functions of beliefs in teachers' thinking. While primarily theoretical in nature,
the framework presented here is grounded in a body of field-based research on
teacher thinking, the Teacher Beliefs Study (TBS).5 AS this study will be referred to in
the course of the paper, its general outlines are described below.
The TBS followed eight teachers across the course of a semester, videotaping
their classrooms and using the videotapes to construct verbatim records of classroom
action. Two types of interviews were used to generate data on the teachers' beliefs.
First, four long, semi-structured, and wide ranging interviews (called 'repertory
grid' interviews)6 focused on the teachers' general principles and beliefs about
teaching, about their students, about student behaviour, and about the community
and organizational contexts in which they worked. Second, four subsequent
interviews focused on the teachers' explanations of their teaching practices. In these
'stimulated recall' interviews,7 teachers were shown videotapes of their classrooms
and asked to describe and comment on the on-going activity.8
Because these interviews were time-consuming both for teachers (who spent an
average of 20 hours talking to us) and researchers (who had to analyse thousands of
pages of interview data), there were only eight teachers in the sample (two each in
eighth-grade mathematics, eighth-grade English, eighth-grade American history,
and seventh-grade Texas history). All of the teachers had at least two years of
experience in the classroom (teachers in three schools were observed).9 Pseudonyms
are used throughout to refer to schools, communities, and teachers.
318 JOURNAL OF CURRICULUM STUDIES 1 9 : 4

The paper begins by delineating six structural features of beliefs that serve to
distinguish them from other forms of knowledge. The functions and uses of beliefs
are then examined.

The structure of beliefs

At least four features —'existential presumption,' 'alternativity', 'affective and

evaluative loading', and 'episodic structure'-can serve to distinguish 'beliefs' from
'knowledge'.10 Two other features —'non-consensuality' and 'unboundedness'-are
useful for characterizing the ways beliefs are organized as systems.

Existential presumption

Belief systems frequently contain propositions or assumptions about the existence or

nonexistence of entities.11 Beliefs in God, ESP, or assassination conspiracies are vivid
examples, but existential presumption also occurs in less obvious ways at much more
mundane levels of thought.
Both of the mathematics teachers involved in the research, for example, held
strong beliefs about student 'ability', 'maturity', and 'laziness'. These were not
simply descriptive terms, they were labels for entities thought to be embodied by the
students. One of the teachers, Mr Ralston, believed that learning mathematics was
primarily a function of practice and drilling, and that students who failed to learn did
so because they were too 'lazy' to do the work. He thus emphasized individual
seatwork and spoke of 'forcing' students to learn by making them do more work and
of motivating students to work by showing them the practical uses of mathematics.
By contrast, Ms Hunt, the second mathematics teacher, thought that learning
mathematics was primarily a function of 'maturity*. She allowed students to work
together in class, on the assumption that the differences in maturity between
students would be small enough to allow effective communication where her lectures
had failed, and explicitly rejected the notion of 'forcing' students to learn on the
grounds that one could not 'force' mental maturation.
The reification of transitory, ambiguous, conditional or abstract characteristics
into stable, well-defined, absolute and concrete entities is important because such
entities tend to be seen as immutable-as beyond the teacher's control and influence.

Alternativity

Abelson makes the point that beliefs often include representations of 'alternative
worlds' or 'alternative realities'.12 Extreme examples of this feature come easily to
mind (e.g., the social or cosmic orders envisioned by Utopian political or religious
movements), but it is also exemplified in many of our common beliefs about
everyday life. Several of the teachers in the study, for example, envisioned and tried
to establish instructional formats or systems of classroom relations of which they had
no direct experience or knowledge (nor were these based on abstract models they had
learned in their formal training).
To give one instance, an English teacher in the study, Ms Skylark, drew her
ideal of teaching from a model of what she had wanted classes to be like when she was
a child-friendly and fun. Although she worked to shape her class to that ideal, she
BELIEFS IN TEACHING 319

had never achieved it; nor had she experienced it as a child. Rather, this was a sort of
Utopian alternative to the sorts of classrooms she was familiar with: she would
sometimes explain her actions in the stimulated recall interviews by recalling
mortifying experiences as a student and expressing her desire to spare her students
such traumas.
Beliefs of this sort can be of great importance in the classroom. They are, in a
strict sense, overriding concerns, and any number of shortcomings and problems can
be justified in terms of their pursuit. Thus it was more important for Ms Skylark to
maintain a friendly and relaxed classroom environment-which for her meant
repeating and re-explaining assignments at length, and allowing students to initiate
digressive lines of action-than to finish an assignment: her classes almost always
ended with the lesson half-covered .Such beliefs are not amenable to falsification - or
even challenge — and failures to translate them into reality in no way diminish their
value.
In essence, then, 'alternativity' refers to conceptualizations of ideal situations
differing significantly from present realities. In this respect, beliefs serve as means of
defining goals and tasks, whereas knowledge systems come into play where goals and
the paths to their attainment are well-defined.13

Affective and evaluative aspects

Belief systems can be said to rely much more heavily on affective and evaluative
components than knowledge systems.14 In some respects, feelings, moods, and
subjective evaluations based on personal preferences seem to operate more or less
independently of other forms of cognition typically associated with knowledge
systems,15 though there is clearly a great deal of interaction between the systems.
Thus, knowledge of a domain can be conceptually distinguished from feelings about
that domain. One's knowledge of the rules of chess and various lines of play does not
depend upon whether one likes or dislikes chess, whether it excites or bores one,
whether one thinks it trivial and decadent or sublime and mystical (though these
attitudes and beliefs would be important influences on how or whether one acquired
such knowledge in the first place, and on how one might be inclined to use it).
Some of the influences of affect and evaluation on teaching are well documented:
much of the literature on teacher expectations, for example, concerns the conse-
quences of teachers' sometimes unrecognized feelings about students. A less obvious
arena in which affect is important is that of teachers' conceptions of subject matter.
The values placed on course content by the teachers in the TBS study often influenced
how they taught the content.
For example, three of the four history teachers felt that teaching the 'facts' and
details of history should not be a primary goal of their courses because, in their
evaluation, students could not be expected to remember such information for any
significant length of time: the content was 'short-term memory stuff as one teacher
put it. They also knew that the content would either never reappear in later grades (in
the case of Texas history) or would be thoroughly re-taught in high school history
classes (in the case of American history). Rather than focusing their energies on
imparting historical facts and details, then, the teachers developed other types of
teaching goals, which, in their view, might have some lasting impact on students: for
example, teaching students 'manners' and how to behave in the classroom, or
320 JOURNAL OF CURRICULUM STUDIES 1 9 : 4

teaching general learning skills, such as how to outline a chapter or organize a

notebook.
Affect and evaluation can thus be important regulators of the amount of energy
teachers will put into activities and how they will expend energy on an activity.

Episodic storage

Abelson suggests that information in knowledge systems is stored primarily in

semantic networks, while belief systems are composed mainly of 'episodically'-
stored material derived from personal experience or from cultural or institutional
sources of knowledge transmission (e.g., folklore).16 Broadly speaking,
semantically-stored knowledge is thought to be broken down or 'decomposed' into
its logical constituents (abstract semantic categories —principles, propositional
structures, or whatever) and organized in terms of semantic lists or associative
networks. Episodic memory, by contrast, is organized in terms of personal
experiences, episodes or events.17
Although suggestive, the semantic/episodic distinction remains controversial18
and provides a weak basis for distinguishing 'beliefs' from 'knowledge* (there are, for
example, a number of 'episodic' models of knowledge).19 A stronger and narrower
claim is that beliefs often derive their subjective power, authority, and legitimacy
from particular episodes or events.20 These critical episodes then continue to colour
or frame the comprehension of events later in time.
There were clear indications in the present study that such critical episodes
played important roles in teachers' practices. Ms Skylark's vision of an alternative
classroom environment organized around friendly relations and 'fun' activities, for
example, was derived as a contrast to her own, very vividly remembered, experiences
as a student. The experiences of Mr Ralston, a mathematics teacher whose
undergraduate training had been in agricultural education and who had spent many
years teaching mathematics for metalworking students in the Job Corps, led him to
believe that students would be more willing to study mathematics if they could see
that it had some 'practical' value —and he tried to organize his instructional activities
according to this assumption. More generally, a number of teachers suggested that
critical episodes or experiences gained earlier in their teaching careers were
important to their present practices.
Such critical episodes are probably at the root of the fact that teachers learn a lot
about teaching through their experiences as students —experiences that have been
referred to as 'apprenticeships' to teaching21 or participant observation22 of teaching
practices. In fact, however, being a student probably rarely entails the reflective and
systematic study that such terms imply. Instead, it seems more likely that some
crucial experience or some particularly influential teacher produces a richly-detailed
episodic memory which later serves the student as an inspiration and a template for
his or her own teaching practices. As discussed later, episodic memory structures of
this type may be highly adaptive for dealing with processes in ill-structured domains
or ambiguous and complex domains.23

Non-consensuality

Unlike the four characteristics described above, non-consensuality is a feature of

belief systems rather than of individual beliefs. Indeed, non-consensuality is basically
BELIEFS IN TEACHING 321

a consequence of the features described above. Simply put, it refers to the fact that
belief systems consist of propositions, concepts, arguments, or whatever that are
recognized — by those who hold them or by outsiders—as being in dispute or as in
principle disputable.
One might well ask how a situation of non-consensuality differs from a situation
in which people simply differ in the amount or quality of their knowledge about some
event or process. One way to answer this is to suggest that belief systems are less
malleable or dynamic than knowledge systems. Knowledge accumulates and
changes according to relatively well-established canons of argument.24 Beliefs, by
contrast, are relatively static (at least in terms of their core applications —see the
discussion of 'unboundedness' below). When beliefs change, it is more likely to be a
matter of a conversion or gestalt shift than the result of argumentation or a
marshalling of evidence. One can say, then, that part of the consensus characterizing
knowledge systems is a consensus about the ways in which knowledge can be
evaluated or judged. By contrast, much of the non-consensuality of beliefs derives
from a lack of agreement over how they are to be evaluated. As already suggested,
belief systems often include affective feelings and evaluations, vivid memories of
personal experiences, and assumptions about the existence of entities and alternative
worlds, all of which are simply not open to outside evaluation or critical examination
in the same sense that the components of knowledge systems are.

Unboundedness

Belief systems can be described as loosely-bounded systems with highly variable and
uncertain linkages to events, situations, and knowledge systems.25 In other words,
there are no clear logical rules for determining the relevance of beliefs to real-world
events and situations. Moreover, these linkages and definitions of relevance may well
be bound up with the personal, episodic, and emotional experiences of the believer.
One way of looking at this feature of belief systems is to say that beliefs have stable
core applications26 (a domain of events and situations to which they are consistently
held to be applicable-these could derive from the 'critical episodes' described above
in the discussion of the episodic underpinnings of beliefs) but that they can be
extended in radical and unpredictable ways to apply to very different types of
phenomena. Knowledge systems, by contrast, generally have relatively well-defined
domains of application, and can be expanded to encompass other phenomena only
through the application of strict rules of argument. What the concept of unbounded-
ness means, then, is that people read belief-based meanings into situations where
others would not see their relevance.
The preceding discussion raises the questions of why beliefs about teaching
with these structural characteristics exist and what role they play in teaching. The
research of the TBS suggests that beliefs and belief systems have two important uses
for teachers —(a) task definition and cognitive strategy selection; and (b) facilitation
of retrieval and reconstruction in memory processes-while serving the overall
function of allowing teachers to deal with ill-structured domains.
322 JOURNAL OF CURRICULUM STUDIES 1 9 : 4

The uses of beliefs

Task definition

Non-consensuality, existence beliefs, and beliefs in alternative worlds make belief

systems very important determinants of how individuals organize the world into task
environments and define tasks and problems. However, the relationships between
cognition, task definition, and performance are highly complex. On the one hand, the
structure of a task-the goals, constraints, and resources that constitute it-may be
said to determine or specify the kinds of processing strategies to be used in
accomplishing it. On the other , the 'task' itself must first be denned: an individual
must perform some prior processing upon the virtual 'task environment' to produce
an actual task or 'problem space'.27
The importance of such task definition is most strikingly apparent in cross-
cultural psychology where non-school subjects typically refuse to stay within the
boundaries of the problem as defined by the experimenter.28 In other words,
subjects may create problem spaces (or 'actual tasks') radically at odds with the
experimenter's conception of the task. Following Schoenfeld, this process of task
definition will be considered a function of the belief systems of those participating in
the task.29 From this perspective, beliefs perform the function of 'framing' or
defining the task at hand. Implicit in this argument is a view of cognitive processing
as entailing several qualitatively different levels or categories of thought.30
First, a microscopic level of internal processing can be identified. This is the level
consisting of the largely automatized and procedural processes of perception which
take place without conscious attention. These nuts and bolts of cognition, consist of
representational structures, processing characteristics, memory mechanisms and the
like.31
A second level of thought could be termed resources, the knowledge possessed by
the individual, that can be brought to bear on the problem at hand.32 Included
within this category would be domain-specific knowledge, facts, algorithms, local
(problem-specific) heuristics and the like. These constitute the tools or tactical
resources of thought. As Schoenfeld argues, there are at least two distinct issues here:
one is possession of knowledge and the other is access to that knowledge.33 One can
have the knowledge necessary to solve a problem but not recognize that the
knowledge is relevant or not know how to apply it to the problem.
A third category of thought, control or metacognitive processes, refers to the
deliberate, conscious control and co-ordination of cognitive resource use in
problem-solving.34 If the category of cognitive resources described earlier can be
thought of as the level of tactics, the category of control processes now under
examination can be thought of as the level of strategy.
Strategic thought, then, involves the conscious selection and use of tools of
thought from a repertoire of cognitive resources to solve a certain type of problem.
But how do people know or recognize the problems they're dealing with? This is the
point at which belief systems—a. fourth category of thought — become important
determinants of task or problem definition. To make this point clearer, let us
consider an alternative and potentially confusing notion of task definition.
It has been shown that 'experts' in a given domain of activity 'see' problems or
task environments differently than novices.35 This sort of task definition differs from
the type we are concerned with here. Expert/novice differences in task definition
BELIEFS IN TEACHING ' 323

refer to the ways in which individuals with different resources and domain-specific
knowledge interpret the surface configurations of components in a well-defined
problem space. Experts, so to speak, perceive basic deep structures underlying a vast
range of possible surface structures. But the key here is that these individuals —
whether experts or novices —are dealing with a task environment where goals,
constraints, and operations are clearly defined. It is, for example, a question of
relative skill or knowledge in a known domain (e.g., chess) rather than a question of
trying to define the game being played, and to decide why one is playing it or what
kinds of consequences might stem from winning or losing. These latter issues are
those with which belief systems are concerned.
We can thus think of cognitive resources, metacognitive control strategies, and
belief systems as progressively more encompassing systems of thought. Task
environments are defined in terms of belief systems. Metacognitive strategies are
employed to select among the available cognitive resources to carry out the task.
(This is, of course, a simplification: there will almost certainly be some feedback and
interaction among the various levels of thought).
The importance of this point for understanding teaching and teacher education
has already been suggested: to understand teaching from teachers' perspectives we have
to understand the beliefs with which they define their work. Consider two of the history
teachers in the sample. One, Mr Larson, saw teaching mainly as a job, a form of
labour, a way of making a living. Ms Marsh, by contrast, looked at teaching as a sort
of moral mission to socialize children and better the community; money for her was
secondary.36 Teaching, in short, took on completely different meanings for these two
teachers, and failure to recognize this would vitiate any attempt to make sense of
what they did in the classroom or why they did it. Similarly, it would be difficult if
not impossible to teach teachers with such different orientations using the same
methods and expect similar results, or any results at all.

Facilitation of memory process

Beliefs, as suggested earlier, frequently involve moods, feelings, emotions, and

subjective evaluations. These features make beliefs important in memory processes.
Mood and emotion seem to be stored in long-term memory; they take the form of
gestalts that can be highly organized for efficient representation and retrieval, and
they seem to require very little in the way of allocated processing-capacity.37 Spiro
suggests that mood and emotion are stored as analogue representations of the
experiential states associated with bodies of propositional knowledge. They function
as a form of background coloration to content representation, the nature of which
'corresponds to the nature of the felt experience'. When events are associated with a
single or dominant experiential quality, their cognitive representation will have a
relatively homogeneous coloration and one can speak of the event as having a
'signature feeling' , 3 8
Spiro suggests that such coloration serves at least three purposes. First, it
facilitates recall: background coloration, figuratively speaking, is visible from greater
distances than is specific content. 'Memory files' can be initially scanned only for
coloration and only those memory areas with the appropriate coloration will then be
'magnified' to allow for the retrieval of content specifics.39
Second, emotional or attitudinal coloration supplies cohesion to elements in
memory. To the extent that content experiences correspond closely to homogeneous
324 JOURNAL OF CURRICULUM STUDIES 1 9 : 4

emotional, evaluative, or attitudinal qualities, the coloration acts as a kind of cohesive

glue holding memories together over time.40
Finally, experiential coloration performs an important function in construc-
tive and reconstructive memory processes.41 It is now generally thought that
memory entails more than the simple abstraction and storage of unaltered and
unedited memory traces. Instead, the representations of events in memory are
partial constructions of events based on an incomplete sampling of the available
information. These incomplete representations are then typically fleshed-out or
reconstructed during recall. Spiro argues that signature feelings constrain re-
constructive processes by filtering out that information which fits logically and
functionally with the event but distorts or conflicts with the stable signature
feeling.42
In summary, then, the affective and emotional components of beliefs can
influence the ways events and elements in memory are indexed and retrieved and
how they are reconstructed during recall. Emotion and affect thus have important
implications for how teachers learn and use what they learn.

The functions of beliefs

It can certainly be argued that teachers' beliefs play a major role in defining teaching
tasks and organizing the knowledge and information relevant to those tasks. But why •
should this be so? Why wouldn't research-based knowledge or academic theory serve
this purpose just as well? The answer suggested here is that the contexts and
environments within which teachers work, and many of the problems they
encounter, are ill-defined and deeply entangled, and that beliefs are peculiarly suited
for making sense of such contexts.

Dealing with ill-structured problems and entangled domains

The terms 'ill-structured problems' and 'entangled domains' require- some expla-
nation. As Simon suggests, there is no precise boundary separating well-structured
and ill-structured problems.43 Instead, problems may be seen as fitting into a
continuum as they vary along the following dimensions.
First, problems differ according to the nature of their goals. Is there a precisely-
definable goal state? Is there a single goal or a set of related goals, or are there a variety
of inconsistent or unrelated goals? Are there criteria available for determining
whether and when a goal has been attained?
Second, how well defined is the 'technology' or set of procedures for attaining
the goal or goals? Can a set of actions be defined that will invariably lead to desired
consequences (or can a probability be assigned to the likelihood of the goal being
attained through a given course of action)? Can one determine, even in retrospect,
what course of action led to an observed outcome?
Third, ill-structured problems are problems that require people to go beyond
the information contained in the problem instruction and use background knowl-
edge or make guesses or assumptions in order to solve the problem. There are no
search strategies that allow one to identify the information relevant to the problem.44
Finally, in ill-structured problems, alternative courses of action at different
points of the problem-solving process are not clearly-defined: it is impossible to
BELIEFS IN TEACHING • 325

define or identify the complete range of optional courses of action at each point in the
process.45 Not only is the problem-solver uncertain of what should be done, he or
she is uncertain of what can be done.
The concept of an entangled domain has to do with instances or examples or
entities which can be identified by some criteria as belonging to a given domain, but
which at the same time do not all share some important sets of criteria and do not fall
into relationships of dominance and subsumption with each other. Thematic
features overlap only partially and incompletely across the domain.46
When people encounter entangled domains or ill-structured problems, many
standard cognitive processing strategies — for example, schema-abstraction or ana-
lytical reduction - are no longer viable.47 Instead, because one could never be sure
just what information would be needed to deal in an adaptive manner with such
domains, one would need to encode as much information as possible in as many ways
as possible. Such knowledge would take the form of rich, contextualized, highly
multivariate descriptions of large numbers of individual cases—knowledge that
could be examined and reexamined from many different perspectives.48 The
episodic cores of belief systems would seem to be good candidates for these types of
knowledge structures. Because belief systems are unbounded, these cases could be
mapped onto a vast range of new events or experiences. Because belief systems are
non-consensually held and often involve ideals of alternative arrangements of reality,
the mappings of these cases onto new settings would be relatively immune to easy
falsification or contradiction. Finally, the structural feature of existential presump-
tion would serve the function of importing stable and predictable contextual
characteristics into ill-defined settings (e.g., allowing the mathematics teachers
described earlier to interpret student actions across many classes or school contexts -
they would not have to invent new theories of student action each time the student
population changes).
The argument that teaching is ill-structured or entangled is by no means new.49
However, it should be stressed that some aspects of teaching in some situations may
be well defined. That much of a teacher's work outside the classroom (as a member of
the school organization or interacting with parents or community members) may be
ill-defined as well (obviously the type of school system a teacher works in, the type of
subject matter a teacher teaches, and the career path of the teacher has much to do
with whether and to what extent the teacher will experience teaching and schooling
as entangled or ill-defined).50

Implications

The model of beliefs presented here has a number of implications for our
understanding of teaching and teacher education. Most obviously, it suggests that if
we are interested in why teachers organize and run classrooms as they do we must pay
much more attention to the goals they pursue (which may be multiple, conflicting,
and not at all related to optimizing student learning) and to their subjective
interpretations of classroom processes. This applies not only to behaviourist-
326 JOURNAL OF CURRICULUM STUDIES 1 9 : 4

inspired process-product models of teaching, but to information-processing and

decision-making models that rely on dubious simplifying assumptions about
teachers' goals51 and to activity structure or task models that ignore the fact that
classroom structures have sources in teachers' beliefs and are maintained over time
only as a result of the goal-directed actions of teachers and students.
This argument is similar in part to the one advanced by Fenstermacher from an
intentionalist perspective.52 It differs from that account in its implications for
teacher education. Fenstermacher suggested that teachers' 'subjectively reasonable
beliefs', once reflexively recognized, could be altered or transformed by being shown
to be 'objectively unreasonable'. However, if beliefs have the structural character-
istics suggested here, then, strictly speaking, they cannot be shown unreasonable or
false: beliefs are not so much sets of propositions or statements as they are conceptual
systems which are functional or useful for explaining some domain of activity.
Because beliefs play a major role in defining tasks, a sort of autodeterminism53 occurs
in which phenomena inconsistent with the beliefs are defined as belonging to
unrelated domains of activity. Thus to show that beliefs cannot adequately account
for aspects of reality other than those they are explicitly attached to (and do 'explain')
is no falsification of those beliefs.54
This line of argument implies that prospective teachers' perceptions of and
orientations to the knowledge they are presented with may be shaped by belief
systems beyond the immediate influence of teacher educators. The development of
beliefs over time, as a product of teachers' long-term comprehension of different
contexts for teaching5 would appear to be similarly difficult to predict, control, or
influence.
At least two lines of response appear plausible. The first would be to address the
features of teaching as an occupation that make beliefs so important. One could, for
example, routinize teaching to the extent that teachers could be taught recipe-like
pedagogical methods, adherence to which could be closely monitored and regulated.
That is, one could transform teaching into a set of well-defined tasks and thus reduce
the role played by beliefs in defining and shaping tasks. This amounts to trying to
improve teacher education by changing 'teaching' itself.
A second line of response would entail trying to change or shape teachers'
beliefs. In part this would mean helping teachers and prospective teachers become
reflexive and self-conscious of their beliefs and, as Fenstermacher suggests,
presenting objective data on the adequacy or validity of these beliefs. However, this
can result in transformations of teachers' beliefs and practices only if alternative or
new beliefs are available to replace the old. This is the crux of the problem: we do not
know very much about how beliefs come into being, how they are supported or
weakened, how people are converted to them, and so on. Socialization, the social
context of the school and other processes and constraints have been suggested as
likely sources of beliefs, but just how they operate on beliefs is far from clear. Just as
important is the fact that attention to beliefs-and especially the idea of consciously
structuring and -inculcating certain kinds of beliefs in teachers - raises questions of
values and superordinate goals that are far from resolved: what should the end result
of education be? What kind of affective climate should characterize classrooms? How
should one conceptualize student characteristics? How should one orient oneself to
the community served by one's school? These are not questions that can ever be
resolved by experimental or correlational studies, but they are questions that must
nevertheless be addressed.
BELIEFS IN TEACHING 327

Acknowledgements

This study was supported by the US National Institute of Education under Grant
NIE-G-83-006, Research and Development Center for Teacher Education, Univer-
sity of Texas at Austin. The opinions expressed herein do not necessarily reflect the
position or policy of the National Institute of Education, and no official endorsement
by that office should be inferred.

References

1. CLARK, C. M., and PETERSON, P. (1986) Teachers' thought processes. In Wittrock, M. C. (ed.),
Handbook of Research on Teaching, 3rd edn (Macmillan, New York), pp. 255-296; SHAVELSON, R. J.,
and STERN, P. (1981) Research on teachers' pedagogical thoughts, judgments, decisions, and
behaviour. Review of Educational Research, 51 (4), pp. 455-498.
2. FENSTERMACHER, G. (1979) A philosophical consideration of recent research on teacher effectiveness.
In Shulman, L. S. (ed.), Review of Research in Education, vol. 6 (F. E. Peacock, Itasca, IL), pp. 157-
185; SCHOENFELD, A. (1983) Beyond the purely cognitive: belief systems, social cognitions, and
metacognitions as driving forces in intellectual performance. Cognitive Science, 7, pp. 329-363.
3. BORKO, H. (1978) An examination of some factors contributing to teachers' preinstructional
classroom organization and management problems. Paper presented at the annual meeting of the
American Educational Research Association. Toronto, Canada.
4. ELBAZ, F. (1983) Teacher Thinking: A study of practical knowledge (Nichols Publishing Company,
New York); JANESICK, V. (1977) An ethnographic study of a teacher's classroom perspective. Doctoral
dissertation, Michigan State University, East Lansing, MI.
5. NESPOR, J. (1984) The interaction of school context and teachers' beliefs, R&D Rep. No. 8023,
Research and Development Center for Teacher Education, University of Texas, AustinTX,ERICED
260079; NESPOR, J., (1985) The role of beliefs in the practice of teaching: final report of the Teacher
Beliefs Study, R&D Rep. No. 8024. Research and Development Center for Teacher Education,
University of Texas, Austin, TX.
6. KELLY, G. A. (1955) The Psychology of Personal Constructs (W. W. Norton, New York).
7. KING, L. and TUCKWELL, N. (1983) Stimulated Recall Methodology. Issues in Educational Research
IX (The Western Australian Institute for Educational Research, Perth, WA).
8. NESPOR (1985), pp. 193-214 (see note 5).
9. NESPOR (1985), pp. 29-66 (see note 5).
10. Cf., ABELSON, R. (1979) Differences between belief systems and knowledge systems. Cognitive
Science, 3, pp. 355-366.
11. Ibid., p. 357.
12. Ibid., pp. 357-358.
13. Cf., ibid., pp. 357-358.
14. Ibid., p. 358.
15. ZAJONC, R. (1980) Feeling and thinking: preferences need no inferences. American Psychologist, 35,
pp. 151-175.
16. ABELSON (1979), pp. 358-359 (see note 10).
17. SCHANK, R. and ABELSON, R. (1977) Scripts, Plans, Goals and Understanding: An Inquiry into Human
Knowledge Structures (Lawrence Erlbaum, Hillsdale, NJ), p. 17.
18. See e.g., TULVING, E. (1983) Elements of Episodic Memory (Oxford University Press, New York).
19. SCHANK and ABELSON (1977) (see note 17).
20. AYEROFF, F. and ABELSON, R. (1976) ESP and ESB: belief in personal success at mental telepathy.
Journal of Personality and Social Psychology, 34, pp. 240-247; NISBETT, R., BORGIDA, E., CRANDALL,
R., and REED, H. (1976) Popular induction: information is not necessarily informative. In Carroll, J.
and Payne, J. (eds.), Cognitive and Social Behavior (Lawrence Erlbaum, Hillsdale, NJ), pp. 113-134;
TVERSKY, A. and KAHNEMAN, D. (1973) Availability: a heuristic for judging frequency and
probability. Cognitive Psychology, 4, pp. 207-232.
21. LORTIE, D. (1975) Schoolteacher: A Sociological Study (University of Chicago Press, Chicago).
22. EDDY, E. (1969) Becoming a Teacher: The Passage to Professional Status (Teachers College Press, New
York).
23. NEWELL, A. (1969) Heuristic programming: ill-structured problems. In Aronofsky, J. (ed.), Progress
in Operations Research, vol. 3 (John Wiley, New York); SIMON, H. (1973) The structure of ill-
structured problems. Artificial Intelligence, 4, pp. 181-202.
328 BELIEFS IN TEACHING

24. TOULMIN, S. (1972) Human Understanding, vol. 1 (Princeton University Press, Princeton, NJ); cf.
FEYERABEND, P. (1978) Against Method (Verso, London); KUHN, T. (1970) The Structure of Scientific
Revolutions (University of Chicago Press, Chicago).
25. ABELSON (1979), pp. 359-360 (see note 10).
26. STEGMULLER, W. (1976) The Structure and Dynamics of Theories (Springer, New York).
27. BEAUGRANDE, R. (1980) Text, Discourse and Process (Ablex, Norwood, NJ); SIMON, H. (1978)
Information-processing theory of human problem-solving. In Estes, W. (ed.), Handbook of Learning
and Cognitive Processes, vol. 5 (Lawrence Erlbaum, Hillsdale, NJ), pp. 271-296.
28. COLE, M. and SCRIBNER, S. (1974) Culture and Thought (John Wiley, New York), p. 168; see also
NEISSER, U. (1976) General, academic, and artificial intelligence. In Resnick, L. (ed.), The Nature of
Intelligence. (Hillsdale, NJ, Lawrence Erlbaum), p. 136.
29. SCHOENFELD, (1983), pp. 336-337 (see note 2).
30. SCHOENFELD (1983), pp. 330-337 (see note 2).
31. See discussions by BEAUGRANDE, R. (1981) Design criteria for process models of reading. Reading
Research Quarterly, 16 (2), pp. 261—315. MONSELL, S. (1981) Representations, processes, memory
mechanisms: the basic components of cognition. Journal of the American Society for Information
Science, 32 (5), pp. 357-363.
32. SCHOENFELD (1983), p. 331 (see note 2).
33. Ibid., pp. 332-333.
34. BROWN, A. (1980) Metacognitive development and reading. In Spiro, R., Bruce, B. and Brewer, W.
(eds.), Theoretical Issues in Reading Comprehension (Lawrence Erlbaum, Hillsdale, NJ), pp. 453—454;
FLAVELL, J. (1976) Metacognitive aspects of problem-solving. In Resnick, L. (ed.), The Nature of
Intelligence. (Lawrence Erlbaum, Hillsdale, NJ); SCHOENFELD (1983), pp. 333-335 (see note 2).
35. CHI, M., GLASER, R. and REES, E. (1981) Expertise in Problem-Solving (Learning Research and
Development Center, University of Pittsburgh, Pittsburgh).
36. See NESPOR (1984) (see note 5).
37. SPIRO, R. (1982) Subjectivity and memory. In Le Ny, J.-F. and Kintsch, W. (eds.), Language and
Comprehension (North-Holland, New York), pp. 29-34.
38. Ibid., p. 31.
39. Ibid., p. 31.
40. Ibid., pp. 31-32.
41. See LOFTUS, E. F. (1979) Eyewitness Testimony. (Harvard University Press, Cambridge, MA); SPIRO,
R. (1977) Remembering information from text: the 'State of Schema' approach. In Anderson, R.,
Spiro, R. and Montague, W. (eds.), Schooling and the Acquisition of Knowledge (Lawrence Erlbaum,
Hillsdale, NJ), pp. 137-166.
42. SPIRO, (1982), p. 32 (see note 37).
43. SIMON (1978), p. 286 (see note 27).
44. Ibid.
45. Ibid.
46. SPIRO, R. and MYERS, A. (1984) Individual differences and underlying cognitive processes in reading.
In Pearson, P. D. (eds.), Handbook of Reading Research (Longman, New York), pp. 492-493.
47. See SPIRO and MYERS (1984), pp. 492-496 (see note 46).
48. Ibid.
49. GREENE, M. (1979) Teaching: the question of personal reality. In Lieberman, A. and Miller, L. (eds.),
Staff Development: New Demands, New Realities, New Perspectives (Teachers College Press, New
York), pp. 27-28; MARCH, J. and OLSEN, J. (eds.), (1976) Ambiguity and Choice in Organizations
(Universitetsforlaget, Bergen, Norway); WEICK, K. (1979) The Social Psychology of Organizing 2nd
edition (Addison-Wesley, Reading, MA).
50. NESPOR (1984, 1985) (see note 5).
51. E.g., SHAVELSON and STERN (1981), p. 471 (see note 1).
52. FENSTERMACHER (1979) (see note 2).
53. STEGMULLER (1976), pp. 196-202 (see note 26).
54. Ibid., p. 19.
55. See NESPOR (1984) (see note 5).