Notes on: Moore R and Maton, K. (2001) 'Founding the Sociology of Knowledge: Basil Bernstein, intellectual fields and the epistemic device. In Morais, A., Neves, I., Davies, B. & Daniels, H. (eds) Towards a Sociology of Pedagogy: The contribution of Basil Bernstein to research. New York, Peter Lang, 153-182.

Dave Harris

The intrinsic relations of knowledge have not been well understood, especially not by the new sociology of education.  It is not just that this neglected to study, more that it was a blind spot for the sociology of knowledge that could not be penetrated.  That is why we need new tools, Bernstein on knowledge structures and grammars which can describe differences between intellectual fields 'in terms of the organising principles of their knowledge formations' (154).  This will lead us to focus on the production of knowledge, which in turn depends on knowledge claims.  This can be seen as completing Bernstein [with a familiar picture of his progress through pedagogy to knowledge structures (155)].  The rereading enables us to see even the early pedagogic work as implying different notions of knowledge, and with 'empirical research' making new objects visible [the actual process seems to be to develop a classification and then to investigate the cells - empirical research is not problematized as we shall see, and it seems to be covered equally by illustrative fictions and unproblematic histories].  We need to look at an epistemic device which, like the pedagogic device, explains why certain kind of knowledge comes to be seen as legitimate, or as 'altering relations between the arbitrary and non arbitrary in knowledge', where the former relates to power relations, and the latter to 'principles intrinsic to knowledge itself', 'relations within'(156).

Bernstein has accused the sociology of education of ignoring the analysis of 'pedagogic discourse itself and its intrinsic features', treating pedagogy as a mere relay for external power relations.  If we look at the organisation of knowledge we can detect some underlying principle to regulate the production of new knowledge and its form.  As this principle varies, so do modes of production, so that certain things become visible and potential objects for knowledge.  It is a difference of principle, 'not just a focus or perspective'.  [Which raises the question immediately of what these principles are - Maton goes on to claim that they seem to be some universal forms, rather like Durkheim's classifications?].  This explains blind spots, which arise from the very constitution of the intellectual field.

Bernstein explains what this principle amounts to.  There are different structures of knowledge, like hierarchical vs. horizontal ones, and differences in grammar between weak and strong.  Thus the natural sciences have an integrated hierarchy of knowledge, while humanities have a horizontal and segmented collection of specialised languages each with their own modes of interrogation and criteria.  The latter include the perspectives in sociology [here, integration does relate to the vertical and collection to the horizontal, which is the commonsense way to see it -  I am sure it is used the other way around in some other Maton pieces].  Development for hierarchical forms involves greater generality and integration, while in the horizontal ones, new languages or segments are added.

Grammar refers to the conceptual syntax which provides empirical descriptions and/or formal models of empirical relations.  Strong grammars can be found even in horizontal knowledge structures which include 'mathematics, logic, and economics'.  Weak ones are found in social anthropology, cultural studies and sociology.  Used together , we can systematically describe differences between intellectual fields in terms of their organising principles.  Now we have to explain the practices which generate the differences, again in terms of 'underlying generative principles'.

Knowledge is both produced and then recontextualized or reproduced.  The former cannot be reduced to the latter, since the two fields of practice are specific, and what might be tacit in one becomes evident in the other.  The pedagogic device plays a major part in reproduction, by transforming knowledge into pedagogic discourse.  Knowledge production highlights new issues, however, by focusing on knowledge claims and relations between arbitrary and non arbitrary elements [the arbitrary now has a broader definition as something which is related to social relations, especially power, and the non arbitrary is justified as something which cannot be so related].  These issues are not so important with pedagogy.  Different knowledge claims can deny that anything is non arbitrary, or impose a particular definition of the non arbitrary as in positivism.  [The issue becomes one of examining how Bernstein defines the non arbitrary].  The basis of knowledge claims is crucial to the existence of intellectual fields, however [really?  Not just an occupational ideology?]. 

The epistemic device is an analogue of the pedagogic device, looking at how generating principles operate.  This leads to Maton on the principles of legitimation, which combine to produce different modes of legitimation.  The principle regulates the combination of the arbitrary and non arbitrary.  The epistemic device shows ways in which 'actors, groups of actors, or institutions may alter these relations', competing over legitimation claims and the status of the field as a result.  This is seen as a 'precondition of knowledge production'(161), having a positive function of establishing knowledge claims [merely as an effect of power?].  The epistemic device is the object and stake of struggle, 'the key to symbolic domination'.

We need to illustrate here, to avoid confusion, like that which is affected Bernstein's work on the difference between knowledge codes, and the actual rules employed by the pedagogic device.  This confusion arises from 'an empiricist tendency to substantialism' [described confusingly as 'asking where the device may be seen rather than when'.  A way to dismiss criticisms of this sort of 'empirical' approach?]

We can illustrate  different strengths of grammar in the sociology of education [which we have already agreed is a horizontal structure].  Changes in the 1970s have been seen as a paradigm shift, using Kuhn.  We can now 'imagine the following scenario'(162) [basically where a horizontal knowledge structure shifts from describing itself as a series of perspectives to one of incommensurable paradigms.  The former still has some elements in common and sees the differences as perspectives.  It's the grammar that is common.  With incommensurable paradigms, there is no agreement about what counts as a question or solution, and no common ground.  The field has been divided into not specialisms but 'exclusively specialised knowers... each with their own distinctive and incommensurable language'. The shift helps Maton bring in his stuff about knowers, but I am not sure that this was the decisive issue between the new and the old sociology of education - there were epistemological differences as well, even if they were not very well clarified.  At the same time, the differences between knowers was also apparent in the old specialist approach, with differences between elite and upstart departments, or with status differences attached to quantifiers].  Overall, grammar has been weakened, [since there are no shared classifications], and this produces 'a comparatively low level of integrative power'(163)

There were still some conflicts between the two ways of characterising the field, since the old guard did not always accept the new epistemology ['or, more precisely, anti epistemology', (164), which enables him to preserve the term for his preferred kind of internal relations].  Claiming to have developed a new paradigm is an attempt to restructure the nature of the field and split its grammar, so it is about 'languages of legitimation', about what counts as knowledge.  Actors are struggling to define the field, to control the epistemic device.  We see in this struggle different principles of legitimation, and the effects include judging how achievement within the field should be measured.  A new relation between the arbitrary and non arbitrary is being proposed, [disguised as] debates about whether or knowledge is to be understood as entirely sociological and historical or as 'ontologically necessary', struggles between epistemic and social relations of knowledge. [Maton is attributing the split tothe struggles between actors, a social relation! He is not even prepared to see internal contradictions and problems with the old approaches, not even as does Kuhn? ]

These are analytically distinct, 'but empirically interrelated'(165).  Languages of legitimation offer different classifications and frames around these two relations, producing the usual configurations of strong or weak classification and framing.  The claim is then that seeing a field as composed of perspectives and specialisms produces strong classification and framing of the epistemic relation, but weak classification and framing of the social relation since 'knowers are not the issue'.  In the paradigms approach, the strengths are reversed, and who makes the claim becomes crucial, 'regardless of the procedures used and the object studied' [surely a substantial exaggeration].  What we end with is a knower mode.  The struggle establishes status hierarchies [undoubtedly present, as in the production of colonies of ethnomethodologists who went around promoting their mates - but ever present?].

Such switches of paradigm have been identified before.  We're not talking about dichotomies, nor preferring one rather than the other - each approach has relative strengths and weaknesses, compared to each other and compared to other fields.  We're talking about changes rather than total switches.  These are not ideal types 'which remain at the level of the empirical [??]; rather, they represent real principles whose empirical realisations are dependent on the enabling context' (166).  There is a tendency to conceive of grammars as absolute states, but this arises from the false dichotomies constructed within the fields themselves.

We can examine two contrasting examples, literary criticism and mathematics, and explain their differences in terms of these modes of legitimation.  We acknowledge that the strength of grammar also depends upon 'the condition of the field in terms of its location within the broader structural dynamic of the education system', which turns out to refer to factors such as educational expansion, state policies, the extent of external regulation.  We're not going to examine those issues here, at the expense of producing 'necessarily partial' accounts (167).  Conventional sociologies of knowledge are required here.

The switch is sometimes seen in terms of a generational conflict, but this would not account for the schismatic radical break being suggested.  The 1960s was a time for announcing radical breaks.  There have been considerable segmental breaks, for example the move towards '"standpoint" theory and "voice" discourse'(168).  Such breaks are so common that sociology looks like it's permanently revolutionary.

The same characteristics affected literary criticism, with the proclamation of a new form in the early 1960s, associated with Kermode.  It featured the familiar 'apocalyptic cosmology', which made all the old approaches redundant.  It was an attempt to close the field.  It implied that the location in time can be the basis for a knowledge claim, just as with postmodernism.  It also implied that only a few knowers could see this new world, located at points in space, time and society.  The knower became more central, and this provided a problem for communication between different groups of knowers [the struggle at one stage involved exclusion of knowers].  Here, social relations are strongly classified and framed, producing little epistemic communities or segments, but there is an inherent instability and a tendency to even greater proliferation and fragmentation.

There was also the notion of a creative fiction, apparently claiming that the world had changed, even though it was the conditions of some members of the intellectual field that had changed [exactly what Bourdieu says about people like Barthes].  The same might be said about those experiencing the postmodern condition.  Such statements are creative fictions, heuristic devices, part of Bourdieu's scholastic fallacy [maybe].  The proposed change is just announced and becomes an article of faith.  Relations with actual objects of enquiry show weak classification and framing, since it is the gaze of the knower which is crucial, the ability to see the new world.  This clearly excludes everyone else and makes everything in the past redundant.

In mathematics, radical breaks or ruptures are relatively rare [requires substantial empirical support here surely?].  Mathematics has perspectives.  It is still horizontal, but with a particularly strong grammar.  The lengthy discussion about Fermat's Last Theorem shows continuing work over a great deal of time, and effective communication with quite different people inside an epistemic community.  There is no particular focus on the knower [although the person solving the Theorem gained a great deal of status].  It is the object itself that is supposed to produce the specialist language and in turn specialist knowers.

There are still mathematical fictions, invented problems and mathematical structures and worlds, but mathematicians 'cannot explore these worlds just as they like'(173), and must follow an agreed strong grammar and criteria solutions to problems.  These are constant, and do not depend on the status of the author.  This helps mathematical work 'transcend specific worlds and endure over time'(174).  The findings are open for use by anybody, and previous work can be developed regardless of context.  The official motivation is only intellectual or mathematical considerations, regulated by the object itself.  The social relation of knowledge is weak in terms of classification and framing.  History itself is at least partially negated, avoiding the impact of historical contexts.  The personal characteristics of Fermat are irrelevant.  Thus development can proceed over time cumulatively, and without its building impenetrable barriers between specialisms.

So both literary criticism and mathematics have horizontal knowledge structures, although they differ in their relative strengths of grammar.  Using these concepts from Bernstein helps us describe these fields 'beyond their empirical and substantive characteristics' (175).  That's why we need to look at the principles, and the 'actual settings for the epistemic device'.  We can also understand struggles as attempts to seize control of the epistemic device, even though they may masquerade as generational ones.  Once knower modes are established, they are difficult to change, 'though the empirical realisations of the principle may be subject to ceaseless change'.  For knowledge modes, there is an emphasis on the 'non arbitrary relation of procedures to their objects of study', which produces an agreed strong grammar and an extended community.  Struggles over the epistemic field are not relevant, since 'it is, so to speak, the social property of the field itself'.  This produces paradoxes such as excessive individualism of sociology.  The affects of particular settings of the epistemic device appear in, for example the relation to the past and whether it is a continuing reservoir or something to be abandoned.

Social and institutional factors are still important 'in the enacted practices of intellectual fields' (176), but there is a separate force and effect of the 'self representations of a field's operations', for example the way in which members of the field define it for themselves [with a reference to Popper here on the importance of the structures of fields in themselves]. Languages of legitimation are important perspectives toward the field and they affect the relation of the community to the field.  Existing approaches to knowledge need to discuss these factors as well, since 'the epistemic shape of an intellectual field has ramifications for its social form'.  The social [actually 'sociological'] nature of knowledge itself has been neglected so far by the sociology of knowledge.

The epistemic device is analogous to the pedagogic device.  The former regulates who can produce legitimate knowledge and how knowledge is selected and transformed to produce acceptable new knowledge.  Together with the pedagogic device, it explains knowledge production, recontextualization and reproduction, although it is less important in the last two.  However all new knowledge has to be recontextualized, and 'all educational knowledge is subject to the epistemic device' (177) [that is, legitimation claims are common in classrooms].

It is not just social relations in classrooms, since the 'epistemological nature of social relations is similarly universal and ubiquitous - if it were not we would not be able to function on a day to day basis'.  This means there is no simple division between social principles' and logically formal epistemological principles, 'as realist philosophy of science has long been recognized': the logic of discovery is social but in a special way.  Truth is also important to every day effective action.  The epistemic device is social, and is most systematically developed in university disciplines, but it is 'necessarily ubiquitous and universal.  It is the precondition of knowledge'.  [Seems to be a commonsense refutation of the new soc of ed?]

Epistemic matters are often seen as tacit , and they may not be highlighted even  in university disciplines such as mathematics.  In other fields, they can produce open conflict.  The issue then becomes one of trying to decide why conflicts like this occur.  The new sociology of education has been largely silent on this, because of its knower mode - only in knowledge modes can claims become detached from their authors and thus subject to debate.  A return to an knower mode threatens this autonomy, and focuses on voice rather than the epistemic device itself.  The arbitrary is everywhere.  Thus the intrinsic features of knowledge cannot be analysed by the new sociology of education - 'weak grammars cannot see the epistemic device'(178).  However, it does have an important role, and it should be the fundamental issue for the discipline.  Only Bernstein has seen this

The next stage will be to see both epistemic and pedagogic devices as part of 'an overarching knowledge device'

back to ed studs