Notes on: Maton K and Muller J ( 2006) 'A sociology for the transmission of knowledges, http://www.KarlMaton. com [NB a version appeared in an edited collection, but Maton says that the published version was changed.  I actually had downloaded it from Academia.ed]

Dave Harris

Bernstein went on to develop an interest in the structure of knowledge as a move away from analyzing pedagogies, and this shift to discourse meant a reconnection with 'systemic functional linguistics'. Early Bernstein work on codes led to a more 'formal analytical concept' of code as'" a regulative principle, tacitly acquired, which selects and integrates meanings, forms of realizations, and evoking contexts"'(2).  This replaces the usual interpretation of the early work on elaborated and restricted code.  Bernstein was always reworking his ideas, aiming to make new objects 'visible' for subsequent research, and then looking beneath the empirical features to uncover structuring principles, including codes, then even further discussion of what generates the principles.  Ultimately, there is an interest in social order and symbolic control, and its links from macro structures of society to individual consciousness.  One path can be traced through this work.

Any linguistic or symbolic ensemble can be seen as linked to its social bases, as 'a material social form of life' which produces various kinds of ordered and disordered language.  Following Durkheim, it was the economic division of labour and the link to forms of social solidarity that produced different symbolic forms.  In class based society, symbolic resources are differently valued regulated and distributed, and sociologists need to understand how this occurs, and through what mechanisms, as societies reproduce and change.  This explains pedagogic relations and the connection to changes in the knowledge bases of society.

Initially, the key concept here was code.  Originally, codes differed in terms of their 'more or less complex lexical, semantic and grammatical features', their linguistic repertoires (5), but this was developed to refer to an orientation to meaning, with codes leading to different modes of communication.  These would be valued differently by schools, partly because they differed in terms of their effectiveness in relating to school values and practices [not analyzed as related to domination as in Bourdieu?].  Thus a restricted coded orientation limits people to context-specific meanings because it has a direct relation to the social base; elaborated code orientations offer a more mediated relation to the social base which 'predisposes that person to universalistic, non-local, context independent meanings', which implies a more specialized context [latter requires security and cultural capital?] .  These qualities become important in modern societies with a specialised division of labour.  A principal transmission takes place through homes and schools, 'but not all homes and not all schools for the same degree', hence class differences. This provided an early connection with systemic functional linguists, and helped move away from the idea that code is not the same as dialect, nor the same simply as middle class speech [although this was how the work was often received].  Bernstein acknowledged this link to linguistics, especially through the work of Halliday.

Further developments followed a more detailed analysis of the elaborated code in the school.  Codes were not just possessed by individuals, but became more general principles or rules.  This connected with the idea of classification and framing, where classification involves relations of power that regulate the connections between contexts and categories, while frames refer to relations within contexts or categories.  This is a distinction between power and control, and it helps flesh out the idea of the effects of orientations to meaning [for example, there can be an invisible pedagogy in the elaborated code, where weak classification and framing leads to the concept of the person rather than the individual specialist].

This has produced observational instruments leading to empirical studies [some listed on page 7], and provides a detailed analysis of the effects of various pedagogic modalities, and can explain why particular groups operate in particular classrooms and schools [as above, where invisible pedagogy operates only with a fraction of the middle class and is effective only in correspondingly 'progressive' schools].  The work has led to a difficult dilemma however—we can change the structuring principles of the school to match the codes already possessed by pupils, or develop ways of providing pupils with the key to the new code.  The problem is that if we do the first, we risk producing a subordinate and lower status form of educational knowledge specifically for working class kids [yes -- but why?] while if we do the second, we imply that there are two forms of knowledge and that the working class kids lack the first one which would be 'beyond the pale in much contemporary social science'(8).

The idea of such codes can be applied to any symbolic group, including work relations, although schools remained as the major focus.  The struggle over invisible pedagogies can be seen as a conflict between different bits of the middle class, but what about the general issues that shapes social structure?  Here we need to consider knowledge production and how pedagogy recontextualizes it - the work on the pedagogic device [nicely summarized in a convenient table, below, page 9, and discussed more fully here]. 


Each set of rules is associated with a field of activity, often associated with specific sites, together constituting an arena.  Again, Durkheim appears on the major split between mental and manual labour, and its 'corresponding symbolic cleavage between sacred and profane symbolic orders' (10).  Again we have an argument about how knowledge is produced and transmitted, and what the consequences are for different social groups, but there is also an implication for the evolution of society, social reproduction and change: conceived as '"an arena of struggle"' over the pedagogic device (11). 

[So there does seem to be some underlying class struggle here despite what Bourdieu says?].  Those with power can set the device to favour their own code modality [the actual wording here is slightly different, though, that they can set the device 'to make sure that the dominant higher status code modality favours their own'], which will put less dominant groups at a disadvantage.  The question therefore becomes the classic one of asking who rules the pedagogic device [still not quite whose code?] and who can impose a hierarchy on code modalities.  In this way, the pedagogic device runs from macro structures of power to individual consciousness, an 'attempt at beginning a grand unified theory'(12), something at a higher level than codes - codes are 'the effects of the device' (13).

[The higher level of abstraction is celebrated in terms of a formula, 13, which I have not reproduced - basically, orientations of meaning are embedded in classifications and frames, in stronger or weaker ways.  Again empirical research becomes possible [one of them includes Singh]

The pedagogic device itself was to lead to a broader analysis still of the issue of knowledge itself.  For a device to be necessary means we need to further analyze pedagogic discourse and the forms taken by knowledge.  The term knowledge is often treated without being analyzed.  Educational analysis tends to represent 'an "over-ideologized" image of knowledge and pedagogy'(14', seeing the only role of knowledge as to produce inequalities, so that all that we can understand knowledge only in terms of the perspectives of the knower or the ruling ideas.  It is true that relations of power do affect the production and reproduction of knowledge, but the 'internal ordering of symbolic forms' is also important [the point that Foucault was to discover right at the end of Archaeology of Knowledge].  Otherwise, education becomes merely a tool of ruling ideas [see also the reservations on this by Bowles and Gintis in their follow up, and the subsequent work by MFD Young].

We need to look at pedagogic discourse and its own specific effects, the relations within it, its intrinsic features, those things that make its specific in education.  The concept of code takes this for granted, and the pedagogic device only showed how discourses were put together, without focusing on the forms of that discourse.  The field of the symbolic production of knowledge had been relatively neglected.  Bernstein even thought that focusing on pedagogy was not sufficient to examine current '"discursive culture"'(16).  We need to examine which knowledges are being produced and distributed, and how they might shape consciousnesses.

Again, Durkheim on the symbolic was useful.  It related both to the social base and to specialized activities of consciousness.  Elaborated codes did not just offer more combinations of terms.  Schools subjects do not just reflect immediately particular forms of knowledge.  Bernstein moves on to look at discourses and knowledge structures [SIC].

The work begins with distinguishing horizontal and vertical discourse. The former referred to everyday knowledge, working with a segmented structure which is context-specific and dependent, joined together in functional ways related to everyday life.  Vertical discourse was explicit and systematic, and so more coherent and hierarchical [Pask's knowledge structures!], featuring recontextualization between general and specific levels.  This difference improves on general 'conflationary' definitions of knowledge (17), found in psychology and in common sense.  However, actors can move between the two in practice.  The real value lies in understanding social structure and the difference between specialized symbolic positions in a division of labour.  It is not just a matter of differences between concrete and abstract thought, or local and official knowledges.  [There is a strong implication that vertically organized knowledge structures are more valuable and better rewarded, and that these are also differently distributed?].

Vertical knowledge structures are seen best in the natural sciences [the first of a series of idealizations, sometimes explicitly based on Popper], which occupy a triangular form, with the most powerful integrating propositions at the top.  Horizontal knowledge structures offer a series of specialism languages and truth tests, as in the humanities and social sciences - the perspectives in sociology would be an example.  Hierarchical knowledge moves by a principle of internal development, extending the base and sharpening the tip of the triangle, explaining more empirical phenomena via integration in increasingly economical ways.  Horizontal knowledge continues to accumulate new languages.  [Shades of Piaget now?].

Horizontal knowledge structures face challenges from intellectual fashion, although they also feature similar focuses and some repetition.  It is difficult to build on knowledge.  They lead to 'epistemologically different understandings of the world' (20) and also offer [similarly diverse] practices, pedagogies and specialisms, while vertical knowledge structures operate within the same agreed language [we are going to get to Kuhn before long!].  There is knowledge the operates in World 3 [explicit Popper here], objective knowledge not confined to subjective states, and this has properties of its own.  In particular, it cannot be reduced to ideology.  Knowledge can still be hypostasised [and put to the purposes of the powerful], but it is a visible object with its own properties and powers: those emerge from social practices but are not reducible to them.  [There is a bit of a weasel about whether the abstract qualities of such knowledge are simply heuristic, helping to make knowledge distinct from social practices for analytic purposes].

This latest development also raises a host of questions, especially:

(1) Whether these distinctions are too clear cut, especially when attempting to describe actual intellectual disciplines.  The characteristics of horizontal knowledge in particular look as if it is subject to constant change with no continuity, but if this so there would be no intellectual field.  Accumulation of new languages does not always cause a crisis in the older ones [we definitely need Kuhn here].  However the issue is whether greater abstraction and generalisability arises rather than a return to themes that have been already addressed but with new assumptions and languages.  The different perspectives in social sciences are the result of different languages, even if they use the same terms such as social class or patriarchy, and this will limit 'development'[23, clear value judgement here].  The issue also turns on whether social sciences can progress as natural sciences do, and here, two particular characteristics of knowledge structures become important, their '"verticality" and "grammaticality"'(23).  The first one relates to developments through integration into more general propositions: there is some work in linguistics apparently turning on the technical qualities of language 'especially grammatical metaphor', specifically relating to mathematics enabling progress in science.
(2) The importance of grammar or conceptual syntax [seems to be the same as the above?], and whether they can develop precise empirical descriptions or formal models of relations.  A 'stronger' grammar can relate alternative theories [a tautology?].  More than one theories do exist inside a vertical knowledge structure, and there may be conflict, but theory choice is different -empirical research can resolve choice if the grammar is strong enough [not at all what Kuhn argues!].  New theories are both in conflict with and commensurate with existing ones, and a successful theory must explain everything that its predecessors did, as Popper argued [and Kuhn denies - Newton's theory of gravity, for example did not explain the precise nature of the force better than its predecessors in one of his examples, and nor did oxygen theories explain more results than phlogiston ones, I recall].  Horizontal knowledge structures have no such agreed empirical test and can only work with critique, a different form of conflict and conflict resolution 'in which the strength of grammar plays a role' (24-25).  Bernstein goes on to argue that economics, linguistics, logic and maths are horizontal, but have relatively stronger grammars and so can make intellectual progress.

Theory choices based on notions of 'consistency and compatibility with other theories (their internal strengths of grammar) and their capacity to explain the results of empirical research (their external strengths of grammar)'(25). This produces a choice based on 'integrative, subsumptive and explanatory power', or what Maton calls a 'knowledge code', based on the relations with knowledge and objects of study.  Any other changes are deemed to be 'ideological rather than rational', based  'more on a "knower code"', based on the status of who is making the claims.

These distinctions could be clearer [they badly need modification according to recent work on science!].  They still remain at the metaphorical or suggestive [highly idealized] stage, without underlying structures.  Sociological relations are also important, and 'a discipline is more than just its structuring of knowledge'.

What is the relation between these knowledge structures and educational knowledge?  Knowledge structures have to be recontextualized, if they are not to be just another version of pedagogic codes.  The framework might be developed to consider such relations in the future, and here notions of knowledge changes might be useful - for example, the subject of English might be structured 'in ways that often debilitate the integration of already learned knowledge', and express an invisible pedagogy rather than making knowledge structures visible (27).  The relations between knowledge in general and curriculum structures also leave a 'space for the play of ideology'  in any recontextualization, but recontextualization itself might be limited by the structure of knowledge as a material basis: Bernstein already argues that evaluative rules are artefacts, but still must bear some relation to parent knowledge.  This again might be explored in the future, and there is some useful linguistic work to assist. There is a final problem of making Bernstein's emphases consistent, combining integrating and collection codes to educational knowledge codes.  Maton (2007)  is on the case.

[The conclusion offers a neat summary of the article, emphasizing the broadening focus of Bernstein's work, tracing school knowledge 'upstream towards its epistemic sources'(30), but insisting on differential distribution and acquisition, a key part of social reproduction.  There is finally an interest in epistemic communities which persist over time, as in mathematics, as offering a 'pure form of communism', (31) apparently as Merton argued.  Here, a knowledge structure seems to have generated a knower structure or scholarly community.  Bernstein wanted to develop a similar one for sociology, attempting to move beyond just accumulation to get to the principles that would enable the building of a vertical knowledge structure persisting across time, and connecting sociology and linguistics].

back to ed studs