Knowledge, Power and Educational Reform

Applying the Sociology of Basil Bernstein

Rob Moore, Madeleine Arnot, John Beck, Harry Daniels, Rob Moore, Madeleine Arnot, John Beck, Harry Daniels

• Publisher: Routledge
• Year: 2006
• ISBN:
  9781134181827
• Edition: 1
• Topic: Education
• Subtopic: Education General

This book is made up of a selection of writings from an international team of scholars, highlighting the contribution made to the field of educational policy and educational policy research by Basil Bernstein's work on the sociology of pedagogy. These contributors explore, analyse and engage with contemporary political reforms of education, contemporary pedagogic debates and the changing nature of professional knowledge, relationships and structures. The subjects covered include:

• particular concepts such as voice research
• the significance of social class in relation to the language, schooling and home cultures
• differences between official and pedagogic recontextualising fields
• formation of different types of identities
• the construction of the learner
• formation of teacher identities and use of pedagogic discourses
• analysis of performance-based educational reforms and its impact on pedagogy.

Part I

Knowledge structures and epistemology

1 On the shoulders of giants

Verticality of knowledge and the school curriculum

Johan Muller

This paper will explore the question of knowledge specialization or ‘depth’. It will proceed by addressing the question: how does knowledge grow? The paper asks further: what does this mean for what, and how, children learn at school?

The first issue to be clarified concerns whether all knowledge ensembles are structurally alike, and whether they all grow in the same fashion. A dominant strain in the social theory literature holds that there is a principled distinction to be drawn, and various thinkers have at different times advanced a map of the territory. As shown in Table 1.1, Paul Dowling (1999) has usefully summarized some principal efforts that have, across the past two centuries, tried to delineate the continuum of knowledge organization and growth.

What is most striking about Table 1.1 is the lack of specificity of the distinctions drawn. Most social thinkers, if not quite all, have delineated the universe of symbolic differentiation by naming linearly opposed poles. Questions remain about the logic of the distinction being made. Are there only two knowledge categories, with all ensembles forming sub-types of one or other pole, which would make the distinction a strictly dichotomous one? Or do the poles name a continuum, with all ensembles differing in terms of a single, or perhaps multiple, classificatory dimensions?

Table 1.1 Distinguishing knowledge continua

Author	Abstract Context-independent	Concrete Context-dependent
Bourdieu	Theoretical logic	Practical logic
Foucault	Programmes	Technologies
Freud	Ego	Id
Levi-Strauss	Science	Bricolage
Levy-Bruhl	Modern thinking	Primitive thinking
Lotman	Rule-governed	Exemplary texts
Luria	Abstract thinking	Situational thinking
Piaget	Science/effective thought	Technique/sensori-motor
Sohn-Rethel	Intellectual	Manual
Vygotsky	Conceptual thinking	Complex thinking
Walkerdine	Formal reasoning	Practical reasoning

There are at least two kinds of problem with these dichotomies. The first forms the point of departure for Basil Bernstein’s (2000) magisterial intervention in the discussion: ‘To my mind much of the work generating these oppositions, homogenizes these discursive forms so that they take on stereotypical forms where their differences or similarities are emphasized.’ In other words, as Moore and Muller (2002) have observed, much of the discussion, instead of elaborating the dichotomous distinction, perpetually collapses the terms of discussion back into the crude simples, which simultaneously distorts the discussion and keeps it mired in ideology. Bernstein’s analysis explicitly sets out to break this vicious circle as we will see below. But a second tendency has also been responsible for retarding development of the issues in the past, which is a pervasive one in nearly all these discussions, namely, to treat the terms of the distinction as delineating differences of activity or practice, rather than differences of symbolic organization, casting the discussion in knower rather than knowledge terms (Moore and Maton, 2001). This reluctance to speak directly about symbolic systems is an old one, and below I will trace its origins to the terms of a debate that began in the eighteenth century at the advent of the age of science (or was it perhaps of much older provenance?), continuing to the present. This debate is about the idea of progress in general, and the idea of progress in knowledge in particular. We are, it would seem, exceedingly reluctant to speak about the social dimensions of knowledge variation, not only in terms of relations between different knowledge forms, but particularly in terms of relations within knowledge forms. The idea of progress haunts us, nowhere more so than in regard to the question of knowledge progression and growth.

Bernstein has, as I have said, intervened decisively in the discussion about the form of symbolic systems, setting out to delineate the ‘internal principles of their construction and their social base’ (Bernstein, 2000: 155). As is by now well known, he distinguishes between two forms of discourse: horizontal and vertical. Since, in his view, only vertical discourse can be said to contain knowledge structures, strictly speaking, this paper will not discuss further the question of discourses and will concentrate on the question of variation between knowledge structures within vertical discourse. Here Bernstein distinguishes between two kinds of knowledge structure: hierarchical and horizontal. It is in this latter distinction that he stakes his claim to go beyond the homogenizing form of account mentioned above (ibid. 161).

For Bernstein, knowledge structures differ in two ways. The first is in terms of what may be called verticality. Verticality is concerned with how theory develops. In hierarchical knowledge structures it develops through integration towards ever more integrative or general propositions, the trajectory of development of which lends hierarchical knowledge structures a unitary convergent shape. Horizontal knowledge structures, on the other hand, are not unitary but plural, consisting of a series of parallel incommensurable languages. Progress in horizontal knowledge structures occurs not through theory integration (or at least not primarily) but rather through the introduction of a new language which constructs a ‘fresh perspective, a new set of questions, a new set of connections, and an apparently new problematic, and most importantly a new set of speakers’ (ibid. 162). Because these languages are incommensurable, they defy incorporation. The level of integration, and the possibility for knowledge progress in the sense of greater generality and hence explanatory reach, is thus strictly pegged.

Before I proceed to discuss the second form of knowledge form variation, it is worth making a few observations on verticality. The first observation is that it artfully incorporates and recapitulates the fierce dispute in the philosophy and sociology of science between the logical positivists and the non-realists, a dispute I shall selectively re-visit below. Bernstein is implicitly asserting that the logical positivists (or realists) were right, but only in respect of hierarchical knowledge structures; the non-realists (Kuhn and after) were likewise right, but only in respect of horizontal knowledge structures. In other words, encoded into Bernstein’s principle of verticality are the terms of debate in the philosophy of science since the romantic revolt of the eighteenth century. Second, though, we should note that the category of horizontal knowledge structures spans a very broad range, from mathematics to sociology and the humanities. Although there is more than one mathematical language, and mathematics is in this sense a ‘horizontal’ knowledge structure, this example makes clear that verticality is certainly possible within discrete languages, verticality of an order perhaps equivalent to that obtained in hierarchical knowledge structures. The germane question then becomes not so much what constrains progression in all horizontal knowledge structures, but rather what internal characteristics constrain it in those that proliferate languages rather than in those where language proliferation is constrained. It was in search of a sociological answer to this question that Bernstein initially began, setting out to provide an alternative to Bourdieu’s sociological reductionism (see Bernstein, 1996).

This brings me to the second form of knowledge variation: grammaticality. If verticality has to do with how theory develops internally, with what Bernstein later called the internal language of description, grammaticality has to do with how theory deals with the world, or how theoretical statements deal with their empirical predicates, the external language of description (Bernstein, 2000). The stronger the grammaticality of a language, the more stably it is able to generate empirical correlates and the more unambiguous because more restricted the field of referents. The weaker it is, the weaker is its capacity to stably identify empirical correlates and the more ambiguous because much broader is the field of referents, thus depriving such weak grammar knowledge structures of a principal means of generating progress, namely empirical disconfirmation: ‘Weak powers of empirical descriptions removes a crucial resource for either development or rejection of a particular language and so contribute to its stability as a frozen form’ (Bernstein, 2000, 167–168). In other words, grammaticality determines the capacity of a theory or language to progress through worldly corroboration; verticality determines the capacity of a theory or language to progress integratively through explanatory sophistication. Together, we may say that these two criteria determine the capacity of a particular knowledge structure to progress.

But for all its richness, this analysis merely starts the ball rolling, so to speak. What it provides is a survey of features of variation, but even the charitable must admit that the poles remain clearer than the intermediate zones of the range. This is partly because the precise nature of and relation between verticality and grammaticality is unclear. A plausible surmise could be the following: that verticality is a categorical principle, consigning knowledge structures to either a theory-integrating or a theory-proliferating category (in turn broken down into a constrained proliferation or an unconstrained proliferation category), while grammaticality is an ordinal principle, constructing a continuum of grammaticality within each category, or perhaps across the entire spectrum.¹ But this does not help us understand the more striking differences between knowledge structures, for example temporal extension, as displayed by Moore and Maton (2001) in the story of the proof of Fermat’s last theorem:

What is so striking about this story is its sheer scale in historical time and in geographical and cultural space. It tells a story of a mathematician in late-twentieth century England effectively communicating with a French judge at the court of Louis XIV, and through him with Babylonians from three millennia ago. It represents an epistemic community with an extended existence in time and space, a community where the past is present, one in which, when living members die, will be in turn the living concern of future members …

(Ibid. 172)

As I will show in the next section, this sense of extension in time and space is the hallmark of the modern idea of knowledge progression. The point here though is that things could not look more different in sociology, which barely dates back as far as the nineteenth century. On the other hand, mathematics shares this feature with literature. Gyorgy Markus (2003) has remarked that the ‘tradition’ in the arts is ‘ever expanding’ and ‘of great depth in time’ (p. 15), a feature he contrasts to science which has a ‘short-term’ tradition, because it is ever ‘evolving’ (that is, progressing) (ibid.). Which knowledge form is nearer to which? The fact is, which forms comprise the middle of the knowledge range is not clear at all. Is geography closer to physics than to biology, for example, and how would we know? Would we count their respective numbers of languages? Could we create an index of theory integratedness? It is certainly the case that empirical study would help to shed light on the theory, but it is also the case that the theory stands in need of some elaboration. This paper is a ground-clearing digression preparatory to that task. In what follows, I will concentrate mainly, though not exclusively, on verticality – on the internal characteristics of the internal language of description.

Why would one want to elaborate a theory of knowledge forms? After all, we seem to have coped reasonably well without one for a long time. Bernstein only turned to the issue towards the end of his work. The contention here is that this lacuna in the study of knowledge and education was not accidental. Rather, I hope to show, it was produced by the failure of social thought to deal with the dilemma of progress. The failure to reckon with the material structural differences in knowledge forms has become something of an obstacle in educational thinking. This can briefly be illustrated in two domains of education practice, namely, curriculum planning and research administration.

Curriculum planning has been thrust into the limelight by global learner performance comparisons, most vividly displayed by the Trends in International Mathematics and Science Study (TIMSS). A central tenet of assessment is that the instrument measuring performance is valid to the degree that it assesses what has been made available for acquisition. TIMSS has made visible the fact that not all children of the same age cohort across the globe learn the same things in the same order at the same level of cognitive demand. This has put a spotlight on the stipulation, sequence and progression requirements of curricula and has begun to suggest that not all subjects in the curriculum have the same requirements. Could this be because their parent knowledge forms are different? To answer this seriously requires greater clarity than we have at present.

As for the question of research administration: research assessments of individuals and bodies of work have made possible comparisons between individuals, faculties, universities and countries. As more and more comes to depend on assessments of innovation and novelty (‘Is this paper really a contribution to new knowledge, or a re-hash of the known?’), the question arises as to what exactly constitutes innovation in different areas of research endeavour, and whether they are all the same. This is only the tip of the iceberg: it soon becomes clear that there are different epistemic cultures, different kinds of collaboration, different publishing traditions, and so on. In short, the globally emergent audit culture compels us to reflect on our knowledge practices, at the centre of which sits the question of their likeness, their comparability, and their compatibility. Once again, we realize how little we really know about how they may be alike or different, and what difference this might make. At the centre of this conundrum lies the question of knowledge progression.

Progress: the very idea, and its sceptics

The foundation of the Cartesian revolution in the seventeenth century was an axiom that appeared to be radically new, namely, that ‘true’ knowledge was characterized by knowledge progression, that that ‘which had once been established did not need to be proved again, that is to say, in which scientific progress, universally recognized as such by rational thinkers, was possible’ (Berlin, 2000a: 28). The conventional account depicts this as the decisive moment in the emergence from the closed tautological world of antiquity, and the birth of the modern (Shapin, 1996).

There are a number of entailments to this view. First, Descartes believed that only in a bona fide branch of knowledge can we find ‘clear and distinct ideas’ (Berlin, 2000a: 28):

The paradigm of true knowledge, according to the Cartesian school, consisted in beginning from truths so clear and dis...

Table of contents

• Cover
• Halftitle
• Title
• Copyright
• Contents
• List of illustrations
• List of contributors
• Acknowledgements
• Introduction
• Part I: Knowledge structures and epistemology
• Part II: Knowledge, identity and voice
• Part III: Professional knowledge and pedagogic change
• Part IV: Policy innovation, discourse and educational reform
• Bibliography
• Index
• Backcover