NCM Brown November 19961 THE REPRESENTING RELATION BETWEEN SUBJECT CONTENT AND OUTCOME STATEMENTS IN THE NATIONAL CURRICULUM NEIL CM BROWN Neil CM Brown Associate Professor Head, School of Art Education College of Fine Arts The University of New South Wales PO Box 259 Paddington NSW 2021 AUSTRALIA Email: Neil.Brown@unsw.edu.au Fax: 61 2 385 0706 THE REPRESENTING RELATION BETWEEN SUBJECT CONTENT AND OUTCOME STATEMENTS IN THE NATIONAL CURRICULUM Abstract This paper examines the various ways in which subject matter has been represented as outcome statements, profiles and other terms of competency in state curricula. The paper describes the different ways in which five alternative pedagogies transport properties between subject domains and outcomes. It takes the view that the meaning of outcomes and standards in curricula is opaque within the function of the pedagogies which generate them. It is argued that the asymmetry of outcomes as different representational kinds runs counter to their popular educational purpose which is to establish a single transparent identity between subject content and a guaranteed level of student achievement. The paper first categorises outcomes as to kind. A realist concept map is then used to meta-represent the ways in which five different pedagogies construct the answers to a hypothetical question about an outcome example in the visual arts. THE REPRESENTING RELATION BETWEEN SUBJECT CONTENT AND OUTCOME STATEMENTS IN THE NATIONAL CURRICULUM I may have said the same thing before...but my explanation, I am sure, will always be different. Oscar Wilde. In Conversation Schwab claimed that one of the signs by which subjects in crisis within curriculum can be identified was evidence of: A flight of the field itself, a translocation of its problems and the solving of them from the nominal practitioners of the field to other men. (1969, p.3-4) By insuring against the judgement of teachers through the stipulation of national standards, outcomes, profiles and performance indicators , educational authorities in many western countries may be forgiven for appearing as if they sought, in Kemmis' words, to `define the process of development and learning as if there were no fundamental debate about these processes', and `to define the knowledges needed to live in the modern world as if these were not also the subject of debate.'(p.55) The concept of a national curriculum is not new. National curricula are familiar in authoritarian states where the reason for their adoption has been twofold.(Meighan 1995, p.27) The interest in national approaches to standards and measurable outcomes is consistent with the view of curriculum as a means for binding former nation states together and for the creation of a sense of national identity. Secondly a nationally consistent program for raising educational standards is perceived as a way of strengthening economic power. Few question the success of curricula in determining these and other cultural agenda. Professor Gough in his article `The Faustian Bargain...' reminds us of the tragic success of the French Ministry of Public Instruction which sought to make the visual arts conform to officially sanctioned categories in the 19th century.(1995) Its `success' led to an artistic revolution against the French academy at the Fin de Siecle. Standards and outcomes reporting is part of a world wide movement towards accountability in assessment which is recognised by its competency based criteria, measurement of fit in relation to work samples, and for what might be referred to as the emphasis on the correct conduct of knowledge. The latter is illustrated by the National Curriculum Programmes of Study of Art in Great Britain in which all bar a few of its `End of Stage Statements' are marked by an increasingly narrow range of performances students `should' or `could' enact to show their level of achievement in a particular subject.(Mason, p.85) The National Curriculum agenda of Government in Australia (1993), Great Britain (1992), and to a lesser extent the National Standards program in the USA (1994), is thought by some, in Kemmis' terms, to foreclose on debate, deriving curriculum and syllabus structures as if from a single rationale of society. (Hausman 1994) `National standards', `outcomes' and their concomitant `profiles' are rapidly firming into a representative of that single rationale of curriculum that Kemmis refers to as being `treated as agreed'.(p.55) Converting subject content into curriculum outcomes, according to the New South Wales Board of Studies, involves the identification of six `stages' of learning. These stages are represented in the form of `statements [that] express the specific intended results of the teaching of the syllabus'.(1996, p.6) Outcomes within the six stages of learning must represent content in a way that `reflects a balance between the acquisition of knowledge and the processes of learning so that students are encouraged to be effective learners'.(p.6) The Eltis Report, an inquiry into `outcomes' reporting commissioned by the New South Wales Ministry of Education, devotes a lengthy passage to an analysis of the different ways in which respondents to the inquiry identified outcomes.(1995, p.33) The Eltis Report concludes that, irrespective of the New South Wales Board of Studies (1991, 1996) official definitions, widespread variation in the interpretation of outcomes by teachers responding to the inquiry was likely to have `implications' for the implementation of outcomes reporting in schools. On the one hand, Eltis remarks, many teachers reported little trouble with outcomes and saw them as little more than a map of probable student learning. On the other hand many expressed concern that if outcomes and standards were to be interpreted as `a checklist of what students should achieve...[then the danger is that]... teachers will teach to the outcomes in order to ensure that students will achieve them'.(p.33) The report concludes that the self fulfilling tendency of predetermined forms of assessment in syllabi constitutes a serious problem for teaching and learning in the curriculum. The deeper question, to which the reported diversity of interpretation of `outcomes' in the Eltis report alludes, but on which it does not elaborate, seeks to find out what kind of entity an outcome or a standard is. Debate over the proper ways to define outcomes and standards has a bearing on the concerns that arise over the way in which subject content is represented as achievement in the curriculum. Categorising Standards and Outcomes Elliot Eisner points out, with some irony that the term `standards' has multiple meanings. Standards, Eisner says, can refer to `icons of student performance', `something that is common or typical', `a rite of passage' and so on.(1993, p.22) He draws a distinction between standards and criteria. Criteria are different to standards, he argues, because criteria make reference to the qualities by which a property can be judged. For example, knowing the nature of "heat" as experience requires the knower to make reference to criteria of judgement. Knowing "heat" as temperature, on the other hand requires consultation with a standard of knowledge such as a thermometer (p.22). The fault in national standards, he suggests, lies with the educational policy that privileges standards of measurement over criteria of judgement in the assessment of curriculum outcomes. There are two important semantic issues addressed by Eisner in these remarks. First, there is little doubt that "standards" and "outcomes" do have multiple meanings as he says. The multiple meanings he refers to represent a range of alternative cultural functions that standards and outcomes are employed to serve. Second, when Eisner distinguishes between heat "as temperature" and heat "as experience" he is making use of two qualitatively different referential systems which arise from differences in the kind of thing being referred to. Heat `as experience' refers to descriptions of heat judged as a value relative to the phenomenal experience of the describing subject; while heat `as temperature' refers to heat relative to the object. In these examples Eisner is variously referring to standards of `heat' as nominal, natural, and non-strict natural kinds. Standards and outcomes as natural kinds It is unlikely that `outcomes' or `standards' would ever qualify as strict natural kinds even if they were possessed by students as objective knowledge about `heat'. Both Putnam (1975) and Schwartz (1978) agree that natural kind terms are applied to things indexically. Schwartz says: In other words the extension of the term is not determined by descriptions semantically associated with the term but by an ostensive ceremony in which one points out some stuff or things that one takes to be paradigmatic and says, `everything of the same kind as this is in the extension'. ...In general, any descriptions associated with an indexical term serve merely to fix the reference. Once a term has been indexically introduced what counts as of the `same kind' as the paradigm will be a matter of empirical research and scientific theorising. (1978, p.183) Esner's reference to heat as `temperature' is an example of an indexical ascription to `heat' as a natural kind. The meaning of heat as a natural kind is to be found in its essence, and, as a reference, to heat `temperature' carries no meaning outside of its paradigmatic relation to heat itself. What we learn contingently about heat through scientific investigation adds to the theory of temperature. Natural kind terms are constructed by their references. Thus natural kind descriptions do not contribute to the structure of the things to which they refer. While references to `heat' as temperature may qualify as indexical, references to `heat' as curriculum outcomes and standards cannot. We would learn little about the properties of curriculum outcomes and standards through their scientific investigation, since an explanation of a curriculum outcome about heat as a natural kind, is not coextensive with causal explanations about `heat' as such.(Schwartz 1978, p.572) Outcomes are implicated in causes relating to representations of standards about `heat' as teaching and learning, not temperature. In other words the study of any particular example of an outcome would imply nothing about outcomes and standards in general. Standards and outcomes as nominal kinds That Eisner is able to accept that `standards' in curriculum have multiple meanings in the first instance, however, is conditional upon them being conceived to exist as nominal kinds. Norris explains nominal kind terms as follows: If [outcome/standard] is a nominal kind term, then theorists might differ over its meaning according to their individual programmatic agendas. Under this scheme of things, the properties associated with [outcome/standard] would be associated semantically with it. Changing the list of properties would amount to changing the concept related to the term, and thus the class of individuals that fall into its extension; and deciding on the list of properties would be a matter of how we wanted language to be used. This decision would be influenced primarily by language user's values. Depending on their agendas, individuals would disagree over [what] should be classified as [outcomes/standards]. (Norris 1992, p.9) According to the Board an outcome must represent subject content as a `knowledge, skill or attitude'.(p.6) Decisions about what knowledges and skills ought to be included within educational outcomes and standards are unlikely to be resolved by reference to an essence of outcomes, but by external values and conventions imported into the reference. Even within the one content domain there would be competition among curriculum writers as to what should and shouldn't be included as an outcome. Different conventions in constructing the conceptual intension of outcomes and standards introduces an asymmetrical relation between outcomes/standards and the properties they possess. Any inquiry into outcomes would therefore have to take the form of a philosophical inquiry into the extension of the concept of curriculum outcomes, since it wouldn't make sense to inquire into the `nature' of outcomes themselves. As nominal kinds, outcomes and standards qualify as artifacts. As artifacts the emphasis is placed upon establishing their meaning rather than their truth. It follows from their artifactuality that no inferences can be drawn about the extension of `outcomes' from the examination of one example.(Schwartz 1978, p.573) They are semantic entities, determined a priori, for which examples carry no implications for changes to their extension. Standards and outcomes as non-strict natural kinds Eisner's representation of `heat' `as experience', however, hints at an epistemological condition that represents heat as a kind of mental outcome or standard. The New South Wales Board of Studies stipulates that an educational outcome must: reflect(s) a balance between the acquisition of knowledge and the processes of learning so that students are encouraged to be effective learners.(1996, p.6) By the introduction of `experience' and a `process of learning' Eisner and the Board in New South Wales respectively, identify `standards and outcomes' not only as semantic objects but also as a set of traits and dispositions existing in the learner. According to the Board it would be insufficient to represent a curriculum outcome about the `heat' of a colour, for example, as a semantic relation alone. Knowledge of warm colours, represented as an outcome, must also be represented as a `learned' kind of achievement. Eisner's suggestion that knowledge of `heat' can be expressed through judgements originating in experience, provides an example of a way that the `learned' condition could be satisfied in the form of judgements about the `warmth' of colours on the palette. From an aesthetic point of view simply `reading' the relation between heat and colour would be considered an invalid outcome, akin to reciting the answer by heart. Unless it could be demonstrated that the `heat' of the colours was able to be judged as an experience, the conditions of `learned' achievement might not be satisfied.1 The introduction of `experience' into an outcome about the heat of a colour does not entail an extension to the concept of heat. It is an empirical issue. It requires a demonstration that a randomly placed colour depends upon the judgements of a felt experience for the reliable belief that it is warm.(Kornblith 1994, p.135) Is the `learning' condition required by the Board within each subject, therefore, satisfied as a nominal extension of a concept of outcomes enacted by the student, or is it, in part, the inculcation of an underlying physical state whose presence in students `as experience', for example, is to be assured by the teacher as a condition of learned achievement? According to Norris it might be both; that outcomes and standards meeting the conditions sketched above are representatives of non-strict natural kinds: Analytic definitions and the mis-representation of outcomes Natural kind terms are: generalisations [that] are metaphysically necessary if true and yet epistemically contingent... `water is H20' [is] not subject to falsification by counter example, even though it is corrigible. (Schwartz 1978, p.184) If we came upon water that was not H20 but resembled water in all its superficial properties we would say it was not water. When we talk of paediatricians as doctors, on the other hand, `doctors' forms an analytic extension to a nominal kind. Being a doctor is not an underlying trait of being a paediatrician and so `paediatrician' cannot be a natural kind. Rather the relation between doctors and pediatricians is an analytic truth established by semantic stipulation. Analytic relations cannot be wrong. Paediatricians are doctors does not refer to a natural, independently existing entity that exists as a doctor by virtue of possessing a number of underlying traits. `Paediatricians are doctors' is a semantic stipulation in the same way that `a square is a figure with four equal sides joined together at right angles' exists as a sematic relation. Both exist in the form of semantic stipulations. On the face of it, stipulating definitive relations between outcomes and subject content in curriculum might seem like a good idea, since it appears to eliminate the ambiguity in teachers' and parents' interpretations of the meaning of the relation. `Transparency' in outcomes reporting is considered to be an essential property of their function as performance indicators. For example, it is consistent with current thinking on outcomes to provide art teachers with detailed work samples which exemplify the standards appropriate to each stage.(Curriculum Corporation, 1993) On reflection, however, defining the relation in this analytic way may prove to be of little advantage in the long term. Along with transparency comes a corresponding redundancy in the information. Analytic definitions deny the representing relation its role in opening up a meaningful difference between the artifacts referred to (subject content) and what they refer to them as (achievements). The uniqueness of work samples as criteria form resemblances with the achievements to which they refer. Like tautologies work samples beg the intuition of meaning which motivates the need for a representation of student achievement in the first place. Even though artworks are traditionally ranked ostensively, the ranked group form their own contingent work sample. Predetermination of work samples outside of this contingent sample would foreclose on the tacit criteria of nominally `good' and `original' properties that are anticipated as being represented in the works. Resemblances and stipulative identities fail to allow for misrepresentation within their representing relation (Dretske, 1986 p.74, Dennett, 1987 p.293). This is because an analytic approach to definition legislates a relation between sense and referent which is unlikely to be misrepresented. In a similar way defining the relation between curriculum subjects and outcomes by using a combination of work samples and stages, aims to fix the terms on which teachers analyse student's progress and thereby, perhaps, insure against teacher error. Easy to apply at first stipulation through stages and work samples can, nevertheless, leave teachers without a clue as to the overarching system which unites the relation (Perner 1991, p.24). Let us suppose, for example that an educational authority has, like the Board in New South Wales, divided the visual arts into five outcome stages. At the moment of division the subject matter defined as stage 3 could just as easily be defined as stage 4 as stage 3. But once the stages are set the divisions are stipulated by work samples of artworks. In the beginning, teachers might have little trouble in interpreting the stages by simply matching achievement to the samples. While ever the definition by work sample works consistently, the students, parents and principal are satisfied. However,at some point the teacher might be presented with an unexpectedly `original' student performance at stage 5 which the teacher misrepresents as an average performance at stage 3. What causes the stage to be wrongly classified? It cannot be caused by the student's performance because student's works cannot cause their stages. The problem is one of classification. Yet how is it possible for the teacher to be wrong if s/he has applied the standard consistently? It is now possible to see why. The `originality' of the student's performance in the example qualifies it as a candidate for stage 5, but its `originality' gives it a lack of resemblance to the stage 5 work sample. Later on, however, the definition was made fallible by the `originality' of the performance, resulting in its relegation to stage 3. Transitive relations of assessment through work samples can never be wrong. A square represented with an extra side is not a misrepresented square but must necessarily be some other figure entirely. A stage 5 performance in art, however, must be able to retain its classification within stage 5 even though it `mis-represents' the stage in some way. By itself and to begin with the stipulation of work samples provides an infallible yet relatively meaningless interpretation of student performance.In this case the property of originality could not be registered by the teacher, resulting in the performance being misrepresented as another stage. The grasp of misrepresentation is the metaphysical realisation that curriculum outcomes bear an opaque rather than a transparent relation to the world. It is the idea that the existence of a misrepresentation is not automatically an untrue or counterfactual account of any state of affairs. It may, in other words turn out to be an untrue account of the world but an account which is nevertheless believed to be true or a `mis-represented' state of affairs; or alternatively, an account made according to an alternative convention.(Perner 1988, p.144) With a development in the tendency for teachers to see outcomes and standards as the representation of a network of terms, held together by a variety of conventional and true notions of form, value, and function, comes a parallel confidence in teachers' instructional autonomy. Outcomes, kinds and pedagogical explanations Teachers rely upon pedagogical notions to provide ready made explanations of things like `outcomes' and `standards', since teachers are aware that a body of subject content carries no implications for its representation as educational outcomes.(Eltis 1995) I have argued that outcomes and standards vary according to assumptions about the kind of things that they are. Pedagogies that admit of outcomes as natural kinds, non-strict natural kinds, or nominal kinds impose differing obligations on the ways in which achievement within a subject content is assessed. Different pedagogies admit and exclude properties of outcomes, as well as vary the relations between them. A sample of children's drawing might be considered a poor achievement as a nominal kind, but considered a psychological `treasure' as a non-strict natural kind. Thus choices about teaching method constrain the way in which subject content is represented to the curriculum. The mutability of subject content in the face of pedagogical differences highlights the importance of considerations given to the relative autonomy of the teacher in making their choice. This the authoritarian agendas of the national curriculum tacitly concede, evidenced by their narrow stipulation of how achievement ought to be represented and their regulation of the pedagogical input allowed by teachers. The metaphysics of autonomous choice The problem with choosing among pedagogical explanations can be expressed as the problem of finding a sufficiently disinterested metaphysical point of view from which to make an unprejudiced choice. Perner describes it as being in possession of representational stories about representations. From Perner's point of view when the stories about `outcomes' and `standards' themselves are made into the referent these stories become meta-representations.(1991, p.35) Foucauldian archaeology is one method that opens up sufficient metaphysical distance for comparing and contrasting meta-represented narratives. Archaeology gains distance through its close mapping of the material traces of educational text.(Foucault 1972) Materialist in this sense, Foucault's archaeological method enables the chooser to remain uncorrupted by the mental assumptions contained within the stories, and to maintain a healthy suspicion of propositions about outcomes that claim to make references to an independently existing world. For Foucault, knowledge, particularly knowledge of human practices has little to do with representations of the truth. Knowledge of human activity, he insists is a self replicating and mutating artifact. Thus textual analysis is intolerant of knowledge references made in any way other than back through a chain of prior texts. Textual analysis must at all costs remain independent of the beliefs they represent. Although appropriate, Foucauldian method is impractical for this project for two reasons. First it is time consuming and requires meticulous attention to historical detail to be effective. Second it is compromised by its emergence as a pedagogical framework in its own right. At the other extreme the realist Richard Boyd provides a distanced if unfashionable alternative.(1988, p.181) An arch realist Boyd swims against the current tide of anti-causal explanations and plurality of local constructs. The centre piece of Boyd's realism is his notion of `homeostatic property-cluster' definitions. Homeostatic property-cluster definitions have been developed within the non-naturalistic Wittgensteinian tradition of ordinary language analysis. Boyd supports these traditions in their attack on the defining of natural kinds as necessary and sufficient conditions.(p.196) He elaborates: According to various property-cluster or criterial attribute theories, some terms have definitions which are provided by a collection of properties such that the possession of a number of these properties is sufficient for falling within the extension of the term.. It is supposed to be a conceptual (and thus an a priori) matter what properties belong in the cluster and which combinations of them are sufficient for falling under the term. In so far as different properties in the cluster are differently `weighted' in such judgements, the weighting is determined by our concept of the kind or property being defined. It is characteristically insisted, however, that our concepts of such kinds are `open textured' so that there is some indeterminacy in extension legitimately associated with property-cluster or criterial attribute definitions. The `imprecision' or `vagueness' of such definitions is seen as a perfectly appropriate feature of ordinary linguistic usage....The property-cluster definitions provides an extremely deep insight into the possible form of natural definitions.(p.196) Boyd continues: There are natural kinds, properties, etc. whose natural definitions involve a kind of property-cluster together with an associated indeterminacy in extension. Both the property-cluster form of such definitions and the associated indeterminacy are dictated by the scientific task of employing categories which correspond to inductively and explanatorily relevant causal structures.... What I believe is that the following sort of situation is commonplace in the special sciences which study complex structurally or functionally characterised phenomena.(p.196-7) Thee are 11 propositions which condition homeostatic property-clusters, the most significant for this project being cited as follows: 1. There is a family F of properties which are `contingently clustered' in nature in the sense that they co-occur in an important number of cases. 2. Their co-occurrence is not...a statistical artifact but rather the result of what may be metaphorically...described as a sort of homeostasis . Either the presence of some... of the properties in F tends... to favour the presence of others.... 3. The homeostatic clustering of the properties in F is causally important: there are (theoretical or practically) important effects which are produced by a conjoint occurrence of (many of) the properties in F together with (some or all of) the underlying mechanisms in question. 4. There is a kind term t which is applied to things in which the homeostatic clustering of most of the properties in F occurs. 5. This t has no analytic definition; rather all of the homeostatic cluster F together with some or all of the mechanisms which underlie it provide the natural definition of t The question of just which properties and mechanisms belong in the definition of t is an a posteriori question - often a difficult theoretical one. 6. Imperfect homeostasis is nomologically possible or actual... . 7. The relative importance of the various properties in F... is a theoretical rather than a conceptual issue.... 10. The causal importance of the homeostatic property cluster F together with the underlying homeostatic mechanisms is such that the kind of property denoted by t is a natural kind... (p.197) It is a realism which: predicts indeterminacy for those natural kind or property terms which refer to complex phenomena; such an indeterminacy is a necessary consequence of `cutting the world at its largely theory independent joints.' (p.199) Frank Keil, who has made extensive use of Boyd's homeostatic property-clusters remarks: Not only does this view suggest that there is more to natural kinds than characteristic features, it also suggests a new way of viewing the nominal kind concepts so common in older studies on concept development. As the contrasts between natural kinds and artifacts demonstrate, almost all kinds may have some set of causal mechanisms that are at least partly responsible for the correlations among their properties. Except for the purest nominal kinds, definitions may not be as fully arbitrary as they seem. They may be arbitrary to the laws of a particular natural science, but not relative to the laws governing human interactions, intentions and culture.(1988, p.55) Homeostatic clusters map the properties and interrelations within definitions. A definition sets property p into causal relation with q, what Kornblith refers to as a `dependant' relationship of properties (1994). The set of properties in a definition are included for contingent, theoretical reasons and these, therefore, may turn out to be wrong. Strikingly, even nominal kind concepts are also represented in homeostatic clusters as theoretical artifacts. Definitions are strung together into networks of sub causal properties or sub concepts, whose presence in the network is not determined analytically as an intension,but are admitted on the basis of their empirical contribution to the causal explanation. Boyd's homeostatic `mechanisms' extend relevancy to external properties implications for establishing counter intuitive cross category links. Properties are added and subtracted throughout the life of the definition and, as he insists, the relevant set may never be complete in the definition of complex entities. Figure 1 sketches how a definition of `outcomes', for example derives its intension as a homeostatic cluster of conjoint properties that relate to each other as a causal mechanism. The lines in the diagram schematise the bivalence of the relationship between one conjoint property in the cluster and another. The nodules `teacher', `student', `parents', `subject content' and so on are, for the purposes of this illustration, coarse approximations of relevant functional properties. How the agency of `teacher' shapes the function of `subject content', for example, explains the extension of the concept as a direction of causality in the representing relation. Definitions vary according to the theoretical constraints on the way in which properties can be imported into the reference and be legitimately shown to represent the referent. For example, a concept of `outcomes' should be able to show whether it is possible to transport `creativity' as a property of `subject content' in art, as a property into `outcomes' of artistic achievement. It should be emphasised that the presence of `outcomes' in the homeostatic network constitutes only one agency among others and is not the definitive property. Note that little attention is paid to intuitive intensions of `outcome' such as those that appear in the Board's definition above. The definition of `outcomes' extends to the homeostatic cluster represented on the map and explanations will outline the direction of causality among its conjoint properties. Thus in some pedagogies the relations between the conjoint cluster terms may allow the transportation of, say, `creativity' into the property of `outcomes', while in others it may not. The power and sophistication of a concept is expressed in terms of its capacity to explain the representing relation between the reference and its referent. Explanations about `outcomes' are stories which are as much about the externally related kinds `teacher', `student', `subject matter, `employer' and `parent', as they are outcomes themselves. Other explanations may borrow functions from the same cluster, but explain their causal relationship and their agency in different ways.  Predicting the referents of outcomes within educational meta-narratives A number of questions can be framed on the concept map which test alternative relations between the functions of `teacher', `student', `employer', and `outcome'. The purpose of these questions is to explore variations in the meaning of `outcomes' among the explanation provided by different pedagogies. Variations in meaning are characterised by sampling the properties which are transported between the sense and referent of `outcomes' within each narrative. For instance, consider the relational question `can a bad teacher make a good outcome?' This question tests the conditions under which the properties of `teacher competency' are transported between the functions of `teacher' and `outcome'. The significance of the representing relation in this question challenges the belief that standards serve as a hedge against indifferent teaching. Or take the question - `can an individual student make an individual outcome?' This question inquires into the terms under which the properties of `student uniqueness' are able to be represented as an outcome, and tests the concern that national curriculum standards may be harmful to the individuality of students. The conditional question - `could a general outcome satisfy the needs of a particular employer?' ponders the different educational terms under which a liberal curriculum has vocational relevance; and so on. The operative terms `make', `satisfy', and so on in these questions are consistent with the causal force that Boyd sees as describing the relations between the conjoint properties in the cluster. Take the following relational question - `Can a difficult subject make an easy Outcome?' This particular question challenges an understanding of the relation between the properties of outcomes and the properties of the subject content which outcomes represent. Sketches of the following five educational meta-narratives give five different explanations in support of their yes/no answers (A1, A2, etc.) in their account of a simple tree drawing task in art. A1.No. Just as there are some trees that are more difficult to draw than others so more `difficult' subjects like mathematics have more difficult outcomes. This is a typically naive realist explanation. It explains the relation between trees and drawing, `subjects' and `outcomes', in terms of a - content > to > outcome - direction of causal agency on the concept map. In this explanation, the thing represented, the tree, is causally implicated in its representation, the drawing. It suggests, for example that hard subjects make hard outcomes. Naive realists omit the role played by students, teachers, settings and other agencies that might be counter intuitively opaque within the referent, and therefore violate the condition of asymmetry in the representing relation. `Common sense' or folk explanations of this kind are most likely to be expected from school age students, and lay parents.(Bruner 1992) Employers who believe that subjects like mathematics and art are of little value to vocational education, may be labouring under the naive realist misapprehension that abstract subject matter causes abstract outcomes and for this reason are of little practical value. Naive realists represent outcomes as something that the properties of a subject cause students to know. Any challenge to this common sense is likely to puzzle naive realists because the referent `hard' subject matter is unable to provide them with a common sense reason why the relation is being misrepresented. A2.Yes. If students understand that drawing trees is an expressive art then the properties of the particular tree being drawn are irrelevant to the difficulty of the drawing. Outcomes which nominate a conceptual grasp of art as an achievement, increase the accessibility of a subject to students. This narrative is typically idealist. It assumes achievement to be a nominal kind. It is founded on the relation believed to exist between subject content and its representation as a token in the student's mind. Idealist outcomes are typically narrated in the psychological language of intentional objects and beliefs. Seen as reflective things outcomes represent a private transaction between student and subject matter. Idealists explain the answer to the relational question as a - subject content > to > student > to > outcomes - direction of causality where students cause the representation of the difficult tree as an outcomes of their conceptual understanding. Thus the drawing of a particular tree is seen as a behaviour caused by intentional belief, rather than as a naive outcome caused by the tree being drawn. Idealist outcomes emphasise the gestation, revisiting, and mental re-description of content, finding absolutely no place for outcomes as end on teleology. The language of outcomes employs the descriptions of `underlying', `basis', and `grounding' that picture the achievement of outcomes in terms of a deeper grasp rather than as successively higher levels of performance. Idealists would argue that outcomes represented as steps and ends foreclose on the potentiality of student's imagination. To attempt the representation of conceptual understanding itself as a competency, outcome or standard, would merely invoke the need for an intermediate set of meta-understandings. A3.Yes. Students are enculturated into tree drawing genres at school. Tree drawing genres serve to impose a structural convention on the particular qualities of the tree being drawn. Different purposes for the tree drawing are co-extensive with different genres of tree representation. In much the same way, semiotic outcomes reflect different cultures of meaning in particular subject content . These structures sort out different references according to the purposes which they serve. Hard subject matter is made easy within familiar convention of reference. This is a typically structural narrative (Goodman 1974, Kress 1993). Structural outcomes are represented as a narrative of translation and communication. Just as the colonial artist Louis Buvelot painted eucalypts as European elms, students are entrenched in their own conventions of tree drawing which they impose as an interpretation on particular trees. The semiotic narrative explains outcomes as a - subject matter > to >teacher > to >outcomes - linkage of causal relations. Easy outcomes are made possible by the clarity with which difficult subject matter can be communicated through shared codes. Genre approaches to pedagogy represent an expansion of the old formalist semiotics of language into a social semiotics. A social semiotics understands subjects content as a series of culturally recognisable genres of exchange. For example, the introduction of functional grammar into the wider curriculum in New South Wales fostered the translation of outcomes, in subjects other than English, into a series of recognisable communicative routines. In functional grammar there are `tree' ways of drawing and `geography' ways of writing - about trees (O'Toole 1994). The differences in meaning between the art and geography is marked by their communicative purpose, and their purpose is shaped not so much by psychological intention as by cultural convention. The communication narrative constructs the information of a subject, rather than seeking to interpret the history of information inherent within a subject. A4.No. Tree drawings in school are simulacra of other tree drawings in school. Profiles and work samples actually develop their criteria by reference to an average school based performance and not by reference to methods of drawing and the artworld. Consequently there is no point at which tree drawings ever reach a referential reconciliation with real trees or even `real drawings' beyond the school. `Outcomes' are educational texts which mutate and quote other educational texts in ways that are unsystematically related to subject content. `Art Express', the annual museum exhibition of high school student's art in NSW, for example, may be little more than a self quoting multi-text, that never properly engages with the subject of the visual arts until the point when it leaves the school and falls under curatorship by staff of the Art Gallery of NSW. This is a typically post-structural explanation (Baudrillard 1983). The specification of outcomes cannot be attributed to the achievement of either students or teachers, specifically because the learner is not considered to be a primary agent in the achievement of outcomes; or, as Paolo Freire points out, even of their failure to be achieved. It is characterised on the map in figure 1 as a - subject > to > outcomes >to > outcomes - direction of causality. The satisfaction of outcomes can be explained by reference to the politics of sovereign and knowledge power, without the need of reference to mental events or students' abilities at all. National standards, in all of their precision, may have hoped to arrest sovereignty over inappropriate `outcomes' from teachers in New South Wales schools. However, from a post-structuralist point of view they are more plausibly explained as the mutation of a Thatcherite doctrine from the United Kingdom. Post structuralists would argue that it remains to be shown how standards, outcomes, profiles and other educational texts can refer to subjects systematically. Many suspicious art teachers have yet to be convinced that outcomes are about subject content and not just about other bureaucratic texts. For a post-structuralist pedagogy, outcomes are not demonstrable as end state performances but about the mechanisms of textual transmission in the politics of education. A5.Yes. Drawing a tree is co-extensive with `knowing how' to make a hard tree easy to draw. For example, `knowing how to choose' from a repertoire of documentary approaches, including photocopying; `how to manipulate' one copy, chosen in consultation with the teacher, which is nearest in similarity to drawings from the late 1970's expressionist revival; then, to be able to `find out' how to apply a graphic mark that satisfies demands for freshness and expressive spontaneity, not to be attempted on the final drawing until a sufficiently spontaneous level of quality is `able to be reproduced', and only after `extensive practice' in mastering late abstract forms of graphic representation; to be `able to make knowing references' to the German artist Anselm Kiefer's sombre metallic planes; and applies a technique of rendering which `refers to criteria' developed in workshops on Kiefer quality rendering, using school based graphite simulation. A gallery owner is left in little doubt as to what a candidate satisfying this outcome can do. The gallery owner would not even need to see a portfolio of works from this candidate since their abilities are stated so unambiguously in the outcome. A student embarking on their journey toward this form of tree drawing outcome is made to feel confident that the level of participation required of them is equally unambiguous. Student's can be secure in the assurance that a high quality artistic outcome is guaranteed. This answer is described in the narrative of `knowing how'. Following Gilbert Ryle it resolves the mind-body problem by collapsing `knowing that' (idealism) into performative, or enactive knowing. In other words, `knowing how' tends to replace the assumptions of conceptual understanding with a range of heuristic to algorithmic performances.(1949) Wherever judgements are required of teachers, they are assisted by criteria represented as a repertoire of abilities. The power of a student's underlying aptitude is disclosed as a set of competencies. Idealism is rejected on the well worn grounds that, for example - knowing the rules of chess is categorically distinct from the aptitude for competent chess playing. The `knowing how' narrative explains outcomes as a network of - subject content > to > student disposition > to > outcome - causal relations on the map. Tree drawing in the accompanying answer (A.5) carries no implication of a coherent body of content, but, through its representation as a sub-routine, or `scaffolding' of performative abilities, designed to complement the underlying dispositions of students to learn, it is made over into outcomes which make it appear as if it does. Concepts, values, judgements, and so on are externalised as performative representations of the underlying ability `to think critically', `to value' and so on. Rather than a metaphor of deeper understanding, the narratives of knowing how use the rhetoric of computer games where achievement is expressed as a stage like movement toward higher performative levels. It is not assumed that the underlying dispositions are presented to students' reflective consciousness. As a Neo Tylerian form of soft behaviour or disposition, knowing how lends itself to criterion referencing because it offers the benefit of revealing the internal mechanisms of thought as performance indicators. Outcomes of knowing how not only present teachers with the right answer to grade, but lay bare all of natural the sub-routines of thinking to mark as well. Sub-routines make the identification of levels easier because of the way in which they separate performances into a temporal sequence. In `knowing how' narratives subject matter is represented as a series of accomplishments, rather than as an integrated network of relations. Instead of the one conceptual understanding, reintegrating and changing qualitatively over time `knowings how' can be stripped out into any number of sequential levels of performance. Even though performance outcomes can range from protocols through to innovations, as enactments they nevertheless represent innate dispositions as good practice. `Thought', for example, is more likely to be narrated as `solving relevant problems by using the school library' rather than as `private contemplation' or as a quality inferred by the teacher from a student's `thoughtful drawing'. The true of subject content is represented as the good of curriculum. Competency based outcomes, as noted at the beginning of this paper, are marked by the language of `ought' and `should' in national curriculum documents, rather than by the language of `is'. Summary These five sketches entail different assumptions about the kinds of things `outcomes' are. There is no implication that one pedagogical narrative is any more valid than another, even though Answer 5 seems to be the currently preferred candidate of national curriculum protagonists. The causal relations expressed in each narrative, which began naively as a subject content > to > outcomes relation, were expanded within the other 4 narratives from a dyadic into a triadic relation. This expansion represents the insertion of an additional function into the homeostatic explanation. Although each narrative is different in its description of outcomes, their explanations deploy similar properties. Teachers can make their pedagogical choices according to the kind of outcome they hope to bring about realising, however, that quite different properties of subject content are admitted and excluded in the characterisation of student achievement as a result. In this realisation teachers will be aware of the ways in which pedagogical decisions frame achievement in the subject content they teach. The professional autonomy of teachers is thus partly understood as the informed ability to choose their pedagogy,. Finally, it seems to me that Boyd's account of homeostatic clusters introduces a useful theoretical power into the explanation of nominal kinds. Beyond its surface application in the five sketches described in this paper, Boyd's realism has potential for a more effective mapping of cultural structures and folk beliefs in education. References: Baudrillard, J. (1983). Simulations, New York: Semiotext(e) Inc. Boyd, R. (1988). How to be a moral realist, in G. Sayre-McCord (Ed.) Essays on moral realism, Ithaca: Cornell University Press. p.60. Bruner, J. (1990). Acts of Meaning. Cambridge. Harvard University Press. Consortium of National Arts Education Associations (1994) National standards for arts education. Reston, VA: Music Educators National Conference. Dennett, D. (1987).The intentional stance, Cambridge Mass: The M.I.T. Press. Dretske, F.A. (1986). A semantic theory of information, Knowledge and the flow of information, Cambridge Mass: The M.I.T. Press. Eisner, E. (1993). Why standards may not improve schools, Educational leadership, February: 22-23. Eltis, K. Focusing on learning: report of the review of outcomes and profiles in New South Wales schooling, The Ministry for Education and Training, August 1995. Freeman, N. & Brown, N. (1993). Children's developing beliefs about art as a basis for sequencing in art education,' Proceedings of the Australian Association for Research in Education . Fremantle, Western Australia. Goodman, N. (1974). Languages of Art. Indianapolis: Hackett Publishing. Gough, A. (1995) `The Faustian Bargain...', Agenda vol.2, No. 2. Foucault, M, (1972). The Archaeology of Knowledge. Hausman, J (1994). Standards and assessment: New initiative and continuing dilemmas. Art Education 47 (2), March, 9-14. Keil, F. C. (1989). Some traditional views of conceptual development reconsidered, in Concepts, kinds and cognitive development, Cambridge Mass: The M.I.T. Press. Kornblith, H. (1994). Beyond foundationalism and the coherence theory, in H. Kornblith (Ed.). Naturalising epistemology (Second Edition) Cambridge Mass: The M.I.T. Press. Kress, G. (1993) Genre as social process, in M. Kalantzis, and B Cope (Eds) The Powers of literacy, University of Pittsburgh Press. Kemmis, S. (1986). Curriculum theorising: Beyond reproduction theory., Geelong: Deakin University Press. Mason, R. (1992). Evaluating the content of art teaching, Artists in the 1990's: their education and values, Issues in art and education, Volume 1, Wimbledon School of Art in Association with the Tate Gallery. Mayer Committee: Employer Related Key Competencies, Commonwealth Government of Australia, 1991. Meighan, R. (1995). National curriculum, Royal Society of Artists Journal . March: 27. National Curriculum Council Consultative Report: Art, York: UK, 1992. National Curriculum (1993) The arts curriculum framework, Australian Capital Territory Department of Education & Training. Norris, S. (1992). Introduction: The generalizability question, in S. Norris (Ed.) The generalizability of critical thinking, New York: Teachers College Press. O'Toole, M. (1994). The language of displayed art, Rutherford: Fairleigh Dickinson Press. Perner, J., (1988). Developing semantics for theories of mind: From propositional attitudes to mental representations. In J. Astington, P. Harris, D. Olson. (Eds.) Developing theories of mind. Cambridge: Cambridge University Press. Perner, J. (1991) The concept `representation', Understanding the representational mind. Cambridge Mass: The M.I.T. Press. Putnam, H. (1975). The meaning of meaning, in K. Gunderson (Ed.) Language mind and knowledge. Minneapolis: University of Minnesota Press. Ryle, G. (1949) The concept of mind, Cambridge: University of Cambridge Press. Syllabus Model Using Staged Outcomes, New South Wales Board of Studies, (March 1996), and Curriculum Outcomes, (September, 1991, p.5). Schwab, J. (1969). The practical: a language for curriculum, School Review, 78:1-24. Schwartz, S. (1978). Putnam of artifacts. The Philosophical Review, 87, 4: 566-574. Schwartz, S. (1979). Natural Kinds and Nominal Kinds, Ithaca: Cornell University Press. Notes: 1. Raises the issue of supervenience, what Simon Blackburn says intuitively is `whatever it is by way of natural or physical states that bring it about that the subject is F...at least the belief that whenever a thing is in some F state, this is because it is in some underlying G state.('Supervenience revisited, in G Sayre-McCord (Ed) Essays on Moral Realism, Ithaca: Cornell University Press p.60). The problem with supervenience is the problem of explanation. It involves the task of establishing the principled link that allows an inference from the underlying state to the supervening state to be made. Typically artworks are judged ostensively by ranking in order one against the other, like judgements made about wine. The tacit ability to rank artworks is assumed as an expression of trust in the existence of an underlying aesthetic disposition to judge, which is universally shared and able to be exercised without mediation.