Preview

Whilst these assumptions are consistent with the learnings of the autobiographical sketch give in the last chapter, I have not felt it necessary, or advisable, to enter into any sort of justifying dialogue regarding my position. This is not a philosophical study, and I have always regarded justification as a loser's game.

So I have presented my philosophical position as a set of assertions with an internally consistent logic; I have briefly described the epistemological, ontological, and axionomic assumptions that have informed this study, and described how that position fits into current post-positivist, interpretivist, and post-modern paradigms.

The chapter ends with a brief outline of the assessment process constructed from my particular position.

Philosophical assumptions : What is knowledge? What is truth?

Any experience is experience (action, feeling, perception) of an event, either directly, or as recalled or as transformed in memory or action. So all experience involves relations. As all knowledge must finally depend on experience, all knowledge involves knowledge of relations; so all knowledge is constructed out of relational events.

To experience an event does not necessitate giving a meaning to that event, but does require a state of awareness or consciousness, from which the event is viewed. For example, an experience may be represented by a pattern or abstract painting which embodies relations without embodying meaning. Giving a meaning to an event requires some theoretical underpinning, some ideas or ideals; some knowledge of relations derived from other events, or possibly, if mathematical relations are construed to constitute meaning, derived from acts of imagination that transcend (are transformations of) known relations. Mathematics can be regarded as a special case of patterning, and whether mathematical propositions or systems have meaning in themselves is moot. I don't think they do. Some post-structuralists want to deny experience that excludes meaning and thus language. My experience denies their denial. Their assumptions refute my denial. Stalemate. But then, I'm writing this thesis.

I use the term meaning to involve more than prediction, which mathematics can sometimes help to accomplish. Meaning involves some reason, some purpose, some intention, some value. Thus meaning is inevitably embedded in language, itself embedded in human discourse. Unless we take a mystical view and define the meaning as the experience itself, or rather as a particular encompassing experience, in which case discourse stops and the world in its oneness pulsates. In this thesis I shall hold to the more mundane view. To do otherwise is not to proceed.

In this epistemology, experience precedes pattern, and pattern precedes meaning. "Whether we are talking about unicorns, quarks, infinity, or apples, our cognitive life depends on experience" (Eisner, 1990, p31). Meaning will then usually in its turn, but not necessarily, pre-empt and distort experience, which will then in its turn influence events. Buddhist meditation is designed to limit this distortion; which brings its participants on this issue close to post-positivists like Phillips (1990), who seem ultimately to define objectivity as the reduction of bias of various sorts.

Meaning is socially constructed because language is socially constructed. What passes for knowledge in common language is a social concurrence in a particular culture about acceptable meanings embedded in discourse. On the other hand, experience is constructed out of relational events not necessarily linked to any particular culture, and the construction of patterns or relations in response to that experience may also sidestep, or transcend, social patterning or common meanings. In other words, I hold the view that creation is immanent in all events, and in all perception of events, and change is more than the imposition of some random variation. Usually, however, we may assume that patterns are also culturally influenced.

Data is a particular form of knowledge constructed by particular people for particular purposes. Such purposes always involve the construction or isolation of events in which the observer is directly, or indirectly through associated theory, involved; for example, measuring devices involve the observer at one step removed. Thus all data, being knowledge, is constructed from events, constructed and/or observed for particular purposes. All data, to be used, must have either a predictable pattern, or a meaning, or both. So if data is to be useful, it must have links to other relational events, or have links to (uneventful) abstract relations.

It follows that, in this world, there are as many potential truths about an event as there are experiences of the event. To the extent that all experiences of the event are the same then there is a case for "the" truth. But how would this be known? Any attempt to know this would involve the sharing of meanings, which are certainly socially constructed and can be as varied as the cultures and relations and metaphors that are used to make sense of them and communicate them. So agreement about one meaning, one truth, represents conformity about social construction as much as it does concomitance of experience.

Ironically, in a social context the idea of multiple truths is unificatory, whilst the notion of one truth is fundamentally divisive; in practice the notion of one truth contradicts the collaborative ethic and supports interaction characterised by entrenched positions. Search for "the" truth is often productive within a closed space of cultural assumption, but does not lead to open inquiry outside that space; rather it invokes defensiveness, and if necessary violence in order to sustain its inviolability. Inevitably it leads to fragmentation and conformity, as contradictory elements break away to form their own "truthful" reality, and all else becomes subservient to "truths" current fashion (Feyerabend, 1988).

One more point about multiple truths; such a claim does not contain the inference of the catastrophic consequence that all "truths," that is, socially acceptable beliefs, are equally useful or sustainable, or that some cannot be falsified. At least at the level of physical definition, it is demonstrably false that I am constructed entirely of green cheese. Such a claim is not a valid contender for any claim to a truth beyond that of a very idiosyncratic and metaphorical form. Truth claims about events can never be proved, but some truth claims can be demolished through procedures of contradiction.

If data belongs to an event, it cannot be attributed to a particular agent or aspect of that event. It is common and comforting to attach data to particular objects or participants in an event, and to the extent that all other participants and relations that constitute the event are held constant and made overt, to that extent attributing the data to a particular agent constitutes a valuable shorthand in description and discourse. For example, to attribute a certain tensile strength to a steel beam is convenient, but has meaning only in regard to an event at which, at a certain temperature, the beam is stretched in a machine until it breaks. The time span within which this (hypothetical) event generates the same data is quite long. But over a thousand years, the steel beam no longer has this property; which is shorthand for saying it will behave differently in the event that it is stretched. Not only that, but any engagement in events will affect the tensile strength in an unpredictable way; if an unbroken part of the beam is stretched again it will be found to have a different tensile strength; as it will after multiple vibrations as part of a bridge.

So experiments in the physical and biological sciences do not produce data about the object, or measure properties of the object being investigated. They produce data about the event that is the experiment. Most experiments describe the behaviour of physical or biological objects under particular boundaried, that is, controlled circumstances. The information they give therefore is not so much about the "natural" world in which we and they live, as it is about the "controlled" world that is the experiment, and sometimes becomes habitualised as technology. Most social research has fallen into this trap of misrepresentation of the source and attribution of data.

Social events, or indeed interactional events of any sort involving living things, have time spans of small duration. Indeed, identical events are impossible to create because social relations, and the participants involved in them, continually change. Even if we could hold all the conditions constant as we do for the steel beam, the data still cannot be attached to the person because, even more so than for the steel, the person of tomorrow is a different person; and part of the difference is attributable to the experience involved in obtaining the data.

It follows from this epistemology that most psychological descriptions of people are shorthand and problematic descriptions of social events, from which most elements that constitute the event are camouflaged. The label is attached to the person even though the events which produced the data involved social interactions. This is an example of faulty labelling. In particular it applies to any notions of skill and competency that do not clearly define the context of their application.

So the issue of objectivity is not that things exist independently of the mind; the issue is whether things (elements) have properties independently of the events used to describe them. To say that a thing is real (has material existence) is very different to claiming that its "properties" are real and belong to it.

Ontology: What is the nature of social reality?

On the other hand, epistemology precedes ontology in that all meanings are socially constructed, and are thus ultimately dependent on social relations and that includes the meanings we ascribe to ontology.

Regardless, the two domains interlink with no inconsistency in terms of the idea of social relations and the idea of knowledge being a function of experience of relational events, and meaning being socially constructed.

Using relations as a primary explanatory factor negates the notion of causality, at least in a simplistic sense. Events are construed as interactive systems where everything effects everything else; patterns of mutual influence replace causality as an explanatory principle. This has been generally accepted in Physics since the work of Einstein and Eddington early this century. It has always seemed odd to me that the more complex the system in which the event occurs - from physics through to biology through to social relations - the more frantically the idea of cause is clung to.

Further to that, the idea of "reality" is similar to the idea of "truth"; a redundancy, an unnecessary complexity, an irrelevant diversion. It contributes to conflict rather than to productivity. It seems more useful to talk about what aspects of social relations intrude most on experience, and are important to the intensity and duration of that experience, and the effects that it generates. In this regard I would make four assertions about social events, conclusions from my own experience and reflection:

• knowledge of social relations (that is, data generated within human interactions), is usefully construed in terms of the power and affect relations of the participants in the event; in particular, asymmetrical power relations generate different data than do symmetric power relations; and positive affect different data to negative affect (Foucault, 1988).
• an event occurs within specific localised power and affect contexts; this is not to suggest that this event might not itself be embedded in power relations (economically, racially, nationally or gender influenced) which push the effects and experience of the event in particular directions, but does put less emphasis on such grand power relations.
• events are dynamic, not static situations; they are characterised by movement, by change. They exist in time, which could be considered one measure of their change. So data about social interactions, which may often be characterised by power and affect relations, will change over time as the power and affect relations themselves change. I assume that any new social relationship (any social event characterised by people who have not met before in that configuration) will initially be asymmetric in respect to power, and moot in respect to affect. The relational changes will affect the data generated through interaction, which includes discourse, and vice versa.
• Fixed societal structures (e.g., hierarchies) crystallise power relations and negate change. To the extent that they are successful they may produce knowledge, consensual interpretations, limited by the very boundary conditions that make its production possible; fixed societal structures also, in time, contradict the flow of interactional life, and produce social pathology.

No knowledge is value free. As Lincoln (1990) puts it, "given the criticism from all quarters, . . . only the most intransigent or the most naive scientist still clings to the idea that inquiry can, or should, be value free"(p82). Being socially constructed, knowledge produced from inquiry is related to the meanings and purposes and structures within which it was composed; and it will tend to confirm or negate those relations involved in its construction, depending on the interests and attitudes and assumptions and awareness of the researcher. Even if data could be produced that was independent of those elements and relations, that very independence is itself a value position, which could be construed either as objectivity, because it has transcended bias, or as ideology, because it camouflages the power relations from which its bias necessarily derives.

As a researcher my task is to contribute to the meaning system that helps me and other people make sense of their experience in the particular class of events with which this study is concerned. They will make sense of it if it is a story that links in some way with their experience, and at the same time is not contradictory to their experience; experience that is, of course, already partly interpreted in terms of other stories.

As an educator my task is to change people; education is nothing if it does not result in change. And as change is inevitable, but may be in many directions, there is obviously an obligation on the part of the educator to specify the direction in which change is intended.

As educator-researcher I must interact with the people with whom I wish to do research or educate. I do this through process (how I do the research), and product (what I produce as a result of the research). If I do not produce the data I investigate, but merely interact with data produced by someone else, this simply pushes the value problem one step backwards; their data was not value free. So if I accept their data without criticism, then I am accepting and perpetuating the values that affected its construction and effects. If I question that data, I question the social values embedded in it, as much as the social effects that are manifested through it.

If whatever I do involves interactions with people, and the construction of knowledge, then whatever I do affects both the meanings of people, and the social relations involved in those meanings. This is not to say that describing "what is" implies approval and acceptance of what is. Rather it is to claim that the very description of "what is" implies a way of viewing the world, a relationship with the situation, an involvement in the construction of the data, that pre-empts the meaning of the data by hiding the value assumptions behind the very mechanisms of its construction; becomes, that is, symbolic violence, unless made explicit (Bourdieu, 1977). Most quantitative research and much qualitative research is in this sense symbolically violent, in that the sources of its power are disguised.

Unless I wish to engage in a value contradiction, it seems necessary to have an awareness of the direction in which I wish to move people's overt and covert experience of social relations and the meaning systems construed within their influence; and to use processes and meanings that are congruent with those purposes.

My autobiographical note indicates that much of my work over the past thirty years has been involved with the nature and practice of violence in its various forms, especially as it affects young people.

My construction of the concept of structural violence (Wilson, 1992) indicates that I regard fixed hierarchical structures, in all their multifarious visible and disguised forms, as inevitably connected to structural violence and hence to social injustice. Due process within legal systems is necessary to alleviate, or control, some of the social fallout, but is not sufficient to ensure social justice at its root manifestation, which requires more equalitarian structures.

Peace and social justice are ideals that have many forms and faces that change over time. On the other hand, physical and structural and emotional and symbolic violence are constructs amenable to more specific definition, and hence more easily recognisable in particular social events. For this reason, I feel more comfortable having as a basic value the reduction of violence, which I could universally advocate, than with the increase of social justice, which is more nebulous because of its many-faceted nature; on this view, increase in social justice that is not associated with reduction in violence would be problematic, involving as it does an internal contradiction.

If beliefs (truths) are multiple, then so must be the values that are implied in those beliefs, or which inform them. How then can any particular value position be maintained as superior to any other?

In regard to the specific events that involve me and others in this thesis, I would answer that while the value of reducing violence is not necessarily superior to others, in the context of this work it is consistent with:

• 1. The learnings (culture and gender influenced as they are) that I have constructed out of my life experiences.
• 2. The ontology and epistemology which I have described, which inform the assumptions on which this study is based.
• 3. A view of life and living that involves ideas of growth, change, and flow at both individual and social levels. As such it is consistent with many views of personal enlightenment and social justice.
• 4. Processes likely to favour the survival of human life on the planet at a time when the technology is available, and primed to destroy it (Schnell, 1980).
• 5. That universal attunement and compassion which is one aspect of the experience described as mystical, as cosmic consciousness, or as the perennial philosophy, which transcends historical and cultural boundaries, and contains a sense of the sanctity of each individual person (Wilber, 1991).

Some doyens in the research game still regard qualitative social research as an exotic rather than a native plant, and as such something to be treated with caution because of its possible ecological effects on what had previously seemed to be a very secure and threat-free environment. Specifically, many testing experts still live in a positivist world (Shepard, 1991). As well, most teachers are quite convinced that their tests measure their student's attainments; the correspondence theory of knowledge may well be discredited, and philosophically empiricism may well have been dead for forty years (Smith, 1993), but in schools and colleges and universities and work places it is alive and kicking. However, a rich literature has developed from the debates involving qualitative research over the last ten years (Burgess, 1985; Eisner & Peshkin, 1990; Guba, 1990 Popkewitz, 1984; & Smyth,1994).

So with some reservations qualitative research is now accepted and respectable, even though practice severely lags theory. The reservations are currently crystallising as sets of questions and answers about how to recognise "good" qualitative research. For example Carr and Kemmis (1985) describe five formal requirements for any adequate and coherent educational science (p158). Criteria and caveats are being constructed that will undoubtedly in time result in a new orthodoxy (Lincoln, 1990). Feyerabend's (1988) assertion that "science is an essentially anarchic enterprise; theoretical anarchism is more humanitarian and more likely to encourage progress than its law-and-order alternatives"(p5), provides as much discomfort in the research world, be it quantitative or qualitative, as in the world of politics or the family. Smith's (1993) work clearly indicates that clarification of the problem of criteria is central to any real progress. It is also necessary if any substantial change in educational practice, and associated structural relations, is to occur.

At this point in time, however, the limits of the field are blurry, and the demarcations between various camps subject to border skirmishes. So at least one reason for my position not fitting into a specific ontological, epistemological, axiological, or methodological tent is that such tents are not clearly differentiated between the encampments. Having said that, it is possible to nominate some camps to which I do not belong, and some camps to which I partly belong, where I would not feel too uneasy sitting in some of their tents.

It is generally agreed that there are three basic positions; empiricist (post positivist), interpretivist (constructivist), and criticalist (Smith, 1994; Lincoln, 1990). It is also agreed that this is an over simplification.

Briefly, empiricists argue that there is a reality out there to be discovered, that it is single and measurable, and that causal laws explain and predict it (Smith, 1994).

Carr and Kemmis (1983) characterise the interpretive approach to social science as aiming "to uncover the meaning and significance of actions" (p92). The interpretive position is that truth is constructed by people, and always involves a social context and social interactions. So truth is relative and multiple. This position has two strands, the ethnographic (Sherman & Webb, 1988), and the ontological strand (Eisner, 1988). The difference is in the way hermeneutics is regarded. In the ethnographic strand, hermeneutics is a method of achieving interpretive explanation; in the ontological strand hermeneutics is more concerned with the idea that all knowledge, all representation is dependent on the primacy of experience (Schwandt, 1990). Regardless, "hermeneuticists of all measure and variety agree that any interpretation of meaning must take place within a context" (Smith, 1993, p16).

Carr & Kemmis (1983) regard post-positivist and interpretivist accounts to be similar in that "the researcher stands outside the research situation adopting a disinterested stance in which any explicit concern with critically evaluating and changing the educational realities being analysed is rejected"(p98). However, some constructivists (Lincoln, 1990), more recently advocate an abandonment of "the role of the dispassionate observer in favour of the role of the passionate participant" (p86). This is a position with which I concur. Smith (1993) elucidates other similarities and differences in the various positions:

The other strand of the criticalist position is the post-structural, post-modern strand, which includes some feminist perspectives. The concentration here is on the construction of social reality through language and discourse, and the way in which this serves dominant groups and interests. The emphasis in research is on discourse analysis, in order to expose such inequities (Smith, 1994). Foucault's work is sometimes attached to this strand, though he himself did not accept the classification. And I would agree. This is important, because the writings of Foucault considerably influenced this study.

So where does my position fit into all this? I am not a positivist or empiricist. I do believe that empirical data can be collected about events; it's just that I don't believe that in relation to social events such data is very stable, can be replicated without considerable error becoming evident, or can be justifiably attached to a particular participant constituting the event. Any such data views that event from a particular position, with particular boundaries, with particular interests and values influencing the collector.

On the other hand truth claims are sometimes explicit, and often implicit, in theoretical formulations or interpretations involving social events. And some such claims can be directly contradicted by empirical data, by effects or consequences that are directly observable.

In terms of ontology, of the nature of reality, I do not fit neatly into any of the camps; empiricist, interpretivist or critical. I am probably closer to being a sceptical mystic. Rather than enter into that potential bog, in this thesis I have bypassed the question of "reality" and begun with the notion of social events, which involve the participants in social experiences.

I am constructivist or interpretivist in as much as I see all knowledge as multiple and constructed. Eisner (1990) agrees that experiences are the basis for cognition and knowledge: "thinking and knowing are mediated by any kind of experiential content the senses generate...our language refers to referents we are able to experience, recall or imagine"(p91). However, as Schwandt (1990) points out, this ontological basis of experience is not common to all interpretivist methodologies.

Perhaps my main point of departure from the criticalist perspective is at the ontological level; certainly I see relations as fundamental in as much as they constitute the mechanisms through which difference and change occur, thus making events experiencable. But I do not wish to "objectify" these into some grand historical schema on the one hand, nor overemphasise their dependence on gender relations or particular discourses on the other. Rather, I see power and affect relations as a "heuristic fiction" that has great generality and elegance as an explanatory and generating principle. However, I am clearly allied with them in their wish to reduce the violation of persons through the transformation of social structures and in seeing social research as a legitimate way to help people make sense of the social world in a way that gives them some leverage to change it for the better. By "better" I refer to a decrease in violence.

A model for the assessment process

Given the ontological position of the above discussion, the assessment process involves (at least) five stages (events) and a context. In actual practice some of these stages may be omitted or fused. Such fusion or omissions may constitute a source of confusion or error.
 

• 1. Test production: An event (experiment, test) is devised to produce data. Such an event will involve an interaction between the assessed person, and instrumentation of some kind. The instrument may exist in the assessor's head, or may be produced as a physical artifact (a written test). The test production process also involves explication of a theory-practice link of some sort, and some prior judgments about a relevant task.
• 2. Test experiment: The person being assessed does the test, by performing what is required in the testing situation. This is the first stage of data production, and this event is completed when the test is completed.
• 3. Data production: The second stage of data construction occurs when the assessor interacts with the testing process directly, or with products from it. eg. a performance or a completed test paper. This interaction involves an interpretation of the data.
• 4. Judgment process: This results in a categorisation of some kind; it involves a comparison of the data with the standard, either directly, or by comparing with data about other students. This process assumes the existence of the standard as a stable and replicable element in the event.
• 5. Labelling process: At least two labels are involved; the name of what has been assessed (described), and the name that describes the level of performance (compared to the standard). The multiple label is constructed from the whole assessment process, and is legitimately attached to those events. In practice it is more likely to be attached to an element of the testing event (the assessed), or to an even more remote theoretical construction related to the assessed (some skill or ability).
• 6. All of these processes are embedded in relations of power which reproduce and invigorate themselves in the processes. And all of these processes (events) are potential sources of error and confusion in the individualised material product of this whole process - the documented labelling and categorisation of the assessed person.

Summing up