The Competence Debate     Graduate Employability      Regeneration     e-skilling & cyberontology     Live Working Papers    Useful Links

Learning as a confidence trick :
exorcising the ghost in the machine

"Learning is the key to prosperity - for each of us as individuals, as well as for the nation as a whole."
(David Blunkett, Secretary of State for Education and Employment; foreword to The Learning Age, DfEE1998: 7)
 

The rise of the discourse of learning

It is difficult to doubt the centrality of the notion of learning in contemporary discussions of social and economic affairs. The current vogue for forming neologisms incorporating the words 'learn', 'learning' and 'learner' should provide fertile ground for students of the history of ideas, particularly within the Kuhnian (paradigms) or the Foucaultian (discourses and truth regimes) traditions. As with the discourse of competence, there appears to be a degree of irresistibility in the manner of its rise (cf Holmes, 1995a); whereas a couple of decades ago there would be discussions education and training, now such discussions tend to be framed in terms of learning . Thus continuing education has been displaced by lifelong learning (DfEE, 1998), debates on further and higher education are based on the notion of a learning society (Dearing, 1997), the journal Management Education and Development has transmuted into Management Learning, and so on. Along the way, students and trainees have been transformed into learners, teachers and trainers are now facilitators of learning, educational aims and training objectives have become learning outcomes, courses are learning programmes, teaching and training methods are now learning methods. My aim in this paper is to analyse the notion of learning as it has now come to be used in such discussions. In particular, my focus will be on learning taken to refer to an individualised active process, something which an individual does.

This idea of learning as an activity is indexed by the change of use of learner from an adjective (eg as in learner driver ) to an agentive noun, ie where the suffix -er denotes the doer of an action. It can also be seen in the way that libraries, especially in educational institutions, are increasingly being renamed as learning centres, that is, places for (individual) learning rather than as places of learning in the sense of a public repository of accumulated knowledge. The analysis here will follow two courses. First, I shall engage in conceptual ground-clearing, examining the logic or grammar of the language of learning; I shall seek to show that the verb to learn, and its various tense forms, cannot denote an activity sui generis but rather connotes an evaluation of other related activities. Secondly, following from recognition of this evaluative connotation, I shall examine the essentially social character of learning, and the implications of this for current issues in education and training, particularly in respect of the skills agenda in higher education and of the management competence movement.

Displacing education and training

Admittedly, modern educational and training theory has perhaps always been concerned with issues of learning, particularly through the application of psychological theory and research. Teacher training courses include the study of psychology as a foundation discipline. Courses for training workplace trainers usually include a session on how people learn, covering some basic ideas about what were taken to be key principles of human learning. Textbooks on training, or more broadly on human resource management, often have chapters partly or wholly devoted to the theory, or theories, of learning (eg Beardwell and Holden, 1997; Harrison, 1997; Reid and Barrington, 1997; Torrington and Hall, 1998). It is understandable, then, that participants to the activities and processes of education and training, especially those who are professionally involved, should be interested in, and concerned to understand, learning. However, what has changed is the manner in which the notion of learning is presented as the central issue for the system of education and training. Gilbert Jessup, chief architect of the edifice of National Vocational Qualifications, asserts that

"The measure of success for any education and training system should be what people actually learn from it, and how effectively. Just common sense you might think, yet this is a comparatively new idea." (Jessup, 1991:3; emphasis added)

"Lifelong learning requires a holistic approach organised from the students perspective. In practice the first change must be one of language. We must use the word learning, not education. ... if we still need the terms further and higher, let us refer to further learning and higher learning." (Wagner, 1998; emphasis added)

It is not uncommon to hear educational and training practitioners insist that they are more interested in learning than teaching/ training, often referring to their roles as facilitators of learning rather than teachers or trainers. Education and training are thus displaced as the focus of interest by this emphasis on learning, and debate on how the notion of learning relates to those of education and training is brought to premature closure. One key element in the rise of this new discourse of learning is the development of forms of certificating what are deemed to be the outcomes of learning accomplished outside of formal education and training settings, usually referred to as accreditation of prior experiential learning (APEL). Universities, particularly the new post-1992 universities, have played a part in this in terms of allowing credit towards an award within a credit accumulation and transfer system. However it is the development of the National Vocational Qualification (NVQ) system that has played the major role, with the separation of assessment of competence from the processes by which such competence may have been acquired. With no formal requirement to undertake any particular educational or training route, individuals may present themselves for assessment on the basis of their claim to be competent as a result of their experience. The specification of the type of evidence required to demonstrate competence remains the same, whether an individual undertake a course of education/ training or they claim prior experiential learning . The primary form of evidence is that of observed performance of disaggregated parts of a job or occupation. Qualifications are thus constituted as certification of learning outcomes, whether the learning takes place through formal education/ training or informally through experience.

A further element in the rise of the discourse of learning lies outside such formal credentialising contexts, particularly in relation to the workplace. Key influences here include the ideas of Revans on action learning (from the 1940s onwards: see Revans, 1982), Rogers on learner-centredness (1970), Knowles on andragogy (1970, 1973), Kolb on the model of the experiential learning cycle (Kolb, 1971, 1984), the self-development movement (eg Pedler, Burgoyne, Boydell, 1978), Honey and Mumford building on Kolb's model (1982), Marsick and Mitchell on informal and incidental learning (1990). The mélange of ideas from such writers now forms the stock of common-sense knowledge often espoused by practitioners within the management training and development field; indeed, many of such writers and their supporters typically display their practitioner credentials in the rhetorical presentation of their ideas and their implications for practice (note 1). Learning is presented as natural, taking place mostly outside formal educational and training settings; indeed, most real ( significant ) learning can only take place when individuals are taking charge of it themselves. They can and will do this, it is claimed, when they face their own real life problems and are not prevented from doing so by constraining forces such as the authoritarian nature of much formal education and training.

Such ideas can be seen at work in the rationale presented for attempts to reform the education and training system. Placing the learner at the centre of the process of learning enables the didactic role of teachers/ trainers to be marginalised. Real life, practical settings and problems now come to be seen as both the best source for learning (as experience ) rather than existing bodies of knowledge ( theory ); they also provide the most valid basis for assessing learning, through workplace performance rather than teacher-set coursework assignments and examination. Moreover, if learning is a clearly-understood process, individuals can learn how to learn and take responsibility for their own learning, particularly through attending to their experience in order to develop, through reflection, generalisations which may be actively tested to generate new experience. In principle, this approach requires no involvement of another to bring about learning. The role of teachers and trainers is now residual, to facilitate the cyclical process whilst retaining the primacy of the learners practical experience. This contrasts with their former role, regarded as overly concerned with didactic transmission of theory . Boak (1991), referring to the Kolb's experiential learning model, suggests

"Perhaps we can see in its rise and popularity the same forces that have pushed to prominence competence - as opposed to book learning and the mastery of theory." (p. 45; emphasis added)

Learning: an active process?

In this new learning orthodoxy we can identify some key assumptions about the nature of learning, which I shall seek to examine in this paper. First, learning is regarded as a process; it involves change rather than a continuous state, some sequence with a precursor state being replaced by an outcome. A corollary of this is that learning takes place over time. Secondly, learning is entitative; that is, learning takes place within individual, monadic entities; such entities are therefore the basic unit of analysis for locating the process. Usually, we think of these entities in terms of individual human beings, but by extension we can also think of organisations as learning. Such entities are differentiated from their environments; whilst the environment can act as the stimulus for learning, and the context in which this is displayed, the learning process itself is, we might say, entitatively located and bounded. Thirdly, significant learning (that which is non-trivial, makes a difference ) is an activity, a process actively carried out by the individual, who can therefore be called a learner ie the performer of the activity of learning. As such, it is amenable to some degree of direction, control and management by the learner; that is, the learner can learn to learn.

The shift in the last-mentioned assumption to significant learning relates to key developments in the theorising of learning over the past three decades. The first two assumptions can clearly be seen in the conditioning theory or behaviourist approach, dominant until the 1960s, whereby learning is taken to be the process through which the pattern of an organisms particular behavioural responses to particular environmental stimuli becomes established through reinforcement. The first and second assumptions can also be seen in the cognitivist approach, which came to prominence in the 1960s, whereby learning is treated as information processing within the learner's cognitive or mental structures. Cognitivism is concerned with meaningfulness of learning, although this is sometimes treated rather narrowly in terms of connections between new knowledge (information) and existing knowledge (cognitive structure). It is in the development mainly from the 1970s of notions of significant learning, with its emphasis upon an active process involving experience, that we see all three assumptions. Meaningfulness is interpreted differently in Rogers (1970) presentation of humanistic theory of learning, and in Knowles (1970, 1973) writings on andragogy, the art and science of helping adult learn . For both of these learning is a self-initiated active process, fully engaging the individual learner in determining what and how they should learn, and how such learning should be evaluated. Experience is central to such a view; this is not the sensory-input of behaviourism, nor the information of cognitive theory, but involves the whole person ie the affective (feeling and emotion) and the cognitive (thinking).

Experience is also central to the model developed by Kolb (1971), termed the Experiential Learning Model (note 2). In this Learning is conceived of as a four-stage cycle. Immediate concrete experience is the basis for observation and reflection. These observations are assimilated into a theory from which new implications for action can be deduced. These implications or hypotheses then serve as guides in acting to create new experiences. In this approach the teacher serves as a facilitator of a learning process that is basically self-directed. (Kolb, 1971: 2) Kolb later provided a working definition of learning as

"the process whereby knowledge is created through the transformation of experience." (Kolb, 1984)

"The last 20 years have witnessed a transition from management training to management learning. The shift of emphasis is away from developing the capacity of the instructor to transfer knowledge effectively to a passive learner, and towards facilitating learning, and helping the learner to learn. This is reflected in the displacement of the dominance of cognitive learning theory by experiential learning theory." (p.142; emphasis added)

One process or many? Learning and language-games

The notion of a learning process sui generis is, on the face of it, intuitively valid. After all, we talk easily of someone having learnt something, a skill, a theory, some important (or even trivial) information, and so on. If a small child at time t1 cannot ride a bicycle (without stabilising wheels) more than a few feet without falling over, yet at time t2 can ride enthusiastically and with barely a wobble over an apparently unlimited distance (subject only to their rapidly changing focus of interest), surely they have engaged in the intervening period in a process of learning to ride? We also say that someone is learning to drive, to speak French, word-processing and so on, thus apparently referring to some on-going process. Although there are differences in the actual or intended outcomes, what has been learnt (past tense) and what is being learnt (present tense ), the fact that we use one term surely indicates a common process?

One problem with this view is that it is clear that there are significant differences in how what is taken to be learning occurs. The common distinction in education and training between knowledge (as information recall), understanding, physical or motor skills, social skills, attitudes, is often related to different types of learning processes. However, it might be replied that at some higher level of abstraction there is some common process, perhaps with some minor differences in the extent to which particular parts of the process operate depending on the type of learning outcome to be accomplished. Thus physical skills may be developed through repeated physical practice, whereas knowledge is developed through repetition of the information to be committed to memory in order to be available for recall. Alternatively, it may be conceded that different types learning outcomes result from different learning processes, but these different processes share the common characteristic of being changes brought about through experience (note 3).

This definitional legerdemain should give us cause to pause, and take Wittgenstein's advice to look and see whether there is anything common to all (Wittgenstein, 1953). As part of his argument that the meaning of utterances must be sought in their use in context, rather than by some fixed correspondence with some order of reality, Wittgenstein uses the notion of language game. To counter the objection that he has failed to say what is the essence of a language game, what is common to all the activities called language games, he asks us to consider what is common to proceedings we call games. If you look at them you will not see something that is common to them all, but similarities, relationships, and a whole series of them at that.(op.cit. 66).Continuing, he invokes the idea of family resemblance:

"I can think of no better expression to characterise these similarities than family resemblances ; for the various resemblances between members of a family: build, features, colour of eyes, gait, temperament, etc. etc. overlap and criss-cross in the same way." (op.cit. 67)

So too with learning ; might we not do better to look for family resemblances between the situations in which we use the term rather than some common defining characteristics of all such situations? One method often used by teachers and trainers to introduce learning theory is to ask the learners (let us keep to the now-conventional terminology) to think of some occasions when they personally have learnt something, to consider how that learning took place, and so deduce some theory of learning. This is the approach taken by Kolb et al (1971), and also by Reid and Barrington (1997). Alternatively, a set of examples is presented; indeed, Reid and Barrington do both. The correct conclusion is, purportedly, that there is indeed some common characteristics, particularly that behaviour is changed. What these approaches fail to do is ask the question what are we doing when we use the term learning in these cases?, or, in Wittgenstein's terms, what language game are we playing? .

In particular, we might point to ( look and see ) circumstances in which talk of learning is merely part of some conversation amongst family and friends, and circumstances in which such talk forms part of official discourse, as in formal educational and occupational settings. The circumstances of a proud parent relating to the child's grandparents that Abigail's learnt how to ride her bike are significantly different from the secondary school teacher's end-of-year report that she has learnt to construct basic sentences in French using the present and future tenses. They differ again from the circumstances in which, in her part-time employment as a post office clerk, her manager records that she has learnt to issue motor vehicle taxation discs. Of course, this does not show that different circumstances necessarily indicate different meanings of the term learnt, learn, and learning . But the a priori assumption that there is a single common meaning of the term learning (and other forms of speech from the root term learn ), free of content and context, is severely undermined by Wittgenstein's language-game argument. Even the reference to change of behaviour fails to perform the task of demonstrating common meaning. Indeed, often we use the term learning when referring to new behaviour, as in the case of the child learning to ride a bike; to talk of this as change of behaviour would appear to be a somewhat forced usage. Besides that, when we explain how we have come to change the way we do something, we would normally give reasons; for example: "I try to avoid the M25 during the rush hour; (I 've found that) you can never tell when there's going to be a tailback somewhere or other". The simple statement I tend to avoid the M25" would tend to evoke the question "Oh, why's that then?", to which the response "I 've learnt to" would be deemed to be insufficient explanation. So too would be the response I 've had experiences which I 've reflected upon, and developed a generalised notion which I 've tested out in practice and had confirmed by other experiences!  

The language and logic of learning

In attempting the elucidate how we ordinarily use the language of learning, one important starting point is the difference in usage of the tense formations of the verb to learn, particularly the past (simple and perfect), eg he learnt, she has learnt, and the present (continuous), eg she is learning . Superficially, these appear to be just straightforward differences in the time at which the utterance is made and that in which the purported process occurs or occurred. Closer examination, however, yields significant differences in meaning.

Taking the past tense formations first, we should note how these relate to statements of capacity, ie can, knows, understands etc. Thus if someone says Jack has learnt to swim, this statement can usually be taken as meaning the same as Jack can swim . Or to say Sue has learnt the difference between latent and manifest function, is (usually) the same in meaning as Sue understands the difference etc. . In particular contexts, there may be subtle but significant difference in meaning, often given by other contiguous utterances: for example, the first statement above might be made in response to the question And what did Jack do over the summer vacation? . But in these examples, such differences make no significant points about learning. We might, in some contexts, wish to emphasise that ability has been acquired, rather than that it was in some way treated as innate, or that was no more than one might expect; eg Ivan has learnt to speak English (he's from Russia) ; I 've learnt to hold my tongue (I was prone to lose my temper) . Such usages tell us about something about the circumstances which pertained hitherto or other particular of the situation, but not about some purported generaliseable process, learning, which has occurred. In general, the past-tense use of the verb to learn may, without significant loss of meaning, be substituted by the words indicated above, ie can, knows, understands, etc.

In his attack on the Cartesian doctrine (or myth ) of the mind-body split, Ryle (1949) subjects just these types of words to analysis. He refers to utterances using such words as dispositional statements, rather than descriptions of occurrences or episodes. The problem is that the para-mechanical notion of the mind in the Cartesian approach, what he calls (with deliberate abusiveness) the dogma of the Ghost in the Machine, has tended to confuse us about the logic of such statements: the vogue of the para-mechanical legend has led many people to ignore the ways in which these concepts actually behave and to construe them instead as items in the descriptions of occult causes and effects . (op.cit., p.112-3) Rather, he argues, dispositional statements should be treated as testable, open hypothetical or semi-hypothetical statements. Hypotheticals ( if ...then... ) are law-statements, making no statement of fact but nevertheless being either true or false depending on whether or not particular matters of fact satisfy the two parts, the if... clause (the protasis) and the then... clause (the apodosis). The point of establishing such laws is to enable us to infer from particular matters of fact to other particular matters of fact, to explain on fact or set of facts by reference to others, or to bring about or prevent certain states of affairs (p.117). Ryle uses the term inference-ticket, by analogy with a season ticket, which licenses such inference, explanation, and manipulation of states of affairs. Hypotheticals must be open, ie not refer to any particular persons or things. By semi-hypotheticals, he is referring to dispositional statements which do mention some particular persons or things, as is the case when we say Steve can swim using the butterfly stroke or Jean knows the Greek alphabet . They are law-like in the way that we use them:

"They apply to, or are satisfied by, the actions, reactions and states of the object; they are inference-tickets, which license us to predict, retrodict, explain, modify these actions, reactions, states." (op. cit., p.119)

Since the use of the past tense of the verb to learn normally carries the same meaning as words such as can, knows, understands, and so on (excepting the examples above, which are not significant for our purposes here), we must treat it as a dispositional word. As such, its function is in the construction of hypothetical and semi-hypothetical statements, rather than in the reporting of some occurrence. That is, to say that someone has learnt something, or has learnt to do something, is to make an inference claim which is used to explain, retrodict or predict performance. We should also note, in passing, that the same applies to the use of the term competent, ie to refer to someone as competent is to make an inference about their performance in the past (as explanation or retrodiction) or future (prediction) (Holmes and Joyce, 1993: 40) (note 4). To ascribe competence, to say that someone can do something, that they have learnt, are merely different ways of expressing dispositional statements. The use of past tense of to learn does not normally describe the execution in the past of some identifiable activity called learning .

What then of the present tense of the verb to learn ? We do say things like Mike is learning to drive, Reena is learning to produce Web pages, Ben is learning the flute and so on. Surely these refer to the activity of learning, in which they are engaged? A problem that arises from this is that, if learning is an activity, it is not observable. It is tempting, then, to locate learning in the mind, as an activity of some incorporeal inner self (in humanistic approaches) or central processing unit (in cognitivism). Of course, there may be observable activities related to what we call learning. Thus, we can see Mike sitting in the driver's seat of a car with the instructor sitting beside him, as they travel along the road making various manoeuvres. But we don t see him separately learning to drive. Moreover, he doesn' t cease learning to drive when the lesson is over. If a visiting aunt asks the 17-year old Mike what he is doing with himself these days, and Mike replies Oh, I m learning to drive, it would be considered strange of the aunt to say Don t tell lies; you re sitting watching the television ! The supposed activity of learning thus seems to take place in some strange dimensions of space and time, such that it cannot be observed and is taking place even when it is not being performed.

Again, Ryle's discussion of dispositions and occurrences can provide us with a method for analysing the present-tense usage. He argues that there is an important class of occurrence or episodic words which, because they are active verbs, have tended to make us oblivious to their logic. These are success or achievement verbs; the examples he gives are win, unearth, find, cure, convince, prove, cheat, unlock, safeguard, conceal . These correspond with task verbs, with the force of trying to . Sometimes we use an achievement verb as a synonym for a task verb (or verbal phrase):
Hear is sometimes used as a synonym of listen and mend as a synonym of try to mend . (op. cit., p. 143) A major difference between the logical force of a task verb and its corresponding achievement verb is that, in using the latter, we are asserting that some state of affairs obtains over and above that which consists in the performance, if any, of the subservient task activity. (ibid.) Thus, for a doctor to cure a patient, she must both treat the patient and the patient must be well again. Ryle notes that there may be achievements without a task performance: for example, success may also be ascribed to luck. In addition, we may use a success verb in anticipation, with the possibility that we will revise the usage in the event of failure:

"[Someone] may rashly claim the expected success, but he will withdraw his claim if he discovers that, despite his having done the best he could, something has still gone wrong. I withdraw my claim to have seen a misprint, or convinced the voter, if I find that there was no misprint, or the voter has cast his vote for my opponent." (op.cit.,p.144)

"When a person is described as having fought and won, or as having journeyed and arrived, he is not being said to have done two things, but to have done one thing with a certain upshot. Similarly a person who has aimed and missed has not followed up one occupation by another; he has done one thing, which was a failure." (ibid.)

As Ryle says, success verbs belong, put crudely, not to the vocabulary of the player, but to the vocabulary of the referee (p.145). So too, I would argue, with learning. An undergraduate studying industrial sociology and learning about Braverman's deskilling thesis is not doing two separate things, but one thing (studying) successfully. To say she has learnt is to say she has studied successfully. We can also apply this to teaching (demonstrating, explaining, telling, etc.) and the now-fashionable notion of facilitating learning . The use of the latter is as a phrase to indicate success in teaching; it is neither the same as teaching (which may be unsuccessful) nor a separate, superior activity.

To summarise and conclude this discussion, then, we can say that the verb to learn, in various grammatical forms, does not refer to some activity performed by someone. Rather, when we examine how we use the term we can see different jobs that it performs within particular contexts. When used in the past tense, its primary use is to make a dispositional statement, one that licenses or warrants inferences about performance, in the past, the present or the (anticipated) future. When used in the present tense, and sometimes in the past tense, it serves to make an evaluation of success of certain well-understood tasks, which are either observable or can be elaborated in terms of observable activities. Such activities include studying, reading, practising, rehearsing, discussing, performed by the individual trying to learn (ie achieve the dispositional state of knowing understanding, being able to), and teaching, explaining, showing, questioning, and so on as performed by someone helping another to learn (succeed in studying etc.). There is no need to assume that there is some common process invisibly taking place within individual entities; there is especially no need to posit some activity, called learning, that is performed by the individual, separate from and in addition to these activities. This is not to deny the usefulness of much of the research and scholarship within psychological studies of learning, but to attempt rein back the enthusiasm with which many people have treated such studies as breakthroughs in an age-old field of human endeavour. The attempts by evangelists for learning to displace well-established terms in the field of education and training by words and phrases (such as learner, learning facilitator, student-centred learning, learning goal, learning programme ) tends to obfuscate rather than illuminate. As Winch (1998, p. ix) puts it:

"The rapid growth of the study of learning as a branch of psychology has been largely responsible both for the increase in information and the decline in clarity on the topic."

Learning and confidence

The above discussion is, of course, primarily a conceptual and logical analysis. Noddings (1984) disputes Ryle's analysis, which she takes to be compatible with certain forms of competence-based education. However, in referring to success verbs and to inference-tickets, there is no specific conclusions that can be drawn about how we establish the criteria for evaluating success, that is, how inference-tickets are produced. This can only be examined in relation to the particular types of social circumstances within which we wish to make such evaluations and inferences. Indeed, Noddings concludes that the logical foundations of competence-based (teacher) education by stating that:

"We shall have to look towards social and political explanations to find out what in the world is supporting it." (op.cit., p.27)

I shall now go on to do this in respect of education and training, particularly with reference to the employment arena. Within employment-oriented education and training, the main issues of inference are those to do with predictions of performance (note 7). That is, when we are deciding whether we would say that someone has learnt something (knowledge, skill, competence, etc.), we are making some judgement about the kind of performance we might anticipate at some future point in time or over some period of time. Another way of putting this is to say that judgements of whether or not someone has learnt (that they are competence) are expressions of confidence or trust that future performance will be as desired (Holmes, 1994). On the basis of such confidence, we award a qualification, recruit a candidate, allow progression on a course, promote an employee etc. Or, more likely, we alone do not take these decisions and actions; we are part of a wider set of actors (or stakeholders ) and other social arrangements in which such decisions and actions are taken. Issues of confidence then become critical in the maintenance of these social arrangements; we might say, then, that learning is a confidence trick ; without confidence and trust, social systems for ascriptions of competence will soon degrade and decompose (note 8).

In attempting to understand how confidence is created and maintained, we can make use of Gergen's (1989) notion of warranting to refer to the way in which individuals and groups attempt to have their version of social reality accepted (note 9). Gergen argues that there is an indeterminate array of versions of reality that may be applied to any given occasion. Different actors involved may prefer differing versions; whichever version comes to dominate may be highly significant for those actors.
Given a range of competing constructions, and sufficient stakes in the outcomes, there may be brisk competition over whose voice is honoured . (op.cit, p73) Formal situations of assessment in the employment arena, particularly selection processes and the award of occupational-oriented qualifications, certainly come within the kinds of situations in which the stakes are high for whatever version of reality dominates, ie becomes inscribed as the official record. Central to such stakes are issues of confidence and reputation. For the applicant for the job or award, whether or not they are deemed to have learnt (are competent) significantly affects their occupational reputation. For the selector, there will be issues of what account they will be able to give to others who may make judgements of the applicant selected, once they are employed; that is, ascriptions of competence (ie having learnt) may be withheld or withdrawn by those others, so affecting the reputation of the selector to make proper selection judgements. In the award of qualifications, the reputations of assessors, educational or training institutions, awarding bodies, and even the awards themselves are at stake; that is, issues of quality assurance are essentially issues of reputation. Gergen continues his elaboration of the notion of warranting by arguing that One of the chief means by which voice is achieved in such wide-ranging situations is through conventions of warrant. (op.cit., p.74) That is, the version of reality being claimed is constructed through conventionally accepted rationales and justifications. Now, in that assessment involves the claim that future performance will be as desired (or not), the grounds for such a claim is likely to be successful only to the extent that it is articulated in terms of such a convention of warrant. That is, the confidence that assessment judgements are valid will only be attained and maintained through the use of conventionally legitimised modes of assessment. We might therefore call these conventions of assessment (Holmes, 1994).  

Performance as a convention of warrant

Conventions of assessment may take a variety of forms; indeed the recent attempts to reform the vocational qualifications system and to change the pattern of assessment in higher education demonstrate this. Central to the claim that NVQs are a superior form of occupational credentialisation is that the assessment of competence is based primarily on observed workplace performance rather than on traditional methods such as written examinations (Jessup, 1991; p. 46). Within higher education, there have been increasing calls for assessment of skills which, whatever subject studied, are deemed to be appropriate for graduates, particularly in terms of their employability . This skills agenda has been promoted through various initiatives, notably the Employment Department's Enterprise in Higher Education and the RSA's Higher Education for Capability, and was endorsed by the Dearing Committee (Dearing, 1997). Yet prior to the mid-1980s, such formulations were not in evidence in discussions of the relationship of higher education to employment (note 10).

It may be tempting to seek to adopt the idea of performance as providing firm ground for making assertions of learning and competence, to enable us to say with certainty that someone has learnt whatever they were supposed to learn. This is the approach adopted within the NVQ system, whereby detailed performance prescriptions form the primary basis for assessment of candidates for the award of an NVQ. If they meet the standards set, they are then deemed competent . This might also seem to be the case in other forms of assessment, with less degree of specification. After all, when a student writes an essay or report, makes an oral presentation in class, or writes answers to questions in an examination, surely they are engaging in performances which are assessed? Protagonists of the skills agenda in higher education would seek to specify the performance requirements for the various transferable skills or capabilities deemed to needed for graduate employability. Isn t it then just a matter of getting the right specification of performance requirements, a subject for an empirical analysis? Such an approach cannot, however, resolve the fundamental issues of risk and confidence. There is no necessity that, just because the required performance has been observed in one context, similar performance will be forthcoming in another context. Indeed, much of the research on transferability presents a pessimistic view (see Bridges, 1992). The best that we can do is infer from past/ present performance to anticipated future performance. This reinforces Ryle's inference-ticket view of the logic of dispositional statements discussed above. However, the social logic of such inference implicates the warranting conventions which operate in the particular contexts in which performance is important. Transferability is thus not an attribute of some purportedly transcendental skill or competence but is immanent in the conventions of warrant connecting the contexts of performance.

The term performance is used profligately by proponents of the competence movement (note 11), but is never properly examined. The assumption is that performance, and/ or the outcomes of performance, can be directly observed; this assumption, of course, runs contrary to the distinction between behaviour, action and act . A description of some activity by a person can be made in purely behavioural terms, usually some bodily movement eg Marie's arm shot up . We might choose to analyse this in terms of causes, eg the heat of some object her arm had just touched, or some muscular spasm arising; the behaviour is unintentional. This differs from the description of an activity in terms of intentional behaviour, an action, eg Marie raised her arm . This differs yet again from the description of the activity as an act, eg Marie made a bid for the vase (or hailed a taxi, greeted her friend, etc.) (see Harré and Secord, 1972; Harré et al, 1985). An action is socially meaningful only in relation to the act which it is intended to accomplish. In terms of those social contexts with which we are here concerned, in respect of issues of learning and competence, it is clear that it is socially meaningful, intentional behaviour, or action, that is important. However, performance cannot be simply represented as action. First, there is the problem of how we recognise actions within the virtually unlimited and inchoate range of behaviours/ movements available to our observation. Secondly, there is the problem of how we judge that a certain action is oriented towards the accomplishment of some certain act:

"There are many kinds of acts that can be performed with the same action, and there are many kinds of actions by which we can perform the same act." (Harré, et al., 1985: 83)

Moreover, as an action is intentional, there is always the possibility that it fails to achieve such intention, that the act is not accomplished. A crucial issue, then, for understanding how performance serves as a warrant for judgements about learning is that of how behaviour is constructed as action, and how success in accomplishing acts is determined.

One way of considering this process of transformation, of what is called behaviour into what is called performance, is through connecting two key terms in current social theory, practices and identity. The general notion of practices can be seen in the work of Wittgenstein (forms of life), Bourdieu (habitus), Foucault (discourse/ practice), Giddens (praxis), Harré (practical aspects of joint action) amongst others. Of course, these various approaches are by no means in full agreement with each other. But we can note the general the emphasis on social order being the continuously negotiated outcome of interaction between social actors but under inherited conditions and with socially-presented resources, particularly, but not solely, language. A particular instance of performance might thus be seen as an instantiation of practice; that is, behaviour becomes performance in so far as it is in accord with the practices local to the social arrangements in which it is enacted. So, for example, a manager's actions are deemed to be managerial to the extent, and only to the extent, that they accord with what counts as such within the social context of managerial forms of organising. This implicates not only managerial ideology (discourse) but a whole set or network of connected practices, including economic and political.

Identity is also a key term of relevance to the question of how behaviour becomes performance . Again, the notion has significant place in the work of the above writers, as well as the social interactionist tradition and much recent social psychology. The tension between the idea of the personal self as experienced, as an acting subject, and the social self, as decentred subject, raises key issues about the role of performance in the production of identity. That is, performance may be seen as behaviour appropriate to a social identity (note 12). Yet if reduced solely to such a formulation, this would seem to destroy any notion that the individual is an acting subject, has no personal identity distinct from social identity. This suggests that we must see identity formation as a dialectical process, whereby the claimon a social identity is iteratively made, and its affirmation is negotiated. Successful outcome results in conventionalisation (Harré, 1983), ie both the reproduction of the social order and the production of the social identity within the individual's biography.

The discursive warranting of judgements of learning and competence, and of their expression in the form of decisions on the award of qualifications, selection for jobs, etc., is thus a thoroughly social process. The notion of performance is understandable in terms of the negotiated construction of an action or actions as instances of practices appropriate to particular social arenas. Such negotiated construction is itself part of the process by which identity is produced, ie claimed and affirmed; in the same moment, the recognition of identity provides the basis for recognition of action as performance, ie an instance of practice.  

Conclusion: Whither the educational theory of learning?

The notions of practices and identity figure centrally in the situated learning theory of Lave and Wenger (1991), whereby learning is examined as legitimate peripheral participation, the process of becoming a member of a community of practice. This view sees learning not merely as a condition for membership, but [as] itself an evolving form of membership. (op.cit.,p.53) The neophyte becomes an established practitioner through the gradual move to more intensive participation:
Moving toward full participation in practice involves not just a greater commitment of time, intensified effort, more and broader responsibilities within the community, and more difficult and risky tasks, but, more significantly, an increasing sense of identity as a master practitioner. (op.cit., p.111) The illustration utilised by Lave and Wenger are fairly close-knit communities of practice: Yucatec midwives, Vai and Gola tailors, naval quartermasters, meat cutters, members of alcoholics anonymous. The work of Grieco (1996) on the example of hop-picking in rural Kent by communities from London's East End appears to confirm the view that learning is best understood as a social process involving issues of identity and practice. In this case, kinship identity was a central feature:

"Apprentice hop labour was thus socially connected to all of the labour around it. Such an arrangement permits the fine-tuning of skill as messages about the quality of the performance of the learner are abundant. Working solely within the social vicinity of the family and its neighbourhood meant that apprentice labour was forced to acquire the working norms of the community." (op.cit., p.158)

The question might arise, then, as to whether such issues are pertinent solely or mainly where the community of practice is marked by strong boundaries in terms of inclusion/ exclusion. The occupational area of management, and the transition of undergraduates to graduate employees, appear to differ from such arenas. However, there are some indications that even in such areas, we can indeed see similar processes. In management development, emphasis has increasingly been placed on such approaches as action learning and mentoring. However, in general little attention has been placed on their essentially social nature; that is, in both action learning and mentoring, the positioning of the novice manager in respect of peers and established practitioners may be seen as critical to the effectiveness of the learning process . Most studies of managers and their behaviour are based on established managers, rather than those who are becoming managers. Hill's study (1992) on new managers, although mainly based within an individual process perspective, demonstrates the crucial role of identity formation based within the practice of management:
"As they began to act like managers, they became managers." (op.cit., p.90) Another study of novice managers (Holmes, 1995c) also clearly supports the utility of an identity project perspective, which pays due attention to the equivocal and uncertain nature of attempts by individuals to achieve recognition and acceptance as managers.

The skills agenda within higher education has, to a large extent, dominated recent debate about learning and the undergraduate curriculum. The empirical support for the validity of the notion of transferable skills is very weak, and the research on which this is based is, arguably, flawed in respect of begging the question (Holmes, 1998). What tends to be missing from the empirical record is the examination of the biographical and occupational trajectories of students after graduation. This lacuna is somewhat strange given the emphasis given by proponents of the skills agenda on the changing nature of graduate employment (eg AGR, 1995). If the higher education curriculum is to provide significant linkage with the post-graduation life, especially employment, of those who enter as undergraduates, then an adequate conceptualisation of such linkage is essential. The notion of graduate identity might serve as a key element within such conceptualisation (Holmes, 1995b).

The application of the notion of community of practice may seem more problematic, given the varied types of occupational arenas which graduates enter. We would need to adopt a less localised understanding, that is, to consider the extent to which there are affinities between such different occupational arenas. The fact that employers in such varied occupational arenas continue to perceive graduates to be suitable for employment suggests that their recruits are prepared for engaging in the practices appropriate to their post-graduation jobs. One important area for consideration in regard to some generalised practices is that of the discursive construction of meaning. This is supported by Bernstein's (1990) emphasis on the role of elaborated coding in complex social divisions of labour. It is also supported by Reich's (1992) analysis of the work of symbolic analysts . There is already considerable work on the role of the manager in constructing and controlling the meanings of events and situations (eg Shotter, 1993; Linstead et al.1996), whilst Abbott (1988) examines the manner in which professions seek to construct problems as amenable to their diagnoses and treatment prescriptions. We might, therefore, consider the well-established practices in higher education of requiring undergraduates to engage in written and spoken discourse not primarily in terms of their correctness, but rather as engagement in the form of life in which the discursive construction of reality is itself the primary practice.

The above analysis raises questions, then, about the validity and value of traditional psychological theorisations of learning, particularly in respect of higher education and of management education, training and development. Attempts to locate learning within individual persons, particularly in terms of an active process, fail to recognise the essentially social character of the task which the notion of learning performs. Learning is a term which is used in the construction of human activity as having a certain character which thereby warrants certain judgements about anticipated activity, and thereby warrants decisions and actions which have consequences within the social order. These consequences include the granting (or refusing) of entry to social arenas and of the right to (continue to) engage in activities within such social arenas. An educational theory of learning should therefore concern itself with its social character.

Notes
 
Note 1
 
In the theory of rhetoric, this would be an example of what Aristotle called ethos.
 
Back to the paper.
 
 
Note 2
 
Of course, the term experiential learning predates Kolb's use, dating back at least to the 1940s, particularly in the field of Organisation Development and the T-Group or Sensitivity Training approach (Cooper and Bowles, 1977; Reynolds, 1979). Weil and McGill (1989) identify four villages or clusters of interrelated ideas, meanings and practices associated with the term, whilst Robbins (1988) discusses the surfeit of synonyms , relating experiential learning to notions of self-managed learning and contract learning . But it is no doubt the simple pictorial representation of the circle in Kolb's model that has served to make it so pop,ar. It graphically depicts learning as an active process, basing learning in the individual's experience.
 
Back to the paper.
 
 
Note 3
 
Psychological definitions of learning typically present it as resulting from (ie caused by) 'experience'. Thus Borger and Seaborne (1966) provide a 'provision definition' of an instance of learning as 'any more or less permanent change of behaviour which is the result of experience' (p.14). Bass and Vaughan (1966) define learning as 'a relatively permanent change in behaviour that occurs as a result of practice or experience'.
 
Back to the paper.
 
 
Note 4
 
Noddings (1984) considers Ryle's approach in relation to competence-based education, but rejects it as an appropriate way forward. However, she appears to equate Ryle's philosophical behaviourism with the manner in which the competence approach engages in the analysis of competence into a set of observable performances. I do not think this is a valid interpretation of Ryle's analysis, which I take to be more concerned to provide a critique of the notion of an inner realm of existence in which can be located the purported causal mechanisms such as competences, and so, by extension, the supposed process of learning. Ryle's philosophical analysis provides the necessary conceptual groundclearing to enable the more significant analysis of the social character of learning and of competence, which this paper goes on to address.
 
Back to the paper.
 

 
Note 5
 
We should also note the use of the term reading when applied to undergraduate study. This usage seems to be confined to older universities; it would seem pretentious if used in relation to study at a newer university. This is yet another mark of the displacement by the language of learning.
 
Back to the paper.
 

 
Note 6
 
The term train is usually taken as a transitive verb, applied more to occupants of lower status positions (or animals). However, we might also not phrases such as training to be and training for , especially in respect of high status occupations eg She is training to be a priest/ lawyer/ doctor , and She's training for the priesthood/ the bar . This reinforces Wittgenstein's exhortation to look and see ,sage rather prescribe meaning through definition.
 
 
Back to the paper.
 
 
Note 7
 
The term prediction is here used in its weak (task verb) sense, trying to foretell accurately, rather than strong (success verb) sense; the essential uncertainty of the future limits the proper use of the latter to retrospective judgement.
 
 
Back to the paper
 
 
Note 8
 
I have taken the phrase 'confidence trick' from Donaldson and Farquar's discussion (1988) of the workings of the banking system, which depends upon the confidence of depositors to maintain levels of liquidity that permit investment of deposits. The cognate term 'trust' is, of course, central to the ideas of social capital theory (Coleman, 1988; Putnam, 1993; Fukuyama, 1995).
 
 
Back to the paper.
 
 
Note 9
 
Toulmin uses the term warrant-using in a similar way as Ryle's use of the term licensing infernce ; Gergen does not make explicit reference to Toulmin's work, but seems to have drawn upon it, and extended it into the analysis of how warranting operates socially, not just logically.
 
 
Back to the paper.
 
 
Note 10
 
The origins of the skills agenda in higher education can be traced back to a joint report by the National Advisory Board for Public Sector Eucation and the University Grants Committee in 1984 (NAB/ UGC, 1984). This asserted that: "The abilities most valued in industrial, commercial and professional life as well as in public and social administration are the transferable intellectual and social skills". Significantly, the report of an interview-based survey of 138 employers of graduates, published in 1985, presents a discussion of employers perceptions of what they value (in a chapter titled A degree and what else?) in terms of the desirable qualities graduates as a group tend to have in comparison with their contemporaries (Roizen and Jepson, 1985). The language of skills in notable by its absence in this empirical study, conducted at the same time as NAB and UGC claimed that employers valued transferable skills.
 
Back to the paper.
 
 
Note 11
 
The term performance-based education is frequently used in the USA as synonymous, or almost synonymous, with competence-based education . (see Short, 1984; Tuxworth, 1989).
 
Back to the paper
 
 
Note 12
 
We have many terms to apply to behaviour which is inappropriate to a particular social identity: eg gaffe, gaucheness, impertinence, effrontery, impudence, insolence, misconduct, mutiny. This particularly applies in terms of social settings intrinsically embued with contested power relations, not least those of management.
 
Back to the paper.
 
 
References