One more time, transferable skills don't exist ... (and what we should do about it).
by Len Holmes, The Business School, University of North London (at time of presentation)
Presented at Higher Education for Capability conference, 'Embedding Key Skills Across the Curriculum', Nene College, Northampton, 27th February 1998
This paper argues that the concept of 'transferable skills' (or 'key skills' etc), and the methodologies by which these are
purportedly identified, are fundamentally flawed. The 'skills project', ie the attempt to identify such entities is shown to have
failed. However, the underlying concern which gave rise to this flawed project may be reframed, to generate an alternative
approach which the paper goes on to address. The notion of a socially constructed graduate identity is presented as the key
concept in such reframing.
If at first you don't succeed ...
The legendary advice to Robert the Bruce, to try, try and try again, may be useful in many areas of human life, but what do we say when millions of pounds are spent pursuing a particular project, over many years, without proven practical outcome? Surely, in any intellectual endeavour our response should be to question the way that the project is framed, to ask whether we have misunderstood the problem we seek to resolve, to widen our perspective on the task. That is what I shall argue in respect of the 'skills project' in higher education. Whether called 'transferable skills', 'personal competences', 'capabilities, 'core skills' or (as in the Dearing Report) 'key skills', these are the purported desired generic outcomes of higher education. Attempts to identify and analyse them have figured in various curriculum reform initiatives at departmental, institutional and at sector levels over the past decade or so. The largest such initiative, Enterprise in Higher Education, involved over 60 institutions with funding of up to £1 million each for five year projects. Yet despite all this effort, there is still no evidence that the skills project has yielded anything but localised schemes of dubious transferable value.
In this paper, I shall present arguments against the validity of the notion of 'transferable skills'. These arguments are based on both the failure of the skills project to deliver practical results, and on the conceptual and theoretical flaws that are inherent in the very concept itself. I will point out the methodological flaws in the project. Whilst the problems with the project have been pointed out before, I will argue not only that transferable skills do not exist, but also that they cannot exist in the way that the skills project assumes.
However, there is a way forward out of the ruins of the project. I shall outline what, in my view, should now be done to accomplish what I take to be the underlying goal, that is, to improve the prospects of students gaining entry to the social arenas to which they aspire to the mutual benefit of themselves, the stakeholders in those social arenas, higher education institutions, and the wider society.
The skills project: much progress, little achievement
There is no doubting that there has been considerable work undertaken on what we might call the 'skills project', since the report on 'Higher Education and the Needs of Society', jointly published by the National Advisory Board for Public Sector Higher Education and the University Grants Committee (NAB/ UGC, 1984). The report stated 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."
A follow-up document continued this theme:
"The personal or non-academic skills of students, which higher education is expected to develop, include the general communication, problem-solving, teamwork and inter-personal skills required in employment." (NAB, 1986: 3)
There soon followed various attempts to elaborate this limited statement, particularly with the funding impetus of Enterprise in Higher Education and in the context of the developing 'competence movement' notably in respect of National Vocational Qualifications and of management development. Various institutional and departmental projects produced various listings of the 'skills' which were deemed to be necessary or desirable. Wider studies include that of Smith et al. (1989) which resulted in a list of 20 'transferable employment skills', and the action research project at Sheffield University which produced a 'model' of 108 'skills' organised into eight categories within four 'zones' (Allen, 1993). The Quality in Higher Education project at the University of Central England came up with a set of 'generic or core skills' which, it was claimed, employers and academics agreed should be demonstrated by graduates. These include: "willingness to learn, team work, problem solving and a range of personal attributes including commitment, energy, self-motivation, self-management, reliability, co-operation, flexibility and adaptability, analytic ability, logical argument and ability to summarise key issues." (Harvey and Green, 1994:7)
The Association of Graduate Recruiters report adopted the term 'self-reliance skills' which was presented as a set of twelve 'career management and effective learning skills' (AGR, 1995). The Dearing Committee reported that:
"we believe that four skills are key to the future success of graduates whatever they intend to do in later life. These four are:
(Dearing, 1997, 9.18)
We now end up with a plethora of different list and 'models' with little indication of how any particular list relates to others. Whilst we might say that there has been much progress in terms of activity, there is no evidence that substantial achievement has been made. The practical question of which list should be adopted, because it is the most valid, remains unanswered. If one aim was to provide employers with assistance in recruiting graduates, by providing better information about applicants than that indicated solely by degree classification and subject studied, then there is clear failure. The diversity of information has multiplied so as to make it unusable. If the aim is to improve educational provision by clarifying (and/ or improving aims and objectives and developing better forms of assessment, the result has been to raise unsurmountable problems of validity. In terms of enabling prospective students to make more informed choice, we must surely conclude that the variety of schemes has resulted in just the opposite.
Studying skills: faulty methodology ...
The plethora of lists of purported 'skills' is understandable when we consider the methods of enquiry adopted in their development. Some lists appear to be wholly the product of the semantic elaboration of some notion of the sort of graduate that 'we' (ie a particular department, faculty or institution, or the higher education system) are attempting to 'produce'. Whilst the outcomes of 'brainstorming' exercises may give a sense of achievement to the staff involved, their conceptual validity must surely be rated low. It is surprising that academics who would challenge such a crude approach to enquiry in their own disciplines can be so uncritical in their acceptance of it in relation to this vital area.
Other lists are the outcomes of what appears to be a more sophisticated approach, usually accorded the honorific of 'research'. Yet even here we find that the methodology poses serious doubts on the validity of the 'findings'. Typically, the approach is to present a list of skills to a target audience and to seek feedback in the form of rating or ranking. Thus Harvey et al report (1992) the use of a questionnaire to employers, asking them to rate the importance of fifteen qualities that graduates might be expected to have. Similarly, Kemp and Seagraves (1995) used questionnaires listing skills to ask both teaching staff and students about how these were part of the teaching programmes and assessment methods. Smith et al. (1989) send questionnaires to recent graduates, with a list of skills. From such methods, conclusions are reached about the importance of particular skills and the 'restructuring' of courses 'with skills at the forefront'.
Of course, what such enquiries presuppose is that the 'skills' terms being used have the same, unequivocal, denotational meaning to all respondents, to the researchers, and to readers of the reports. Such a presupposition is unsustainable, as the study of 'managerial skills language' by Hirsh and Bevan (1988) shows. Nor would some form of 'glossary of terms' help, as the terms used to explain particular skills terms would themselves be subject to multiple interpretations.
Even if we were to allow that the meanings of the skills terms were stable between researchers, respondents and readers, the methodology of the skills project would be problematic. Asking employers, teaching staff, or students to rate or rank skills, or to say anything at all about them, does not guarantee that such ratings, rankings, or comments are correct. All they tell us is the opinions about the purported objects of enquiry. Valuable though such opinions may be, they cannot be taken as descriptions of those objects. Indeed, it is not uncommon to use the same methodology to assess the 'correctness' of respondents' views or opinions. It is quite possible that the researchers use such an approach to assess students' knowledge and understanding in a particular subject, yet presumably they would not base their own understanding of the subject on the answers provided by the students! ... and a flawed concept
No matter how valid a methodology might be, if the concept being studied is itself flawed we can only end up with invalid findings. Yet it is clear that the concept of 'transferable skills' is rarely subject to critical scrutiny by protagonists for the skills project. Indeed, Griffin (1994) refers to the 'aura of untouchability' of the concept of transferability:
"That transfer takes place was and is so powerful an assumption as to be deemed beyond discussion: what we think or can do just does transfer from one situation to another." (p.134)
Despite a number of published critical analyses of the concept (Bridges, 1992; Griffin, 1994; Gubbay, 1994; Holmes, 1995), the proponents of the skills project have noticeably avoided responding to such criticism. When the Times Higher Education Supplement published a critique by Hyland (1997), the only response was in letters attacking the author rather than a reply to the argument presented.
I shall not rehearse the well-reasoned arguments presented in the texts cited above except to reiterate Bridges' (op cit.) challenge that without a theory of social domains, the notion of transferable skills is neither intelligible nor applicable.
"We need to ask perhaps what would make one social context different from another to the extent that it might constitute a challenge to the transfer of skill, so that the fact that people showed a capacity to apply something learned in one context to another, demonstrated that they has mastered a transferable skill."
Similarly, Wolf (1991) argues for 'the primacy of context' in our understanding of 'core skills', the term cognate with 'transferable skills' as used in National Vocational Qualifications.
"... these skills are by definition inseparable from the contexts in which they are developed and displayed, and ...they only make sense (or, rather, the same sense) to those who have the same recognition and understanding of those contexts." (p. 194)
Proponents of the skills project might argue that 'transferable skills' is a theoretical construct and should not be taken necessarily to refer to some empirically real objects. I suggest then that we ask what kind of theoretical work this construct is supposed to perform in the context of educational theory. Presumably it is intended to provide an explanation of the links between performance of individuals in different contexts. That is, we wish to explain how performance by an individual whilst undertaking education (including various forms of assessment, both formative and summative) can provide the basis for anticipating performance in post-graduation employment (and also, perhaps, postgraduate study). The attempted explanation is that there are certain individual qualities or attributes which result in performance, that is, they cause or are used in performance. The lists of purported qualities tend to be a mixture of characteristics of some inner self (eg 'flexibility', 'self-reliance'), which suggest a causal link, and 'skills' which an individual has and uses, suggesting an instrumental link. I have caricatured this as a 'homunculus with a toolkit' model, whereby education seeks to develop the 'inner self' and provide it with a set of tools for use in various tasks and situation (communicating, problem-solving, etc) (Holmes, 1995, 1996).
However, when we ask how such qualities are identified the skills project seems to point to the very performances which the qualities are supposed to explain. How do we identify that someone has 'problem-solving skills'? Presumably by observing that they solve problems. But there are no empirical objects which are denoted by the term 'problems'. Instead, there are various situations which, under particular circumstances, we call 'problems'. To call a situation a 'problem' is to assert that it is undesirable, not in itself, but relative to some more desired state of affairs, and that it is possible through some action to change the current state of affairs towards that more desired state. 'Problem solving' is an abstract term we use to refer to the action of attempting such change, not an internal instrument used in action. That is, 'problem solving' is a classificatory, not explanatory, term. This applies equally to the other varied terms that comprise the various lists of purported 'transferable skills'. They are abstracted ways of talking about performance.
If transferable skills were explanatory terms we would be faced with the further theoretical task of explaining how a person decides which skills should be used, and how, in the different contexts to which they are deemed to be transferable. One attempt at this is to suppose some further 'meta-skill' (or set of them) which we might call 'transfer skills' (Blagg, et al, 1993). But how then does an individual decide that a situation requires the application of such a skill to transfer a particular transferable skill to the situation? Presumably there must be an infinite series of such 'meta-skills'! Such a concept is surely meaningless.
Now you see it, now you don't!
Up to this point, I have discussed the notion of 'skills' as if any particular individual either has or does not have each of the purported entities. This is the way that much of the discussion of skills tends to be presented. Employers say they need this or that skill (Green, et al, 1992); students say they have or have not received help in developing a skill (Kemp & Seagrave, 1995). Higher education should deliver the required skills to students. However, one of the key tasks that skills are supposed to perform is to differentiate between the states achieved by any particular student, ie before, during and at the end of their course. A second task is to differentiate between graduates, so that an employer can carry out their recruitment more effectively. These tasks require that the concept of a skill allows for qualitative differences, 'more', 'less', 'better', 'worse', etc. This raises the question of how we measure points on the qualitative scale. Can someone have a zero-level of a skill? Is there some finite top level?
More problematic is the question of the stability of a skill. For the notion to be of any use, it must be the case that an individual's possession of a transferable skill, to whatever level, is relatively stable. Once developed to a particular level, it would stay at around that level within certain parameters (eg of time). Changes in quality of performance in specific situations that are proximate in time might be understood in terms of other factors, such as motivation or stress, but ceteris paribus, we should expect a high level of consistency. Otherwise, the concept of skill cannot explain performance.
Yet we know that this does not represent the reality of social life. Members of a work team often disagree on the best way of dealing with a situation. Does this indicate an absence or low level of 'problem-solving skill' for one of the parties involved? If so, which? Or does it indicate a low level or absence of 'communication' skill? Or 'teamworking' skill? Does it indicate 'leadership' or its absence? Even if we could adjudicate on this situation, what do we do when, moments later, the team move on to another matter which is dealt with effectively, harmoniously, and speedily? Have the skills suddenly returned?
It is clear that the concept of 'transferable skills' requires us to accept a whole weight of implicit assumptions which themselves require explanation. Far from being a simple concept, its explanatory power is dissolved by the plethora of explananda it conjures up. The most sound conclusion that suggests itself at this point, using Occam's razor, is to reject the notion. In terms of explaining the relationship between education and graduate performance, 'transferable skills' not only do not exist, they cannot exist!
Rescuing the skills project: talking of persons
Of course, to say that some supposed entity does not exist does not imply the denial of the phenomena which the notion of that entity is intended to explain. During the eighteenth century, European science used the idea of phlogiston, a 'fiery substance' within combustible material, to explain how combustion took place, ie as the escape of phlogiston. Joseph Priestly coined the term 'dephlogisticated air' to refer to what we now call oxygen. The whole theory of combustion based on the notion of phlogiston was overthrown by that proposed by Lavoisier, that combustion involved the chemical combination of oxygen with the material. This radical overthrow of one attempted explanation by another did not change combustion itself; wood, paper, cloth, etc still burnt! It was, however, a major advance for the nascent science of chemistry.
What, then, are we seeking to understand? What is the empirical reality for which we need to develop a mode of analysis and explanation, with practical consequences, but which the notion of transferable skills has proven to be flawed? It is, surely, that individual human beings, persons, attempt with greater or lesser success to engage in activity in different contexts in such a way that we say that there is some significant similarity about that activity in those different contexts. There are two key elements in this: persons who engage in activity, and the language or discourse used about the activity.
The conventional skills project appears to be based mainly in a cognitive psychology paradigm (or, perhaps, a pastiche of it!). Attempts at explaining human activity involve the hypothesising of 'modules' of cognitive processing within the individual's mental structures. Thus problem-solving is undertaken by some problem-solving mechanism. In contrast to this, the discursive psychology paradigm that has developed over the past couple of decades provides a different basis for explaining human activity:
"Psychology is the study of how, why, when and where active people use signs for all kinds of purposes: thinking, planning, anticipating, judging, lying, doing calculations, forming consortia, denigrating rivals, presenting representations of the past, and so on." (Harré, 1995: 157-8)
This approach takes seriously the socially embedded nature of all human activity, including the social-embeddedness of individual persons, ie their identities. It also takes seriously the 'discursive turn' in the social and human sciences, language as a social practice through which the social world is constructed and not merely described. There is now a considerable body of theoretical and research literature adopting this approach (eg Potter and Wetherell, 1987; Edwards and Potter, 1992 ; Shotter, 1993; Harré and Gillett, 1994; Harré and Stearns, 1995; Smith, Harré and Van Langenhove, 1995), which surely merits attention in the quest to develop the social relevance of higher education. The approach certainly appears to me to be consonant with the educational approaches found in situated learning theory (Lave and Wenger, 1991) and in attempts to utilise the insights from the work of Vygotsky (Daniels, 1993).
I have attempted elsewhere (Holmes, 1995, 1996) to outline an alternative to the conventional skills project, utilising these developments in psychology and educational theory. Essentially, this would focus on the processes through which students become graduates, that is how they achieve a graduate identity. By 'graduate identity' I am referring not just to the point of the award of a degree, but the later accomplishment of an affirmed social position in the arenas where being a graduate is socially significant, primarily in employment and/ or in postgraduate study. An 'identity' is not one of a fixed essence but a negotiated outcome of the individual's claim on the right to the desired social position and the evaluations made by gatekeepers to such a position. It involves an identity project (Harré, 1983), extending over time and space, in which the outcome is not determined by internal attributes but subject to various communicative processes of claim (by self) and affirmation or contestation (by others).
In this graduate identity project, the discursive warrants (Gergen, 1989) adopted for claim and affirmation/ disconfirmation are critical. That is, what counts as the demonstration that someone is indeed worthy of employment as a graduate (or worthy of postgraduate study) is constructed through particular linguistic forms which have become conventional ways of articulating what is required or desired. These conventions of warrant (Gergen, op cit.) have connotational force, not denotational reference. This is surely evidenced by the way that employers tend to talk about 'the sort of person' they seek and the wide variety of different ways of expressing what seem to be similar qualities. Green et al. report that one employer stated that they wanted graduates who were able 'to make an early impression on the organisation' (p.2); this does not seem to refer to some 'early-impression-making' skill!
Towards new investigations
Taking such an approach would suggest that what we should be doing is to investigate the processes by which the graduate identity formation process takes place. One aspect of this is the way that graduates accomplish entry to employing organisations and the way they achieve the transition from graduate recruit to that of an established position in which the correctness of the recruitment decision is not in doubt. How do they represent and account for themselves in this trajectory? How do they try to 'make an early impression'? How do they handle challenges to their right to be considered 'capable'? How do different graduates do this in different circumstances? We should examine what happens when a graduate fails to achieve the identity, ie fails to gain a position as a graduate or loses such a position - a 'spoiled identity', as Goffman (1963) would have put it. We should examine the graduate identity formation process in relation to other identity formation processes, particularly those of ethnicity, gender, and class.
We should also examine how the 'gatekeepers' to the graduate identity engage in evaluating whether a particular individual should be accorded such recognition. Such gatekeepers include, of course, graduate recruiters but are by no means confined to such a group. They include the managers of the graduate, and also 'significant others' such as work colleagues, subordinates, other staff, and more senior personnel. Again, a strong hint of the importance of the evaluations of other staff is provided, again, in the report by Green et al., who state that 'other qualities' employers would like to see include 'humility', a reflection of the anecdotally reported claim that graduates have too high an opinion of themselves!
Such investigations should take account of the way that the language in which people engage in real social interaction is highly complex, and rich in connotation, not limited to the precise use of clearly defined denotational terms. Pre-structured questionnaires are just not up to this task. Rather, we need to examine primarily the extended discourse that occurs in natural settings and open-ended interviews. Developments in approaches to discourse analysis, coupled with the advances in software for computer-based analysis of qualitative data, provide much more sophisticated resources for tackling such a project
Into the curriculum
This discussion has taken us onto wider terrain than that in which most teaching staff in higher education are engaged, with their practical concern for design and delivery of programmes of education for real-life students. This has been necessitated by the problems inherent in the conventional skills project, its failure to deliver its promise because of the conceptual and methodological flaws. However, there is, I believe, a welcome twist to the story. By adopting the revised approach suggested, we would achieve a richer, grounded set of accounts of how real graduates move from being students to a desired position, whether as an graduate in employment or as a postgraduate student. These accounts could then serve as exemplars for undergraduates, in planning their own journeys to achieving their graduate identity. Rather than hearing restricted, abstract accounts of the need to develop this or that skill, students will be able to hear how their predecessors have engaged in the real-life process of achieving their desired status.
Moreover, teaching staff will also be in a better position to understand what happens to their students upon graduation. Rather than being faced with the limited information provided by first destination reports, or abstracted listing of skills, teachers will be able to consider how their own teaching and assessment programmes and practices relate to the tasks faced by students in their post-graduation lives. Such programmes and practices can then be enhanced and modified, where necessary, to help students become better prepared to enter and progress within the social arenas to which they aspire. Indeed, it may well be the case that existing programmes and practices are, in fact, well suited for this and do not need significant change in the way that proselytes for the skills agenda claim. Whatever the case, any programme of change in the practices of teaching staff in higher education is likely to succeed only if those staff perceive the proposed change to be valid, in conceptual and practical terms.