Inadequate Conceptualisation
Objection 1, that EER relies on inadequate conceptualisation, indicates a major problem that affects not only EER but the social sciences more generally. It is also a fundamental problem that we will be revisiting in Chapter 4 when we consider more closely the concept of an adequate explanation. If EER is conducted in order to enable us to understand educational practices and the concepts of which they are expressive, then we need a conceptual apparatus that enables us to understand and explain why those involved in education act and think as they do. In order for this to be possible, it is necessary to understand their reasons for acting as well as those factors that impel them to act. In order to do either of these things the researcher needs to, as far as he possibly can, see the world through the conceptual framework that the research subject (teacher, parent, pupil, administrator, policymaker, public) sees it. In order to do this the researcher needs to understand the concepts relevant to reasons for acting. This in turn entails a deep engagement with the normative activities (Baker and Hacker 19853) that characterise educational practices. Concepts cannot be understood apart from the normative context of their employment. It follows that we have to understand that context in its own terms, using the concepts employed by those within the practice. Doing this involves investigating and understanding those concepts and relating them to the purposes of the investigation under way. This is why Peter Winch in The Idea of a Social Science (Winch, P. 1958) proposed that the social sciences had greater affinities with philosophy than they did with the natural sciences. Like philosophy, social science was to be to a large extent a conceptual investigation. Winch’s suggestion was initially received with outrage as an attempt to send the social sciences back to a pre-scientific era. However, the intervening sixty years have only served to show the prescience of his claim and EER in particular bears many of the marks that result from not paying sufficient attention to Winch’s arguments.
In addition to this hermeneutic problem it is also too often the case that the explanatory concepts invoked by educational researchers to structure their own investigations are themselves inadequately framed, leading to further confusion about what it is exactly that they are supposed to be investigating. Thus, framing investigations about language use and educational achievement or about skills without adequate conceptions of what ‘verbal deficit’ might mean or without an investigation of what is meant by ‘know-how’4 can also lead to poorly designed empirical investigation. In the latter case, not only is the research design compromised, but also the investigation of the concepts of the research subjects.
So those who complain about the poor quality of conceptualisation to be found in EER often have plenty to complain about. But it is important to note that their complaints do not relate to the possibility of ever finding anything out through EER, but rather to the inadequacy of many of the efforts made so far. If conducting an activity is more valuable than not conducting it, the fact that it is currently poorly conducted cannot be a good reason for not conducting it at all, but rather a reason for conducting it properly, even if that means a more cautious and less ambitious approach is needed in the future.
It is generally speaking true that EER is too often (although by no means always) inadequately conceptualised. We will have cause to look at some important examples in subsequent chapters. However, this is symptomatic of a larger problem in the social sciences, that much of the empirical work carried out is inadequately conceptualised. This can mean one of three things:
1 That the value position from which the research is being conducted is not sufficiently defined and is insufficiently self-reflexive. It is inevitable that educational and social science researchers more generally come to their work with ‘baggage’ in the form of presuppositions, prejudices and values of their own which may place barriers in the way of their seeing the practices that they are studying in the way that they are seen by the agents themselves. A failure to acknowledge this and to take account of it can lead to misunderstanding of the point of the practices being researched.
2 That there are inadequate framing explanatory frameworks. For example, neoclassical economics may employ a concept of rationality that is inadequate to explain human behaviour in the relevant contexts. Here the problem is, at bottom, at least partly philosophical since an explanatorily adequate concept of rationality requires philosophical development not empirical research, although the use of examples is very often a good way of testing the adequacy of a conceptual framework. A related problem in educational research would be to oversimplify the motivational factors underlying parental choice of schools in, for example, India (Gurney 2017).
3 A related but different difficulty is that an inadequate investigation has been made of the conceptual framework and epistemological presuppositions of those being investigated by social scientists. They thus run the risk of misconstruing the phenomena which they seek to describe. This appears to be the case in, for example, Evans-Pritchard’s account of the poison oracle in Zande society, where it is assimilated to a form of inadequate scientific reasoning (Evans-Pritchard 1936). A similar phenomenon in education would be when researchers misunderstand parental attitudes to schooling, failing to realise that the parents being researched set greater store by different educational processes (Brice-Heath 1983). In this kind of case, insufficient attention is paid to the ways in which subjects of research perceive their practices and how those practices link up with other practices within a culture and thus what it makes sense to say within those practices (see Rhees 1970).
A fourth issue, closely related to the one just discussed, is concerned with the ways in which empirical enquiries in education are conducted. Although they are empirical, this does not mean that they should solely concern themselves with events, processes or states of affairs connected with educational practices. There is very often an important interpretive or hermeneutic role to be played in such studies if they are not to fall into the trap of oversimplifying or even misconstruing what they are studying. A good example, which we will look at more closely in Chapter 10, concerns the use of terminology connected with professional know-how and vocational education, where a combination of linguistic ‘false friends’ and a form of linguistic imperialism can lead to an inability to see what is important in the practices under scrutiny.
These pitfalls bear out a point that Peter Winch made in, 1958, that many of the important theoretical issues that have been raised in relation to the social sciences ‘belong to philosophy rather than social science’ (Winch, P. 1958, p. 17). This continues to be a problem in empirical educational research, but does not obviate the need for it.
Inappropriate Methods and Flawed Explanations
Many of the complaints about EER do not focus on the conceptual issues, but rather on whether or not research designs are sufficiently technically robust to sustain the conclusions of their investigations. A particular complaint is that there is a tendency to overgeneralise. Thus Robin Barrow writes on the inadvisability of teachers using EER: ‘…for the teacher should never act on a generalisation, since for all he knows the children he faces, or some other aspect of his situation, are exceptions to the generalisation’ (Barrow 1984, p. 186). Although Barrow points to a real difficulty for EER, he cannot use this claim to assert its impossibility. The statement above is itself an educational generalisation and, if true, refutes the claim that it makes, since at least one educational generalisation is then true. If false it entails that some educational generalisations are true. EER would indeed have difficulties if any generalisation were forbidden. Fortunately that need