Educational Explanations. Christopher Winch

self-evident fact should pose a problem for the very possibility of EER (Clark 2011) as is sometimes claimed. Upon examination, it appears that there is a clear distinction to be made between a social phenomenon and value-laden practices: ‘As we have seen education is a good. Social phenomena are not… Second “education” being value-laden does not denote a range of empirical phenomena which is its proper domain as “social science” does’ (Clark 2011, p. 50). Consequently, according to Clark, educational practices cannot be studied as social phenomena. If this is so, then EER is clearly impossible, since it is the very attempt to do just that. However, the only plausible way in which one could arrive at such a conclusion would be if one held that social phenomena are by definition value-free. But only some conceptions of social science would be committed to such a view, notably those within the behaviourist tradition, or perhaps a form of hermeneutics that sees values excited within the imagination of the researcher in rather the same way as Hume envisaged ethical and aesthetic judgment to arise (see for example Hume [1757] (2008)). This is a way of conceptualising social phenomena which is rejected in this study, and there is little evidence that Clark himself subscribes to it either. By their nature, educational practices are value-laden. They are also phenomena in the world, albeit not ‘raw’ physical phenomena. An educational researcher may not identify with the values that are expressed in an educational practice, but s/he is entititled to explore what values are expressed in a practice whether or not s/he agrees with them.¹⁸

Needless to say, there are preconditions for any successful investigation of educational practices if one wishes to understand them. These are not always easy to obtain. First, the researcher has to recognise that what he regards as a good may not be by the participants in the researched practice. He cannot judge the value of the practice by his own standards if doing so entails that he rejects its character as an educational practice. Second, he will need, if this is the case, to have the patience and imaginative sympathy to try and understand what the values expressed and goods aimed for in such practices actually are, and how they relate to other values held within the society.¹⁹ Third, he will need to be able to look back on his own values and to see if what he has learned about the practice being studied tells him something about himself, the values that he espouses and the practices and the society in which he originates. This last requirement might seem strange. However, a researcher who is unable to develop in this way is one who is also unlikely to develop the sense of detachment and imaginative sympathy required by the first two conditions. In particular, such a researcher will lack the ability to make crucial connections between the attitudes and practices of his own society and that of the society or practices that he is trying to understand. None of these desiderata can be assumed to be easy to achieve. Nevertheless they are essential to successful EER.

      CONCLUSION

      We have looked at six different ways in which the value of EER can be called into question.

      It is inadequately conceptualised.

      1 It uses inappropriate methods and flawed explanations.

      2 It has difficulties generalising beyond particular contexts and in establishing the continuing identity of particular phenomena.

      3 Most of the findings of EER are false.

      4 Common sense can do just as good, if not better a job than EER.

      5 EER is irrelevant to value-laden practices such as education.

      We have seen that although 1–4 are often true of EER, they do not constitute sufficient grounds for abandoning it. The incoherence of objection 5 shows that we need more than ever to ensure that EER is of good quality. 6. depends for its plausibility on the view that one cannot systematically research value-laden practices, a view which is highly implausible.

      Notes

      1 1 E.g. Barrow 1976; Carr, W. 2006.

      2 2 Clark’s arguments are somewhat different and will be dealt with later in this chapter.

      3 3 I have used the term ‘normative activities’ following Baker and Hacker to capture the practices which allow for application of terms like ‘correct’ and ‘incorrect’ to them. As many of these are informal and often fluid, the term ‘rule-governed’ with its connotations of something codified as an account of how they work, is slightly misleading.

      4 4 A further complication is that there may be interlinguistic conceptual variation which is difficult to appreciate if the researcher has neither the linguistic resources nor the philosophical training to be aware of such nuances. See Chapters 9 and 10 for fuller treatment.

      5 5 We will look in more detail in Chapter 4 at the criteria for good explanations, but we can be reasonably confident at this stage that they will not need to meet the standards set by Barrow in order to be viable.

      6 6 By ‘theory’ I mean a body of supposed knowledge organised and connected, usually derived from systematic observation, description, analysis and evaluation. This is not an exhaustive account of what is meant by ‘theory’ but will do for the kinds of empirical theories with which we will primarily be concerned. See Winch, C. 2010, Ch. 6, pp. 111–113 for more on this.

      7 7 If P can be explicated as (p and q and r…) and r is false then it follows that not-(p and q and r …) even though p, q and other propositional components of P might be true.

      8 8 One might say that Popperian falsificationism takes this point to an extreme degree, if one maintains that the falsification of a single observation report deductively implies the falsehood of the hypothesis that generated the observation statement or even the laws underpinning the hypothesis. As we shall see, however, there are legitimate ways of blocking such drastic inferences.

      9 9 We need to be much more careful when talking about probabilistic confirmation as for example Carnap (1962) and Salmon (1975) (both available in Achinstein ed. 1983a) have shown. It is possible for example that a research hypothesis h and an auxiliary hypothesis a be separately refuted and the conjunction handa be confirmed (Salmon 1975, p. 107ff.). Given that EER using quantitative techniques such as various forms of statistical inference is engaged in probabilistic confirmation and refutation, these paradoxes are potentially of huge importance (see Ch. 8).

      10 10 It goes without saying that all points about contradictories and contraries apply to propositions as well as theories.

      11 11 Two formally incompatible explanatory theories P and Q may be jointly false in C1. However, P may be true in C2. If this is so, since the truth of P entails the falsity of Q (they are incompatible in any context) then they are contradictories in C2. They have not become compatible through a change in context.

      12 12 For examples of the latter, see both MacKay (2006) and Grant (2011).

      13 13 There is much more on this topic in Chapter 8.

      14 14