process described by Wittgenstein when told about the claim that ‘nothing will make us leave the paradise that Cantor created for us’, referring to the discovery of the transfinite cardinals, to which Wittgenstein is said to have replied, ‘I wouldn’t dream of driving anyone from paradise. I would just show them that it wasn’t paradise and they would leave of their own accord.’ (Diamond 1976). Another example would be through the demonstration of a formal or geometrical proof.18 It can also occur through the kind of ‘paradigm shift’ in a scientific community described by Kuhn (1962), whereby cumulating dissatisfaction may lead to a radical shift in perspective which may involve conceptual upheaval. An example used by Peter Winch (2015) concerns Socrates’s exchange with Polus in the dialogue Gorgias. Here, Winch argues, Socrates is not trying to persuade Polus to change his mind, but to get him to understand what the state of his mind really is (op. cit., pp. 116–117), namely that Polus must really believe that it is better to suffer than to do wrong, despite what Polus sincerely says. It does not involve introducing new facts but by changing Polus’s perspective on facts already known.
Persuasion may also involve more practical considerations, such an appeal to collective or individual interests. At the other end of the spectrum it may involve the display of superior power of a cultural or technological kind (consider the role of Western technology in undermining religious belief both in the West and in countries colonised by Europeans). Finally it may rest on the threat or use of force.19
THE ROLE OF CONCEPTIONS OF RATIONALITY IN EDUCATIONAL RESEARCH
If persuasion of these various kinds is established in ways in which beliefs are changed and concepts undergo shifts, what room is there for rationality in such processes? Are we doomed to doubt the rationality of both research subject and their researchers? We have, it should be acknowledged, held fast to the concept of truth, but has this been done at the price of giving up on rationality? There are paradigms of rationality which are more or less dominant in different cultures. In ours, deductive reasoning holds particular sway. However, it is by no means the only way of thinking about rationality. Hollis (in Hollis and Lukes 1982), for example, argued that something like basic logical presuppositions would need to hold in all cultures in order for them to be intelligible at all. Such a position rests, in turn, upon a conceptual scheme in which particulars and persons are categorial primitives (Strawson 1961).
But we also know that logical contradictions can be tolerated, not just in ‘primitive’ societies but also in our own (Evans-Pritchard 1936). Does this undermine the claim that there is an underlying logic to all societies? Not necessarily. That contradictions may be tolerated in certain contexts does not entail that ignoring them has no consequences. They tend to exist in those circumstances in which the benefits of tolerating them outweigh those of dispensing with them, for practical reasons such as the maintenance of the stability of a way of life. One could say that although humanity has a common constitutive rationality, various forms of practical rationality (to do with the determination of ends and means to those ends) may suspend the formalised canons of constitutive rationality in some circumstances. If such contradictions are exposed, it is the researcher’s task to first investigate whether they really are contradictions and, if they are, to enquire why and to what extent they are tolerated and, perhaps, to seek analogues in their own belief system.
We may also wish to note the role of authority and authoritative argument in determining reasons. It is possible therefore to distinguish between the constitutive rationality appealed to by writers such as Hollis,20 the theoretical rationality associated with formal systems, the practical rationality associated with the determination of ends and appropriate means to achieve those ends and critical rationality or the stance towards authoritative determination and the scope of the use of arguments from authority adopted in different cultures and different contexts.
More needs to be said about critical rationality in this context. Arguments from authority presuppose that the premise citing the authority be not disputed.21 Argument and justification from authority are indispensable for any intellectual division of labour. However, the prestige and scope of arguments from authority may well differ from epoch to epoch and from culture to culture. We may, therefore, possibly distinguish between cultures and societies in terms of their stance towards arguments and justifications from authority and thus to critical rationality.
Thus Evans-Pritchard (1936), in his discussion of the Zande poison oracle, although noting the care the Zande took to ensure that the poison substance (benge) was not tampered with, were not inclined to dispute the pronouncements of the oracle, even when they were not apparently borne out by events. Zande tended to attribute such an outcome, not to a mistake on the part of the oracle, but as maladministration on the part of those consulting the oracle. Evans-Pritchard’s interpretation of the oracle practice involved seeing it as a form of primitive and ineffective scientific enquiry and thus, by the standards of modern academic scientific method, lacking in rigour and effectiveness.
There are two possible (and complementary) responses to this. The first is to point out that Evans-Pritchard may have been drawing the wrong analogy and would have been better drawing an affinity between Zande oracular practices and religious practices such as prayer for intercession to be found in our own society. Then consulting an oracle might not seem so outlandish to an educated European. The second point is that there is a gap between our idealised view of scientific practice and what actually happens within science practices. Kuhn (1962), Lakatos (1970) and others have drawn attention to scientists’ ability to tolerate inconsistencies and even contradictions for practical purposes22 and for everyday scientific practice to depend heavily on the authority of a research paradigm and the principal personal representatives of that paradigm as being decisive in determining whether to accept or reject the result of an empirical enquiry.23 We should not therefore deceive ourselves as to our readiness to be critically rational and sceptical of arguments from authority on all occasions, even in those practices where critical rationality is apparently at a premium, while at the same time being prepared to be robust in maintaining critical rationality in our own practices. We are susceptible to the toleration of contradictions and dubious authorities just as other people are.
A good example of where debates about such issues can lead to widespread confusion relates to the role of literacy in our understanding of rationality. There is a long tradition dating at least from the work of Lévy-Bruhl (1910) which suggests that one of the marks of a ‘primitive mind’ is exclusively oral culture in which constitutive rationality, let alone critical rationality, cannot properly exist. Such a case has been argued for in a less extreme manner by for example Goody and Watt (1963), Olson (1977) and even Stubbs (1980). It has been disputed by Labov (1969), Finnegan (1973) and Winch, C. (1983, 1990). Debates about the enabling nature of literacy show us how intimately related are conceptual and empirical enquiries within the broad field of educational research. Without some clarification of what is meant by ‘rationality’, ‘argument’, ‘symbol’, ‘context’ ‘surveyability’ and ‘genre’ just to take some important terms, it is difficult to make much progress in understanding the terms of an investigation into the enabling powers of literacy, let alone conduct empirical investigations into these enabling powers.
CAN A CRITERIAL CONCEPTION OF TRUTH BE DEFENDED?
In this last part of the chapter I wish to consider some possible problems for the view that has been developed so far.
The first is that criteria change over time (they are revisable) and so what was taken to be true at one time may not be at a later one. We need to remind ourselves that to say that p is true is not to say that it corresponds with reality, but that the ways in which its truth should be determined have been satisfied. This will, in many cases, involve ‘looking at reality’,24 but it does not follow that it means investigating whether or not it corresponds with reality, for reasons we have already noted. Our responsibility is to ensure that our criteria are as robust as they can possibly be,
