In our argument so far we have tried to show some kind of logical connection between certain fundamental values and premisses about human nature on the one hand and certain kinds of educational practices on the other. The precise character of the logical connections which we have drawn may deserve closer scrutiny in a more technical philosophical work than this one pretends to be. They are certainly not all of a piece. In some cases we have suggested that there might be quite a tight logical entailment, such that a commitment to principle A necessarily implied a commitment to practice X. In other cases the connection was looser - X might be one form of practice but perhaps not the only one which reflected or was consistent with the principle. Though we have from time to time referred to evidence of the actual consequences of one kind of practice or another, it has not in general depended upon such evidence. In slightly more technical terms our argument has been very largely a priori in character. Inevitably an argument of this kind will be an incomplete argument. There will be a number of questions of direct relevance to the issue at hand which remain unanswered. In particular, the question 'What are the actual effects of mixed ability grouping on children's social relations, educational careers, or classroom learning?' is not one which can be answered on the basis of a priori reasoning. This requires experience, observation and evidence. We can examine philosophically the reasons for seeking certain intended or hoped for effects, or the desirability or otherwise of other effects when known. We can even

observe logical features of practices of a kind which provide good (though perhaps not sufficient) reason to predict what their effects will be. But in the end philosophers overreach themselves if they try to prescribe educational (or any other) practices without consideration of the actual and not just the intended or anticipated outcomes of those practices. The arguments we have presented so far need, then, to be considered alongside empirical evidence available in other sources relating to the effects of different forms of grouping and different styles of group teaching.