These then are the four main strands in Aristotle's thought concerning the law, or in other words, the four elements he might have distinguished in his conception of law. The analysis I have attempted seems to me to reflect both Aristotle's view of the complex nature of law and also what he would look upon as the different grounds for its validity. I think that the several elements in his doctrine are fundamentally independent of one another, and similarly that they do not compete with one another since they embody answers to different questions concerning law. Also the recurrent theme of my own comments has been the assumption that 'law' is a complex term, comprising in its application a number of different definitions concerning rules and validity, authority and obligation, sources of law, and the like. In my opinion it is the merit of Aristotle's conception of law that he appears to recognise the multiple meaning of the word ‘law’ and, accordingly, the need for a multiple definition.

Some words enter the language with an uncommon aptitude both for uniting things already observed but formerly severed by separate terms and for fostering the recognition of things unnoticed before. Indeed, they often unite things that ought still to be left discrete; and even among those properly united, clarity may require the acknowledgment of many distinctions. I shall here consider such a term and the various kinds of things to which it can and cannot (in the same sense) refer: projection.

Just before the second world war, in a paper read to the British Association, Morris Ginsberg talked about the failure of social philosophy and the social sciences to work together in the universities ‘toward the rational ordering of society’. Some time after the war Alexander Macbeath complained to British sociologists of his own vain search for a social philosopher who could teach in a course on public administration. Then a few years later A. E. Teale told an inter-professional conference at Keele that people who teach and train teachers, those who train social workers of all kinds, were disappointed when philosophers professed themselves unable to help those who had to ‘equip students with the skill to change prevailing moral attitudes and standards’.

The Question of punishment and its justification has been a major preoccupation in recent philosophy of law. The reasons for this growing concern are not difficult to discover. Both philosophers and jurists have become increasingly sceptical of traditional theories of legal punishment. Each of these inherited theories was designed to establish criteria for the recognition and appraisal of punishment as a legal institution. However, alternative theories emphasised different and often conflicting criteria. Some theories emphasised moral desert and retribution, while others appealed to the social utilities of punishment. Failure to reconcile these diverse philosophical viewpoints, through sustained debate over two centuries, has been a major source of intellectual dissatisfaction.

Mr A. Quinton has attempted to show that the Kantian doctrine of the necessary unity of space rests ultimately on very general contingencies (Philosophy, 1962). More recently, Mr Swinburne has tried to argue the same point for time; and thus he asserts that ‘Kant's unity of time is no more an unalterable necessity of thought than his unity of space’ (Analysis, June, 1965). I wish to defend Kant against both charges, by showing that the charming stories which Quinton and Swinburne tell, in order to show that there are conceivable circumstances in which space and time would not be unitary, do not succeed in showing this at all.

The publishers in this country of The Perceptual Process by A. Campbell Garnett (reviewed in PHILOSOPHY, Vol. XLI, No. 158, October 1966, pp. 371–3) are George Allen and Unwin Ltd., not the University of Wisconsin Press.