Berkeley on Abstract General Ideas

Berkeley on Abstract General Ideas* ROBERT A. IMLAY THE MAINPURPOSEOF THIS PAPERis to show that Berkeley's solution to a problem for which other philosophers had introduced abstract general ideas leaves his ontology open to serious attack. A subsidiary purpose is to show how he might have accepted abstract general ideas as a solution to the problem. But first of all what was the problem? The problem was to explain how we can know a property to belong to all particular triangles, for example, when we have used only one particular triangle in the demonstration. A standard reply, and one given by Locke, was that we have an abstract general idea of triangularity which stands for the oblique, the rectangle, the equilateral, the equicrural and the scalenon triangles without being any one of them. As a result, to demonstrate that a property belongs to such an idea is, ipso facto, to demonstrate that it belongs to all the different kinds of particular triangle which it represents.t Berkeley, as it is well known, rejected Locke's solution to the problem. He denied that anyone could have an abstract general idea of triangularity which stands for the oblique, the rectangle, the equilateral, the equicrural and the scalenon triangles without being any one of them. Indeed, he argued that in Locke's description of the idea that philosopher attributes logically incompatible properties to it.2 A sympathetic reading of Locke, however, will not generate this conclusion. As one commentator has pointed out, Locke never said even in the passage Berkeley quotes against him that the kind of abstract general idea in question is composed of inconsistent parts. Rather, he said that it is composed of some parts of several inconsistent ideas. And there is absolutely no reason to think he believed that the parts themselves were inconsistent. 3 On the other hand, Locke did say that such an abstract general idea was something imperfect that could not exist,a Berkeley's rejoinder is that if it could not exist it must be a logical impossibility, * I should like to thank Messrs. Peter Danielson and John Martin for their extremely helpful criticisms of an earlier draft of this paper. John Locke, An Essay Concerning Human Understanding (New York, 1959), ed. Alexander Campbell Fraser, IV, vii, 9. 2 Works of George Berkeley, ~ls. A. A. Luce and T. E. Jessop (Edinburgh, 1949), 9 vols., Hereafter cited as Works, II, 32-33. s R. I. Aaron, lohn Locke (Oxford, 1955), 2nd edition, pp. 196-197. 4 Ibid., p. 196. [321] 322 HISTORY OF PHILOSOPHY since that and that alone presumably is beyond God's power to create. 5 Moreover, he agrees with Locke that it could not exist. But why did Berkeley believe that it could not exist? The answer is not hard to find. If you hold, as both Berkeley and Locke did, that triangularity as such does not exist, although the different kinds of particular triangle do exist, then it will begin to look as though triangularity is identical with the different kinds of particular triangle. As the result of such an identity, moreover, the view that an idea of triangularity is not reducible to an idea of one of the kinds of particular triangle may seem to amount to the view that an idea is not reducible to itself. And this is a manifest contradiction. 6 It is one thing, however, to deny that an abstract general idea could exist. It is another thing to explain what they were originally introduced to explain. Berkeley, of course, realized this and tried to give an alternative account of how we can know a property to belong to all particular triangles when we have used only one particular triangle in the demonstration. He retained the concept of representation to be found in the abstractionist account but assigned the representative function to a particular idea as opposed to an abstract general one. Thus, he argued that the idea of an isosceles rectangular triangle whose sides are of a determinate length may be extended to all other rectilinear triangles of any size whatsoever. And the reason why one can do this, according to Berkeley, is that neither...