With Locke the question comes into even greater prominence. On what conditions can I have knowledge? How can I be certified that my ideas--the subjective images in _my_ mind--have a reference to something objective and real? Locke's answer is, not unnaturally perhaps, somewhat prolix, and wanting in fundamental precision of principles. After dismissing the view that, even before experience, there are certain common ideas spontaneously and by original endowment present in all human beings, he goes on to show how we can sufficiently account for the ideas we actually find by supposing in us an almost unlimited power of joining and disjoining, of comparing, relating, and unifying the various elementary 'ideas' which make their way into the empty chambers of our mind by the senses. As to the source, the channel, and the nature of these sense-ideas, Locke is obscure and apparently inconsistent: though clearly it should be all important to know how an idea can be caused by, or spring from, a material thing. When in his fourth book he comes to the question of what is the reality, or the _meaning_ of our ideas, he does not really get beyond a few--rather dubiously reasoned-out--conclusions that, although strictly we cannot go beyond 'the present testimony of our senses employed about the particular objects that do affect them,' we may for practical purposes allow a good deal to the presumptions of general probability. But Locke had also begun to criticise our ideas, in his account of their formation out of the 'simple ideas'-- (which neither Locke nor any other atomist of mind has succeeded in making clear)--which the several senses give, and by observing or reflecting on what goes on or is present in _our_ minds, _we_ 'form,' he says, various ideas. In a style of discussion which is on the borderland between vulgar and philosophical analysis-- (never quite false, but nearly always inadequate, because it almost invariably assumes what it ostensibly proposes to explain,) Locke tells us how we get one idea by 'enlarging,' another by 'repeating,' as we please, the bounteous data of the touch and sight. But amongst the compounds there are some of more disputable origin. There are some--e.g. ideas of punishable acts or legalised states--which are 'voluntary collections of ideas put together in the mind independent from any original patterns in nature,' These, though entirely subjective, are entirely real, because they only serve as patterns by which we may judge or designate things so and so. It is worse with the idea of power, which we only 'collect' or 'infer,' and that not from matter, where it is invisible, but only in a clear light when we consider God and spirits. Still worse, perhaps, is it with the idea of substance, which is a 'collection' of simple ideas with the 'supposition' of an 'incomprehensible' something in which the collection 'subsists.'