Everything that exists is a particular thing. Frisky, Tiger, Felix, and Snowball are in the external world, but there is no cat. At the same time, most of our terms are general rather than particular. How, Locke asks in Book III, chapter iii, do we get these general terms? Since words refer to ideas, general terms, naturally, refer to general ideas. General ideas are produced through a process of abstraction. We take our ideas of Frisky, Tiger, Felix, and Snowball, and we attend to what is similar in all of these, discarding what is different. From what is common to all of them (fur, soft, meow, arched shape, etc.), we form a new idea. This is our abstract general idea of cat (also sometimes referred to as a "partial idea"), and we attach to it the general name "cat." Locke does not give much of an argument for this claim. He simply asks that we all inspect our own ideas and see whether there is any difference between our idea of man and our ideas of "Peter," "Paul," etc., other than that the one is more general and the other more specific. It is patently obvious, he claims, that our idea of man is just these other ideas minus what differs between them. Locke gives three reasons why we form abstract general ideas and their correlative terms. The first reason is that it would simply be too hard to remember a different word for every particular thing that exists. In addition, having a different word for everything that exists would seriously thwart the communicative function of language, since different people all have experience of different particular things. Lastly, it would subvert the aim of science, because science tries to make general claims about the way things are. Chapters iv and v are discussions of general names of mixed modes and of simple ideas.
Locke does not really give much of an argument for his theory of abstract general ideas, and so it is difficult to come up with objections to the argument itself. On the face of it, though, there does seem to be a problem with his account of how we get general ideas. When attending to the ideas of Frisky, Snowball etc., is there really anything exactly the same about them? Cats, like all particular things, are individual in all of their properties. If the mind is sifting through particular ideas looking for matches between them, it will be hard pressed to find any. Locke, however, could probably give a satisfactory response to this worry. The mind attends to things that are similar enough. The mind is not a precision machine and has lots of room for judgment calls, and it is therefore capable of attending to similarities while ignoring the particular features that distinguish even these.