Generally we regard both Locke and Berkeley as "nominalists". But correctly speaking, they should be called "nominalistic realists", because they restrict what exists to what is particular, and yet admit that predications extending beyond some original paradigm specimens may be valid. But Berkeley blames Locke bitterly for "abstract or general ideas", which are, according to Locke, the bonds between particular things that exist and the names they are to be ranked under. His attack on Locke results from two theoretical bases, one of which is his ontology and the other is his theory of "signification". On the latter we should pay attention to his pragmatic point of view. Though Berkeley rejects "abstract ideas" in the Lockean sense, he doesn't necessarily deny our ability to abstract. I suppose the abstraction level he rejects is not "abstraction" but "generalization" to be exact. And, with Berkeley, the epistemological condition that precedes the ability to abstract within the limits of his permission and "general ideas" in the Berkeleian sense, is our ability to perceive "likeness" between particulars, in my opinion.