Lexeme vs. Concept
The object of philosophical inquiry are concepts. Concepts are the crystallization of a particular problematic within a particular range of conceptual coordinates (conceptual plane of immanence). Words, on the other hands are just collections of sounds or letters. The only people who are afraid of words are fascists, GLAAD, and PETA. (That is to say, the only people who are afraid of words are fascists).
The key thing here is that a particular word and a particular concept do not always correspond, and in the history of philosophy the same or related word has been used to designate very different or even opposite concepts. A perfect example are the words “subject” and “object.” In contemporary philosophical discourse, these two words correspond more or less with their contemporary grammatical sense: the subject is the one that does, the object is the one done to. But in Medieval philosophy, these two words have precisely the opposite sense, because the basic principle of relations between entities was not, as it is today, their identities, but rather, their predicative relationship. Thus in Medieval philosophy, “subject” almost always means “the one done to,” what today we would call the “object.”
The confusion between word and concept is a frequent source of conceptual error among undergraduates and other easily excitable thinkers. Take one course on Derrida, and suddenly every use of the words “presence,” “truth,” and “speech” are like red flags. Take one queer theory course and anything “heterosexual” or “heteronormative” is to be dismissed and derided. We are also easily tempted to judge particular thinkers “incompatible” through this very error, as the above example shows. If you go purely by his critique of “extension,” it’s easy to assume that Hume and Spinoza are entirely incompatible; except that Hume is explicitly critiquing the word “extension” as used by Descartes, while the same word means something very different in Spinoza.
Ironically, Hume falls prey to this very error himself in his discussion of Spinoza (a different part of the Treatise than the critique of extension). In his critique of Spinoza, Hume attacks Spinoza’s use of “substance,” “mode,” and “attribute,” suggesting strongly (and derisively) that Spinoza simply indulges in the same metaphysical errors and inconsistencies so common among the Scholastics when they use these terms. But what Hume fails to realize (and probably had no possible way of knowing) is that when Spinoza uses these classical metaphysical terms, he uses them in a very different way than the Scholastics do. This is because, even though both the Scholastics and Spinoza ultimately trace the use of these words to Aristotle, they derive their concepts from two very different traditions: “substance” means something entirely distinct in the history of Latin Christian philosophy than it does in the history of Arabic-Judaeic philosophy, on which Spinoza draws in his use of the term.
In short, words are not concepts. And only the latter are proper objects of philosophical critique.