Consider the following two sentences.
Bertrand Russell said that "the essential business of language is to assert and deny facts".
Bertrand Russell believed that language is basically assertive.
The first can be established merely by providing a reference (e.g., "Russell 1961: x", and the associated entry in the bibliography). I have claimed that Russell has written an eleven-word sentence. (As I pointed out yesterday, that claim is almost true: my memory got ten of those eleven words right. I'll address this issue on Friday.) The second, however, requires an argument, in which the first may play a role. The fact that Russell wrote a particular sentence will serve as evidence for the larger and much more interesting fact that he held a particular belief.
Notice that my claims about Russell's writing and Russell's mind themselves express beliefs—my beliefs. I believe that Russell wrote those words and I believe that he held that belief. Through my writing, I am hoping to persuade you to hold those same beliefs (what I believe about Russell, not what Russell believed about language). Now, I don't expect you to trust me blindly, certainly not when I'm writing for scholarly purposes, and that is why I will provide the reference for the quotation. You may or may not go back to my source to check my work, but the presence of the reference itself moves us beyond merely "blind" trust. After all, you can now assume that I at least looked at the page in question, and you can assume that one of my readers (perhaps one of my reviewers) has checked or will at some point check it. There's a fact in the world that corresponds to my claim and to the belief I want you to form, and I've told you exactly where you can see it for yourself.
Russell's state of mind in the early 1920s when he wrote those words is more difficult to establish, to be sure. But it is the presumption of scholarly prose that such states of mind are real and knowable. There is a kind of "fact of the matter" about what he meant. I'm not here talking about a general theory of mind, i.e., a philosophical position about the knowability of other people's state of mind. I'm talking about the knowability of the beliefs, opinions, ideas of other scholars, whose work we cite. We are naive, common-sense realists about the words they have written. And somewhat more sophisticated hermeneutic optimists about their meaning. That is, while we will always grant that there can be different interpretations, and while we may even grant that some of these disagreements are ultimately unresolvable (perhaps because of "the play of différance", perhaps because of "the death of the author"), we don't think that there is an entirely arbitrary relationship between the words a scholar puts on a page and the meaning that the scholar intended. Moreover, as scholars, we regularly invoke the intended meaning, committing both the original author and our fellow scholars to it. It's possible to get Russell's beliefs wrong, and being able to do so is an important part of being a philosopher.
I'll pick up the thread on Friday. Let me conclude today by marking two important limiting cases of this argument. First, when I say that there is a "fact of the matter" about what a scholar means, I do not mean that this fact is "empirical" and to be determined by the application of a "scientific method". I'm with Richard Biernacki on this point. Second, I want to stress again that this is neither a theory of mind nor a theory of language, nor even a theory of writing. It is a presumption about scholarly writing. It does not, for example, apply to the work (and perhaps not even the mind) of Gertrude Stein.