Speech vs. Writing Cagematch

Derrida is setting up a writing vs. speech opposition in which he claims that people privilege (value? esteem? trust?) the latter over the former. This is clearly in anticipation of a point at which he will demonstrate that writing has been getting the shaft all these years and deserves just as much respect as speech. Fair enough, except he's unclear about the axis along which writing and speech are being contrasted. I can come up with three ways in which Derrida might be claiming that the one has been erroneously valued above the other.

1. People in Western society have historically trusted the spoken word more than the written word. The average guy on the street is inclined to believe someone when they're speaking face to face, but to treat a written document with a extra measure of skepticism.
2. Linguists in particular have paid too much attention to speech as an object of study and have not given writing its due.
3. Linguists and philosophers have claimed that writing isn't a semiotic system unto itself–it is just a way of recording speech, and speech is the only true semiotic system in natural language. Writing is just notes on a staff, and not music.

Axis (1) is a broad historical claim, so broad in fact that I don't know what to make of it. Axis (2) is also a historical claim, though of a much narrower scope. Axis (3) is more a philosophical than a historical claim. (In the sense that you could believe in the semiotic autonomy of writing even if everyone else did too.)

I'm not sure if the most charitable reading picks one of these or some combination of the three. However, if this point remains muddy, Derrida's oh-no-no-it's-writing-that-matters counterpunch is going to be something of a wet dud.