A bit more on visual goings on in Wittgenstein's later work

Actually, this is just a bit more about some of the philosophical consequences of the ubiquity of linguistic awareness thesis (ULA) that propped up the previous post.
Consider, for example, Eugen Fischer's Wittgensteinian claim "Pictures rather than propositions determine most of our philosophical convictions" (Philosophical Delusion and its Therapy: Outline of a Philosophical Revolution, p.21). ULA implies that this cannot be right (so it also looks (sic) as if the envisaged revolution can only be an abortive one). For only a linguistic interpretation of a picture can do the work of forming convictions by interacting with our stock of beliefs. And, a picture in itself cannot determine such an interpretation (something that Wittgenstein seems to be committed to anyway). Here, of course, we will want to say that pictures have some causal role. But, it appears that we cannot define it non-arbitrarily at any useful level of generality. This holds for the general relationship between our beliefs and the world if we think of that relationship as being primarily non-linguistic.
Fischer tends to explain 'pictures' in terms of what might best be called linguistic models - so the visual aspect turns out to be redundant and the relations that are important are inferential.
My main concern here, I guess, is that, frequently, the philosophical use of visual language might itself be the product of thinking that requires therapeutic intervention - something Rorty cottoned on to in Philosophy and the mirror of nature (though he also slipped into pictorial mode all too often).