Dear scholars of literary studies,
you can't ignore the text. You have proven that in years over years of scholarly practice, paying attention to the smallest details of the text (I remember discussing the function of the punctuation mark in William Carlos Williams' "The Red Wheelbarrow"). But you have been blind to 50 % of the text. The human perception of a textual factum works in two ways; you consciously only acknowledge one of them. When we look at a text we can decide (!) to see the semantic or the formal structure, to decipher to "hidden meaning" or to apply an image-like perspective and see the text in its form.
It is not irrelevant in which type you set your words, even if the majority of you strongly believes that. It is not "just the outer appearance". Such a thing as "typographic dispositivs" (See Wehde's transfer of Foucault's terminology for (typo-)graphic purposes) do exist. I don't think that the type and the composition is accidentally set or that the producer "probably wasn't aware of what he was doing." If the producer – who is often enough a literary scholar – cares about the signals he sends or the meaning he shares, then he should know what he is saying with all of his utterances. A lack of awareness is not appropriate for scholarly research and that applies not just to the areas of research a scholar is randomly interested in. Every engagement with every text should have the janiform ambigramness of the object in mind. There are two sides of the story. You cannot state that form equals content and therefore only see the content. It is not a logical utterance as in a = b > b = a. It is rather a dynamic dichotomy with interplays and mutual reactions which constantly changes as society changes.
There are rules and norms that consolidate themselves in society and therefore in the perception of every subject, meaning: us. I therefore believe in the motivation of every sign, not in the arbitrariness. I therefore disagree with Ferdinand de Saussure's utterance that all signs are arbitrary and conventional and point into the direction of "social semiotics" – a movement fostered by Kress/van Leeuwen who set the motivation and the convention as the prerequisite of a slightly changed theory of signs. The consciously stand in the tradition of semioticians like Roland Barthes or the previously named de Saussure, but instead of simply adopt and apply what they said forty years ago, they are trying to assimilate their theorems to current issues and needs. I don't think that these issues and needs are actually that current. They existed forty years ago as they do now but the theory needs to evolve and finally grow out of its infancy.
From a believer in text.