We are coming to the end of section A, my friends.
When the fundamental concepts of art theory are accounted for on an a priori basis and therefore are valid independent of any experience this does not of course mean that they could be deduced in a purely rational way independent of any experience. As the fundamental artistic problems … already represent the contrast between form and volume, time and space, presupposed a priori in a specifically sensible sphere, so the understanding of these fundamental problems presupposes an empirical (visual) manifestation. (55)
Oh, good: this solves my “manifest vs. a priori” difficulty from last time. What Panofsky is saying is that, while the basic concepts are legitimate a priori — we know they hold good before looking at any individual object — they are not totally abstract, conceived prior to the basic manifestation. Rather, they were developed to fit the absolutely minimal conditions of an experience of visual art: if we’re not looking at some manifestation that deals with haptic and optic, surface and depth, and fusing and splitting, then we are simply not looking at a work of art (or else not looking aesthetically) at all. A manifestation of these concepts is just what visual art is.
What applies to the understanding of the fundamental problems applies to a still greater extent to the understanding of the individual problems; artistic problems are only understandable by means of their solutions, that is, the works of art themselves. And if those concepts that formulate the problems claim a priori validity this claim does not mean that they can be found in a priori form but only that they can be legitimated a priori. (55)
Just as you shouldn’t look directly for the fundamental problems in the individual artwork itself, you shouldn’t try to refute the fundamental concepts by showing that they can’t be found in the work in question. The goal of empirical research is not to attack the prevailing theory, but to work with it and extend it (see below) in order to get at more of whatever is interesting about the object. (Important, maybe, for the purposes of my ACLA paper: “manifestation” for Panofsky is prior to “form.” We get our concepts straight about the pure manifestation before we advance to looking more closely at the form.)
So, again, art theory maintains a sort of benevolent autocratic relation to art history, helping it improve itself but not susceptible to coup d’état from mere researchers: “the system of concepts for a science of art, insofar as it is discovered and developed a posteriori, will continue the results of empirical research, but its existence, insofar as it is valid a priori, is independent from these results and cannot come into conflict with them” (55). Again, the potential problem with this, from a contemporary scientific point of view, is that in practice it’s unfalsifiable (wonder if Panofsky ever ran into Karl Popper?). But let’s grant Panofsky this theoretical autonomy for the nonce and see where he goes with it, how he imagines it interacting with the course of empirical research:
A new discovery or a new observation can constitute proof either that one of the already established and a priori legitimized artistic problems in a specific case has been solved differently from what had been supposed so far or that one of the questions formulated by existing concepts had in this case been answered wrongly (for the intention of art theory is nothing other than formulating questions). (55)
What the stability of the fundamental concepts will give us is a ground from which to change our understanding of particular artworks. If we hold the basic problems steady, then empirical work (discoveries, analysis, etc.) can allow us to see that a work has solved one or more of those problems differently (from some other work, or from how we used to think it solved them).
For example, if against all odds a pictorial landscape from the fifth century BC should be discovered, an amendment of art historical understanding must be made, which in itself would not affect the theoretically constructed system of art concepts. Or again a new discovery or new observation can prove that a specific artistic problem has not yet been recognized as such, that the questions formulated in an earlier conceptual system were directed away from an important characteristic of certain objects. For example, when an architectural monument of a completely unknown form comes to light or when an already-known work of art unveils a new aspect of a new problem to the eyes of a scholar who studies it, the relevant problem needs to be shaped to a formula and brought into relation to those problems already known, which constitutes an extension but not a destabilization of the art-theoretically generated system of concepts.
This is close, as I read it, to Alexander Nehamas’ use of the concept of interpretation: if it comes to light that James Joyce actually wrote Augustine’s City of God, say, then we need to rearrange our entire understanding of literary history around that fact. But the new discovery, transformative as it is, doesn’t fundamentally change (“destabilize”) the way we think about literature, it just means we have to shift everything around and reconsider (“extend”) the links between various different texts and authors. (This is all in "Writer, Text, Work, Author," I believe.)
At this point I think we’ve all so fully internalized the Kuhn/Bachelard/Foucault idea of “paradigm shifts” or “epistemic breaks” (and the attendant uses and abuses of its “revolutionary” implications) that this is actually kind of hard to get, and certainly once gotten it’s easy to argue with. But it's also very attractive to anyone halfway interested in reconciling theoretical thinking with historical research. For now, I think it's enough to get clear about what Panofsky's vision was, before assessing whether there are contradictions in it (there probably are) or whether it might have some special utility for literary scholars today (it probably could).
That’s the end of part A. Deep breath. I am very much looking forward to an Erwinless weekend. Next week, maybe, I’ll start on B.