(I’ve already run through all the pictures of Panofsky Sr. to be found on the internet apparently, so here’s one of his son Wolfgang, who is a renowned physicist in his own right.)
So Panofsky holds that empirical researchers use two orders of concepts — sensible and suggestive/demonstrative — in their practice, but that these exist on a different layer than the fundamental theoretical concepts he himself formulates:
To repeat, one has to differentiate between two layers within the standard art-historical concepts: a superficial layer of solely suggestive or demonstrative concepts, which only define and describe the sensible qualities of artworks, and a deeper layer of concepts that actually characterize style, which interpret those sensible qualities along the lines of stylistic criteria in that they interpret certain complexes of sensible qualities as the manifestation of principles of design and those principles of design as the differentiation of a uniform stylistic principle. (60)
Both of these kinds of concepts, remember ("sensible" and "suggestive/demonstrative/[first-order] stylistic") are still pragmatically generated art-historical concepts, and not properly theoretical concepts of the kind that Panofsky wants to propose and perfect.
But what gives this guy Panofsky, and art theorists more generally, the right to impose their more "fundamental" concepts on researchers? “Now, it is worth asking what grants art history the right to operate interpretatively in this way, as any interpretation, if it is not simply capricious, demands the existence of solid and legitimized standards of definition to which objects of interpretation can be related” (60-61). (Note Panofsky’s neo-Kantian concern with “right,” with justification, grounding and legitimation — the kind of questions Habermas revives in the 60s and 70s.) “The answer, which should perhaps now be obvious, is the imperative of those artistic problems formulated in the fundamental concepts of art theory. Those fundamental problems [i.e. haptic vs. optic etc.] represent the standards of legitimation to which the sensible qualities of artworks can be related. These problems help observation navigate while collating the sensible qualities within certain complexes of manifestation and interpret them according to stylistic criteria. Only by being oriented towards them do we have the ability to conduct successfully this enterprise of collating and interpreting” (61)
So the point of fundamental theoretical concepts is to “orient” empirical research: this is what differentiates Panofsky’s position from that of scientific positivism, which would be mere, pure data collection, without the necessary second step of “collating and interpreting.” This helps us arrive at a practical definition of what the critic/theorist actually does, in relation to what the historian/researcher does: she does “stylistic characterization,” as opposed to a “sensible classification.” (By the way, I don’t think Panofsky thinks these necessarily have to be different people, or that one can’t move from gathering and classifying data to interpreting and characterizing it; but he does think you can only wear one hat at a time.)
This act of “characterization” is pretty close to what Kant called “judgment” (?). It means not just description of the artwork, however exhaustive, but an assessment of whether it is satisfying, which Panofsky reformulates as a question about whether it solves fundamental problems:
The stylistic characterization “within this work of art lies one of the fundamental artistic problems — or even all the fundamental artistic problems — which have been solved in this or that sense.” We only get to this point after first recording a complex of manifestations (for example, the extremely “textural” mode of surface characterization or the extremely unifying mode of color design), with the help of demonstrative concepts (doughiness or the addition of grey in all local colors) and then by interpreting this “texturality” or unity through the prism of the fundamental problems. Only by this process can the observation of sensible qualities … reach the sphere of problems that actually characterizes style and whose specific task is to relate the sensible qualities of works of art to artistic problems. (61)
For Panofsky, understanding of art, when it’s done correctly, is a process: the theorist/critic looks at what the historian/researcher has gathered, and at what she has already observed about it (with those pragmatic “suggestive or demonstrative concepts” that emerge out of the research experience), and goes on to relate it to the fundamental problems, to what is agreed to already exist within the sphere of art history. And this entails saying how this particular work has “solved” one or all of the fundamental artistic problems. (There may be a direct line of descent from this neo-Kantian definition of the critic to the self-understanding of a Clement Greenberg, or after him a Michael Fried.)
This brings us back to the distinction suggested earlier between characterization and classification: “Only from this perspective is it clear why the double precondition of stylistic characterization — that in certain sensibly perceivable complexes of manifestation an artistic principle of design become apparent and that the principles of design themselves to be governed by a single underlying stylistic principle — actually exists in its own right. Only if we can take these complexes of manifestation as solutions to artistic problems can we understand them as the manifestations of principles of design, that is, as testimonies for a certain position on fundamental and individual artistic problems. Only if we can detect within these fundamental and individual problems the shaping of a single ur-problem can we grasp a single stylistic principle in operation within and above all principles of design, that is, a certain position on these ur-problems” (61-62) (I’ll leave it for now, but I’m uncomfortable with this talk of “ur-problems.” For one thing, EP makes a weird move, within a single sentence, from “a single ur-problem” to plural “ur-problems” — is this a translation glitch? Or an inconsistency within Panofsky’s system? Does he really think it’s productive to think of all artworks as solving a “single ur-problem”? If so, what is it? For modernism, maybe it’s autonomy; this seems to be what Bourdieu thinks.)
But, yet again stipulating Panofsky’s terms:
only from this perspective is it clear why we have the right to deduce stylistic differences from the differences of sensible qualities and a stylistic identity or relation from a correspondence or similarity of sensible qualities. We can do this because (and on the condition that) the difference of sensible qualities provides testimony for differing positions on one and the same artistic problem (to which a unified artistic subject could only respond in the form of one position); we are allowed to do this because (and on the condition that) the correspondence or similarity of the sensible qualities provides testimony for an identical or related position on one and the same artistic problem (to which differing or, rather, different artistic subjects could only respond with different positions). (62)
Gaaah. This is an insanely dense and defensive sentence (I mistyped it about four times just now, getting lost in the parallel clauses). So, it seems that what Panofsky thinks is that the role of art theory is to mark individual artworks (and maybe, on another level, artists?) as positions within a larger space of art as a whole. (I may be overreading this passage based on a false transfer from some of Bourdieu’s writings, but just run with me here.) A description of an artwork’s “position” will need to take into account historical facts about it (when it was made, why it was made, the biographical details of its maker if any are known), but also, more importantly, its position within the space of art history, that is of all other artworks. And this latter space can only be known if we orient ourselves according to fundamental theoretical concepts. If we want to have a hope of understanding artworks with reference to a universal, then something like Panofsky’s system will allow us to get from a thick empirical description (a classification, and even a preliminary characterization) of an individual object to the whole set of artworks ever produced in any culture. And though Panofsky opposes the hastier kinds of “short circuit” thinking about art, his scheme does allow us to get from specifics to the general (Art) without tracing everything with endless cultural-historical research.
But Panofsky doesn’t just make this as a suggestion for art historical research, he makes it as a claim about art and style: “every stylistic characterization (no matter whether it wants to or not, no matter whether it knows it or not) necessarily relates — both in choice and organization of its observations and above all in its creation of a conceptual system — to the contents of that sphere of problems that I tried to outline in the first part of this essay. A stylistic characterization that denies this relationship to the fundamental problems either denies it incorrectly or is incorrectly termed a stylistic characterization” (62). (Hmm. Here he makes the twentieth-century theoretical move par excellence — “no matter whether it wants to or not, no matter whether it knows it or not” — taken over from Marx, Freud and Nietzsche by basically everybody who wants to make a categorical statement without arguing for it. I am skeptical about this descriptive, as opposed to prescriptive, aspect of Panofsky’s argument, but this is neither the time nor the place (well, I guess it actually is the place) to talk about that.)
But then he goes back to the methodological level: “Yet the art historian qua art historian cannot understand these problems and their interdependencies. He who solely deals with empirical manifestations is in the opposite situation from the art theoretician, who aims to develop fundamental art-theoretical concepts and to differentiate art-theoretical specialist concepts. The art historian … can define the empirical manifestation but not the solution of the artistic problems, while the art theoretician only supplies the solutions of artistic problems without being himself able to understand their empirical manifestations in actuality” (62). So the historian and the theoretician need each other: that’s the take-away from Panofsky’s essay (this section of it anyway; no, he’s still not finished). With a rhetorical emphasis on the historian needing the theoretician, to be sure (this is against Dorner, presumably):
Empirical manifestations, with which art-historical study in its strict sense is concerned, do not show the fundamental problems at all; the problems are not apparent within the phenomena but are required behind the phenomena and therefore are only understandable by a form of investigation that is not art history but art theory. It is only the nonsensible conceptual speculation of art theory that can establish the goals by which the work of empirical researchers must navigate its every step. Only art theory can understand those artistic problems that are assumed in art-historical practice and can provide those artistic problems that are assumed in art-historical practice and can provide the fundamental and specialist concepts for this enterprise. (63)
So theory is the concept factory. Researchers mine the ore of art, give it to the theorists to make tools out of, and then the theorists give it back to the researchers to use to find more art with.