Michael Maranda



Michael Maranda
originally published in New Literary History 29:3 (1998)

A statement of fact: “James Jarves, a nineteenth century collector, sold the Madonna with the Pomegranate (Madonna with Child), a painting by Sandro Botticelli, to the Yale Art Gallery in 1870.”

   In this apparently authoritative art historical assertion, I have made what could be called a factual error. The painting in question was not painted by Botticelli, and thus the attribution is not credible (at least, not in the terms of current art historical scholarship). It is quite clearly indicated, both on the label to be found in the Yale Art Gallery which accompanies this painting, as well as in the catalogue where this information was most recently verified (by myself — although if requested I can show you the catalogue in person), that it is attributed to “an associate of” Botticelli. [1] An honest mistake? Perhaps. If I contextualize this statement with a slightly different argument, I can turn this incorrect fact (of attribution) into a credible “historical fact.” In the process, I wish to ask implicitly what it could mean to get a fact wrong.

                * * *

   Before considering the implications of “getting the facts wrong,” I want to direct this discussion to interrogating what exactly is a fact. A good place to begin would be an essay of Carl Becker, aptly titled “What are Historical Facts?” [2] Becker writes that, perhaps because “historians feel safe when dealing with the facts … the facts of history come in the end to seem something solid … like bricks or scantlings, so that we can easily picture the historian as he stumbles about in the past, stubbing his toe on the hard facts if he doesn’t watch out” (WH 42).

   One wonders if such an image of the fact as a solid object with which the historian constructs just as solid an essay ever had much valence. Nevertheless, while the “linguistic turn” in literary studies affected greatly the scope and tenor of scholarship in that discipline, it appears that — despite the great promise of the work of Hayden White — the practice of writing history has not altered in many significant ways. [3] Perhaps not for all intents, but apparently for all purposes, the “linguistic turn” in historiograpy has not been that much of a turn at all. The usefulness of historiographical studies influenced by literary studies is in the understanding of a rhetoric that, essentially, has remained the same. If this “turn” has not aided us directly in the writing of history, it has given us tools with which to read one.

   Becker points out that there is never a simple “fact” — ”[t]hat is to say, [for example] a thousand and one lesser facts went to make up the one simple fact that Caesar crossed the Rubicon” (WH 44). A “historical fact” is the condensation of the innumerable “facts” which constitute it — all not “historical facts” unless they carry the same symbolic weight as, to retain Becker’s example for now, the “crossing” itself. Becker continues by noting that “[w]hen we really get down to the hard facts, what the historian is always dealing with is an affirmation — an affirmation of the fact that something is true” (WH 47). What enters into the historian’s realm is not the event of Caesar crossing the Rubicon. What enters is the affirmation of the truth of the phrase, “Caesar crossed the Rubicon.” In other words, a historical fact is not what has happened but what is said to have happened. [4] This affirmation occurs in conjunction with the textual source either while being written by the historian or read by the reader.
To restrict our discussion to “historical facts” would beg the question concerning the status of these lesser facts in a historical narrative. To aid in understanding the nature of these facts, we can go to the work of Frank Ankersmit and his discussion of historical reality. The core of Ankersmit’s argument consists of a claim that historians do not “analyze a previously given historical reality, but define it first. Historical reality is not a datum but a convention created by the reality effect.” [5] The “reality effect,” a term coming from Roland Barthes’s semiological literary analyses, refers to the tension between predictive and notational aspects of a text; the predictive providing the frame (in the sense of an internal armature, such as narrative) upon which the notational is draped (the details that rely upon the frame for significance). Notational details come in conflict with the predictive structure of the text because of their apparent irrelevance concerning the narrative. This tension gives these details the illusion of being the surfacing of the “real” into the text and thus they serve to signify the very act of signification. In other words, they say “I am real” and no more. [6] Similar to naturalist painting, it is the insignificant notational details, which have no meaning separate from the predictive armature, that give a historical text the illusion of truth. The illusion of truth in the historical text is, ultimately, the illusion of the past, or “the reality of the past is an effect caused by a tension in and between historical texts” (RE 140).

   If, as Ankersmit claims, the reality effect is operative in the historical text, we must determine what sorts of textual markers play the constitutive parts of the predictive narrative and the notational details. [7] The predictive structure is the narrative argument that the historian makes, what might be summarized in an abstract of the text. The notational details, however, are less obvious.

   The disciplinary justification for the condensation which forms “historical facts” is described by Dale Porter as follows: “If one accepts the historian’s belief that immersion in documentary evidence is the best path to knowledge, one can realize that the shorthand [label (for example, “historical facts” in Becker’s terminology)] refers to masses of detail that this historian is well aware of, but prefers not to mention.” [8] The “historical fact,” then, is a symbol of all the (absent) lesser facts that constitute it. [9] The lesser facts which constitute the “historical fact,” if we follow along with Porter’s description, are indexical signs which point to the research that (we assume) the historian has accomplished. We can call these lesser facts, then, referential facts as they appear to be pointing outside the historical text (to either primary or secondary sources).

   Becker’s claim that when writing of “historical facts” we are writing about words (WH 43) can be applied without much difficulty to referential facts as well — we are still writing about words, and it is only through the scrupulous transcription of quotations that we can claim we have gotten the facts right. We prove this by providing our readers with the proper references, which they can trace themselves, if they so desire. The introduction of direct references into the historical text for this purpose, as both Stephen Bann and Partner point out, can be traced to Gibbon. [10] The result of the acceptance of and subsequent demand for direct references is the transference of these references into the most factual part of the text.

   Borrowing from the discourse of the philosophy of science, in particular the work of Bruno Latour, [11] we can call these referential facts “black boxes”: “The word black[-]box is used by cyberneticians whenever a piece of machinery or a set of commands is too complex [to describe or understand]. In its place they draw a little box about which they need to know nothing but its input and output” (SA 2-3). Latour extends this metaphor from the self-conscious use by cyberneticians into a general rhetorical strategy of scientists. The result is a series of implicit footnotes, where the references to basic ideas — such as the constitution of water — are subsumed to an implicit agreement within the discipline to accept these shortcuts. Thus, any fact that grounds a scientist’s argument will be accepted as “given” and thus as “true” despite the fact that any of these “black boxes” can be opened up by critics to reveal that the (lesser) facts that construct them are not necessarily grounded themselves. The more “black boxes” that scientists utilize, the more difficult it is to dispute their arguments (WH 80). [12] The effective scientific argument itself aspires to become the “black boxes” of future scientific work. Translating this sentence to the present concern — the “historical fact” of one historian becomes a referential fact in constructing the next historian’s “historical fact.”

                * * *

To see the insignificance of these sorts of fact in action, I use an example from Michael Holly’s Past Looking: “As Theodor Mommsen predicted, Burckhardt’s studies would still ring true ‘though every sentence in them should stand in need of correction by advancing research.’” [13] On a literal level, reading this assertion of Mommsen in an aggressive manner appears to support Ankersmit’s thesis. On a pragmatic level, reading this assertion as an assertion of Holly does so as well. Looking to the indicated source of this quotation that appears to be attributed to Mommsen, however, we find the following words of James Nichols: “As the years have passed, however, the work of the specialists [philological contemporaries of Burckhardt who were more concerned about getting the facts right] has been rendered obsolete by more specialists, whereas Burckhardt has himself demonstrated Mommsen’s prediction that his works would still be read and still be true though every sentence in them should stand in need of correction by advancing research.” [l4] Holly’s use of this statement, which if Nichols had offered a citation we should be able to trace to Mommsen, is not in the least lessened by the “true source” of the quotation she uses. It is the interpretation of Burckhardt, not the fact of who interpreted him in this manner first, that should be considered relevant.

   Nevertheless, in a manner consistent with Holly’s reading of the metaleptic rhetoric of Burckhardt, we can ask whether Mommsen could be considered as the originary source of this statement. Specifically, we can look to Burckhardt himself to find him claiming, with regards to Machiavelli’s Storie Fiorentine, that “[w]e might find something to say against every line …, and yet the great and unique value of the whole would remain unaffected.” [l5]

   While quoting Burckhardt in this manner might seem to be guilty of the classical argument to authority fallacy, Latour notes that this fallacy is, in a certain sense, the very (pragmatic) basis of any scholarly discipline when speaking of such facts (SA 31-32). [l6] The very repetition of this statement (whether claiming to quote Mommsen directly or not) in a multitude of sources is its own proof. I could have used the following quotation from Ankersmit in its place: “We tend to regard a text consisting of true but irrelevant statements as ‘less true’ than a relevant text which contain some factual errors” (RE 136). As a “historical fact” (the “unreliability” of Burckhardt, on the effectiveness of an argument despite factual unreliability), the marshaling of these facts (the quotations and references that I have given) are useful only insofar as they make the necessity of my arguing this point less pressing. They give my argument authority by presenting my reader with a series of “black boxes” that are also, conveniently, “present-able.” What is even more convenient is that I have no responsibility to present them other than in this literary fashion. [l7]

   The claim I am making is that the notational details that are implicated in the reality effect in historical writing are not, as in literary works, descriptive details such as the barometer in Flaubert’s Madame Bovary. [18] Instead, they are the referential facts that are used to ground the “historical facts.” These references, while pointing to texts outside the present one, only do so through a subterfuge of false referentiality — it is as if the presence of a reference is used as a self-justification for the statement to which it is tied.

   I might have misquoted one of the sources in my Burckhardt example, or perhaps my use of them is even a misrepresentation of the argument that could be found in the source. I might have, that is, gotten one of these facts wrong. Readers could return to these sources, and dispute me on the level of facts. As soon as they offer their own reading of any of these references, however, they are opening up the black box of “historical facts,” at which point not only is the relevance of my misrepresenting the source severely restricted, but the very act of disputing my argument will also serve to legitimate my account as being worthy of recognition (as the next paragraph will discuss). The considered (and intentional) misrepresentation of sources (in moderation) might very well be one of the most useful tools that critical historians have at their discretion. Used indiscriminately, however, it can lead to an attribution of the entire argument to an unreliable narrator. Note that it is not one mistaken fact that puts the account at risk but many such mistakes — the status of the singular fact is not so relevant as the general impression of incompetence that a plethora of such “errors” would give. A rejection of this sort would be as likely to occur (if not more so) if the text being considered were rife with orthographic and grammatical mistakes.

   Fortunately, the significance of a historical account is not a series of facts. The facts that are used in a historical account must be arrayed in a narrative which is “something more” than a series of sentences. Ankersmit calls this “something more” narrative substance. [19] This narrative substance draws in and coheres a group of facts together, implicating them within a whole. More properly, it is the drawing in and cohering of a group of facts that is the narrative substance — in other words, interpretation; the modalizing and contextualizing of the facts. The relationship, or, more properly, the nonrelationship of facts to interpretations, is quite clear. “Facts only (dis)prove statements about the past. Only interpretations can (dis)prove interpretations.” [20] Through this quotation we can see that the ability to defend or reject facts will not have much direct relevance for the manner in which a historical narrative discusses and elucidates the past. This ability is not irrelevant, but neither is it particularly significant.

   If this explains where the tension of the reality effect exists within a singular historical text, it does not address the tension that Ankersmit claims exists between historical texts. Where this tension between texts might lie is suggested in his “Six Theses on Narrativist Philosophy of History.” In one of his theses he again brings up this notion of the space between historical texts in claiming that “[h]istorical insight … is only born in the space between rival interpretations and cannot be identified with any specific (set of) interpretations” (41). Historical insight, then, is being attributed here to the discipline and not to the individual text. Extending the implications of this quotation, we could even claim that a singular historical text cannot be understood as a history! It is only at the point when (at least) two conflicting accounts of the same historical event, epoch, era, or period are given can we even begin to speak of a historical argument. If there were merely one account available, there is no basis from which to dispute the narrative (nor the “facts” contained therein), and thus there would be no historical insight offered.

   Such a possibility of a singular historical account not being a history might seem to be absurd unless we understand that every historical text carries within it a metacommentary concerning the nature of the historical project. Imbedded within any scholarly composition is a critical framework that enables the very possibility of the project. This framework consists of assumptions concerning how a particular discipline functions and what truths it can legitimate at the same time that it is, if only implicitly, an argument for a particular approach toward that discipline. This argument can take the form of an overt polemic that challenges or defends the status quo or it can be submerged beneath the surface of what might be considered “art history in practice.” This is the basis from which James Elkins argues against the (mistaken) interpretation of “traditional” art histories as being somehow devoid of theoretical concerns. [21] It would be difficult to argue the contrary unless one assumes that there is an ahistorical a priori “History” to which all historians unconsciously aspire in their writing — in which case my argument would be bankrupt and a singular historical text, sans metacommentary, could indeed contain “historical insight.” Considering the existence of this metacommentary in all historical texts, however, it is possible to see how this interaction between (at least) two historical texts can (and must) occur no matter how disparate the “objects” of study.

   To see if there really is such a thing as a disciplinary reality effect functioning in art historical discourse, I would like to return to the concrete example with which I began this essay, namely the Madonna with the Pomegranate (Madonna with Child). In “Botticelli Recovered,” Frank Kermode describes how, in the early nineteenth century, the work of Botticelli was reevaluated by painters, collectors, amateurs, and dealers from a lesser painter of little note into a major figure of the fifteenth century. [22] According to Kermode, Botticelli was used to construct a legitimate lineage to ground contemporaneous (for example, nineteenth-century) practices opposed to academic art, which, of course, were all based on more “classical” Renaissance artists. In the process of this (contingent) reevaluation, many of the works of his “school” were misattributed to the “master” himself and thus, ironically, his resurrection was enacted on what are now considered, for the most part, fakes and misattributed paintings.

   James Jarves, the collector, was apparently of the class of amateurs who were invested in this turn toward what were, at the time, nonclassical artists of the early Renaissance. [23] In the 1860s it would be highly likely that this particular painting would have been thought of as an “original” Botticelli. Jarves most likely purchased what he thought, [24] and what accepted authorities of the time thought, [25] was a Botticelli and sold it in this spirit. My attributing this painting, in light of the protestations of the current collection handbook and label to the contrary, might very well be a factual error. On the other hand, my “historical fact,” that Jarves sold a “Botticelli,” is not. I am not making a claim based upon who painted the panel, but rather on what name the panel is given. Whether this painting was indeed painted by Botticelli or not is particularly irrelevant as — as counterintuitive as this may sound — his actually painting it is insufficient to make it a “Botticelli.” To give it his name requires the affirmation of a proper authority, and I hardly qualify on that level. As such, this (unknowable) fact has no bearing upon the name which it carries.

   The actual facts that I served up in the previous two paragraphs, the facts that I can “get wrong,” are indicated by the references themselves. In making reference to Jarves’s public beliefs concerning this panel, I was making a statement of fact — a statement that includes within it an internal caveat through the indication of the source. The fact is that a claim was made on the status of this painting, and the existence of this claim can be verified.

   Concerning the Botticelli panel itself, however, we are faced with another level of referential objects — those that are usually assumed to be the facts, of which I intend to show are really the “facts,” of an art historical work. In stating that any reference to this panel directs the viewer to previous textual references, I am sure that a literalist reading of this narrative would counter that, despite (or perhaps as a result of the textual web of references, there is still the material object, the Madonna with the Pomegranate (Madonna with Child) to which all these references ultimately return. In response to this accusation I would admit that indeed there is such a panel, one that, after all, I myself have seen and even surreptitiously touched (when the conservator who was showing it to me turned his head)! I have no doubt that the panel currently on display at the Yale University Art Gallery consists, for the most part, of the same material substance as the one that Jarves sold in 1870, that Mignaty of Florence (a painter and close friend of Jarves’s), probably overpainted some time in the 1850s, [26] that Hammond Smith of New York restored in 1915 (EI xxv, 128), that Charles Durham of the Fogg Museum cradled in 1930 (EI xxv, 128), that Andrew Petryn of the Yale Art Gallery cleaned and in the process removed the work of the previous three in 1952, [27] and that, ultimately, I touched in 1994. I am going to argue, however, that all these references do not return us to this panel. Rather, as I now intend to argue, they lead us to it in a complete reversal of the causal chain.

   Its existence as a panel, as a painting, and indeed as a material object is a hard fact to deny. It would not even be possible to apply Joan Scott’s argument against the “experiential fallacy” to deny its existence. [28] And yet, on at least two levels, I can deny it.

   First, as the list of actions performed on this painting indicates, the object has not remained constant. In each case, there have been additions and subtractions from the surfaces of the panel — most notably in the tempera and varnish which, if this work is significant, it is significant for. Its value is not based on the wood that it is painted on! To follow through on this question would require engaging the question of when an object ceases to be that object. If I replace the leg on a table, is it the same table? If I then replace the other three legs, is it still the same table, and finally, if I then replace the top to the table, is it the same table? This question lends itself, without too much alteration, to a consideration of the practices of conservation — a consideration that is beyond the scope of this essay.

   Second, and more relevant to the argument at hand, “it” (the panel) exists not despite but as a direct result of the textual web of references which appear to refer to it. [29] The attribution similarly shifted throughout this period, from an “original,” [30] to the work of “an apprentice,” [31] to a “pupil of,” [32] to a “follower of,” [33] to the “school of,” [34] to the work of the “studio of,” [35] to a “close follower of” (EI 128), and even to the slightly awkward suggestion, after the restoration of 1952 (which, incidentally, Seymour mistakenly dates to 1954 [EI 128]), that “it is … possible that in it the hand of the pupil merges with that of the master” (RI 35), and finally, as already mentioned, in the most recent handbook of the collection as an “associate of” Sandro Botticelli. Is this, or is this not, the same panel? Each of these attributions has very different implications for the status of this particular, supposedly constant, panel.

   For rhetorical purposes we should focus on the most radical shift in its attribution — from an original Botticelli to an apprentice. Obviously, when William Rankin (the first to publish specific doubts concerning the veracity of the Botticellian attribution) claimed in 1895 that it is not an original Botticelli, he was not claiming that until his review of the panel it was a Botticelli, and ceased being so at his discretion. In fact (but not in “fact”) this is exactly what happened. His claim that it was the work of a “poor apprentice” (SE 148) is a “historical fact,” not a fact. It is not self-evident who painted this panel — there is nothing to present to the viewer other than the painting itself and without a reasoned argument as to why it should or should not be accepted as an authentic Botticelli it will remain a mute object. To state the obvious, if it were self-evident, there would have been no need for the continuous reattribution of the panel.

   Nevertheless, it seems obvious to Rankin that this “second-rate” panel, ipso facto, could not be by the hand of Botticelli (SE 147). His judgment is not in any way altered by the amount of overpainting that this collection received at the hand of Mignaty, as Rankin is not adverse to pointing out the cases where the overpainting rendered a painting unidentifiable — although, more often than not, overpainting appears to indicate to him that the panel was originally made by an “imitator” rather than a “master.”

   One exception to this demotion of severely overpainted panels in the Jarves collection is more indicative of where the authority of Rankin’s adjudications come from: “Ascribed to Giorgione is a small ‘Circumcision’ (no. 77), in which there is probably not a square millimeter of the original color remaining. It strongly recalls the Louvre ‘Holy Family’ attributed to Giorgione … but is abler in composition, and was perhaps once a Giorgione” (SE 150). If it is not already (logically) apparent, we can see here more clearly that the case for attribution cannot rest on an immanent reading of a singular work for the same reasons that an isolated text cannot constitute a historical argument. An attribution can only come through comparison with other works which, for the moment, are accepted as unproblematic.

   All references (not only Rankin’s) which presume to refer directly to the Botticellian or Giorgionesque panels do so only obliquely, working through already established textual descriptions of what would constitute a painting as an original Botticelli or Giorgione. The demotion by Rankin of the Madonna with the Pomegranate (Madonna with Child) to the work of an apprentice is based upon previous literary (for lack of a better word) descriptions of what counts as a Botticelli — previous descriptions that are not referenced in this particular text, but we must assume that Rankin is familiar with the literature of attribution. This is merely another case where the reader is asked to accept the “masses of detail that this historian is well aware of, but prefers not to mention” (EP 7).
Should we delve into the presumed authority of Giorgione’s Holy Family at the Louvre, we would see that that too rests upon other attributions based on Giorgione’s “style” and so on. This process is not at all stable, on a hermeneutic level, for should it transpire that the Holy Family be itself not a Giorgione [36] this would not in any way put into doubt Rankin’s estimation of the Circumcision, as the relevant comparison is to textual descriptions of the work. Tracing the web far enough, it will eventually be seen that there can be constructed a basic textual description of a Giorgione which is fairly stable, no matter which panels are accepted as being authentically “his.”

   To complicate this scenario more, in looking for a confirmation of this attribution to the Holy Family, one can find that in 1905 its attribution was far from clear. Although the Louvre maintained that it was a Giorgione, Mary Knight Potter states that “a pretty general opinion exists that it is not [Giorgione’s].” [37]

   In light of this, we could express doubt in Rankin’s familiarity with the literature of attribution. Although the Potter was published ten years after Rankin’s article, as a guidebook for the general public, one should assume that her ascription is based upon prevailing opinions of the previous decades. For example, at around this time, Paul Konody and Maurice Brockwell ascribe it to Cariani. [38] One could put Rankin’s, Potter’s, and/or Konody and Brockwell’s competence with regard to the existing literature in question. Nevertheless, to ask who, in this case, is right is to miss the point. As with many of the assertions being made in this essay, these facts are not as important as the illusion that an article gives of its author’s competence.

   Describing the process of naming thusly is to inscribe within this argument concerns about the very status of names. It would be useful, then, to return to the example of the reconsideration of Botticelli in the nineteenth century, but to look at it through Saul Kripke’s analysis of naming practices. [39] In doing so, I have to clarify one very important point. When Kripke analyzes the proper name, he assumes that the name “Nixon” (his example) refers to a specific individual (Nixon) and not to a socially constructed representation that is, only contingently anchored to a specific being (“Nixon” just happening to be correlated to Nixon). In contrast, within the discipline of art history, the name Botticelli does not refer to a specific individual of flesh and blood but, rather, to a body of work that is attributed to a literary figure — the author-effect of Foucault or implied author of reception theory. [40] In other words, because the disciplinary definition of what constitutes a Botticelli panel precedes the art historian who looks at his work, the description of a Botticelli leads to the painter, and not the reverse. As what we are interested in is this literary figure, it would make more sense to utilize Kripke’s analysis of species names and not proper names. Although the example of a species name which Kripke uses is gold, I will be substituting “Botticelli” where appropriate.

   The reevaluation of Botticelli in the nineteenth century should be considered as the source of our contemporary, canonized Botticelli — the origin of the name as we understand it, then, is in the nineteenth century. A description of a particular, unique, and valued painting style was used in the construction of the referent “Botticelli,” but the description is not interchangeable with the name. To schematize this process, we should imagine a scene with some of these early nineteenth-century proponents of Botticelli (including, more than likely, Pater) pointing to a body of work, and saying, “[Botticelli] is the substance instantiated by the items over there, or at any rate, by almost all of them” (NN 135). [41] The assumption of this scene is that the group of works attributed to Botticelli were, somehow, of the “same kind.” The need for a precise definition of what made them such was not necessary in the creation of the idea of Botticelli. “To the extent that the notion ‘same kind’ is vague, so is the original notion of [Botticelli]” (NN 136). What works were then considered part of this corpus was contingent and yet, in a very literal as well as the more technical senses, a priori. [42] That all these works were not to be found in a single collection, this pointing to the work was accomplished through a condensation of the corpus into a general textual description. Once the description establishes a fixed reference (Botticelli the painter), the particulars of that description (the various works themselves) are no longer necessary (again, in both the literal and technical sense). The textual definition, on the other hand, is analytic and thus necessary — but not a priori. [43] With this description, “[m]any who have seen little or no [Botticellis] can still use the term [properly]. Their reference is determined by a causal (historical) chain, not by [seeing any specific panels]” (NN 139). [44]

   At this point in Kripke’s analysis, however, the difference between gold and Botticelli becomes apparent. Kripke points out that after the category “gold” is established, “[s]cientific investigation generally discovers characteristics of gold which are far better than the original set” (NN 138). For gold, this would be its atomic structure, number 79 on the periodic table. With Botticelli, however, it is the original description correlated with “his” name that is used to purge that very corpus of some (according to Horne, most) [45] of the work that was constitutive of that very description. Rankin’s demotion of this panel, for instance, is based upon the “fact” that “Botticelli never did such hard outlines or such feeble hands” (SE 148). The name “Botticelli” comes retroactively to constitute that to which it only appeared to refer (BM 210). The content of this name is not, then, representational of what a Botticelli panel should be — it performatively establishes what can be accepted as a Botticelli a posteriori (SO 99).

   Zizek writes the following in his psychoanalytic revision of Kripke’s account: “ [I] t is the element which represents the agency of the signifier within the field of the signified…. [I]ts role is purely structural, its nature is purely performative — its signification coincides with its own act of enunciation; in short, it is a ‘signifier without a signified’” (SO 99). [46]

   If this description sounds somewhat familiar, it should. It is a restating of the role of the notational detail, the reference which refers to signification. The names of the art historical text are the notational details, and the insignificant aspect of any reference (to either a primary or secondary source) is a reference to a name. That the name is insignificant is also indicated in Jacques Ranciére’s description of the general structure of the historical text — and its relation to the content which fills that structure: “[T]he materials are nothing without the architecture…. What we avoid considering is simply this: history is, in the final analysis, susceptible to only one type of architecture, always the same one — a series of events happens to such and such a subject…. We are [inevitably] confronted by the leap into the void, against which no auxiliary discipline’s rigors offer a guarantee: we must name subjects.” [47]

   The implication of Ranciére’s analysis is that what is always interchangeable is the main “subject” of a historical narrative. Whether our historian is dealing with historical persons, objects, or events, the predictive structure remains the same — at its most basic, something happened. The base narrative supersedes the naming of the subjects involved. To whom (or what) it happened is irrelevant as far as this basic structural claim of the discipline is concerned. The naming of a subject does not bring that subject into being, and so we can take Ranciére’s structure of history and the previous analysis of names and conclude that in the historical text that which is not significant is which names are used.

                * * *

In order not to mislead my reader into thinking that this unstable process of attribution is in any way alleviated by contemporary attribution studies under the influence of much more technologically complicated apparatus, we need only look again at the “Rembrandt Project” — one of the more celebrated projects of contemporary connoisseurship. As steeped as this project is in scientific analysis, it still requires an authoritative countersignature to perform the attribution. [48] Svetlana Alpers has written that “ [i]t is clear within the volumes [of the Rembrandt project] themselves … scientific or technological means … assist, but do not replace, connoisseurship or the judgment of the trained eye.” [49] Seen in this light, it appears that the situation has not changed much, aside from a need for more instrumental supports to make the claim being made appear reasonable. [50] Ultimately, there has to be a voice that speaks, a voice authorized through the acknowledgment of being associated with an “eye,” and names the work under consideration. In the final instance (which is never really final) the call comes down to a performative “act of connoisseurship.” Contrary to the standard cliché, saying does seem to make it so.

   Admittedly, this argument so far has concentrated on the level of attribution and, as surely as the situation when a historical project is limited to the stringing together of facts, if the art historical discipline is limited merely to attribution studies, its relevance for cultural history (or, indeed, history) would certainly be suspect. The implications of this process, however, reach over and beyond the mere act of attribution. If we think of attribution as the action of identifying an object, this process would also include the more generally useful project of identifying the age, the period, the style, and the provenance of a work. Naming the work is merely a special case of the necessary step of knowing what one is writing on — as basic to the discipline as the ability to cite texts properly. It also is the basic strategy for distancing the object, of placing it into the past, as what is effectively of concern is not the “present state” of the object but its (fictional) original state. This “original state” is thought to be extrapolated from the material of the archive.

                * * *

This description is not to discount the rather important factor in the methodology of attribution, the implicit (although at times overtly explicit) valuation of the work under consideration. The estimation of the Madonna with the Pomegranate (Madonna with Child) by Rankin is an example of a rather explicit case. The attribution of a work for the purposes of the study of a historic era, on the other hand, must by necessity value a work implicitly — is this work “relevant” for the period under discussion, will it advance my argument? If the answer to both of these questions is yes, it becomes a “significant” work. The main example that I have been using, the Botticelli panel, is extremely useful for my argument, and thus I am valuing it as a significant work for the purposes at hand. Its very insignificance for “art history,” as the work of an unknown associate of Botticelli, makes it that much more relevant and significant for my purposes. Similarly, to use a rather tortured description, Rankin’s Giorgione’s Holy Family became useful for my line of argument, and thus I valued it as an exemplary attribution. As valuation plays such a role in all attributions (either as primary sources or as objects of study) then Kermode’s analysis, which focuses on the relation of the canon (as structure) to art historical writing and thus on valuation in general, can play a significant role in my argument.

   I want, then, to quote Kermode who is quoting Gombrich who is quoting Warburg, who, through this labyrinth of quotations, notes that “what is made of [historic artifacts] … depends upon the subjective make-up of the late-born rather than on the objective character of the classical heritage” (BR 26). The implication of Kermode’s analysis is that ultimately even the most archivally based argument is based upon present concerns … a matter of “opinion.” What is found (or not found) in the archive is not so much the “truth” of the panel but the historical legitimacy for it, its attribution, and its significance.

   Kermode describes the relation between the work and its subsequent criticism in the following manner: “The success of interpretive argument as a means of conferring or endorsing value is, accordingly, not to be measured by the survival of the comment but by the survival of its object. Of course, an interpretation or evaluation may live on in the tradition on which later comment is formed, either by acceptance or reaction; but its primary purpose is to provide the medium in which its object survives” (BR 67). He seems to view the role of criticism as regaining (or activating) the relevance of significant works of art. The flow is from work to text in order to get back to the work in the final instance. This flow of significance might be likened to some form of alchemy, whereby the base material of wood, pigment, damar varnish, and egg white (in the particular case of the Botticelli panel) is alternately made into gold (as a “Botticelli”) or lead (as an “associate of Botticelli”) — suggesting that, perhaps, my substitution of Botticelli for gold in the Kripke quotations above was not such an arbitrary decision. [51]

   While Kermode’s quotation seems in agreement with the argument I am attempting to make, a closer look illustrates otherwise. “[A]n interpretation[’s] or evaluation[’s] … primary purpose is to provide the medium in which its object survives.” What seems most significant in this phrase is the role that a canon-defending argument is granted. On the surface, Kermode appears to give to the object primacy, yet at a deeper level he suggests a fragility of the work of art so described. What a canon defending argument does is the necessary work to naturalize the “master-work” at the same time that it creates it through naming it. The work as such, as a manifestation of the object that is more properly its medium, has to disappear in order to reappear as an effect of the text.

   In this manner, the art work appears located both in the “past” and in the present. It is in the past by virtue of its material existence; for example, it was made before the now. It is in the present as an object that must promise the possibility of being presented. This present-ness of the art object, however, ultimately is caught in the rhetoric of the historical explanation. That is, its present-ness is as a historical object. The undertaking of art history places any work (even a contemporary work) in the past.

   To look at the Madonna with the Pomegranate (Madonna with Child) as an exemplary archival object, by which I mean all primary sources — the names of the hermetic text — we see that they only gain significance through their reference in referenced texts. It is in those cases where they have been “black boxed” most proficiently that they masquerade most effectively as facts (whereas they are truly “facts”). [52] When a painting is used in the furtherance of other arguments (as opposed to being the focal point of an argument), its status must be unproblematically established either through referencing other, authoritative works (such as with the case of Giorgione’s Holy Family above) or through a digression that usually takes place within the footnotes of the text. [53] Its own contested status cannot be acknowledged as such a concession would throw the validity of the main argument into doubt in a manner related to, but not identical with, the risk of being caught with erroneous facts.

   On the level of the discipline, what occurs is a rather strange affair, whereby the further the referenced object is from the argument, the more apparent authority the referential act carries with it. It is in the externally directed (and displaced) facts (and I am not sure in this case whether /facts/ should be in scare quotes or not) that external artifacts become notational themselves. In the discipline as a whole, then (Ankersmit’s “between historical texts”), present (and, one might add, presentable) artifacts (and not just their names) could be considered as the notational details, insignificant without the predictive framework of the discipline. It is the alternately necessary present-ness of the archival material — necessary for the discipline to exist as a whole — in the face of the simultaneous distancing of this same material within the predictive narratives of individual texts which is the source of the tension which real-izes the discipline. Perhaps the most effective example I have given of the tension between the distancing and presenting of an object can be seen in Rankin’s qualified attribution of the Giorgione which “was perhaps once” an authentic work which now no longer apparently “exists.”

   My use of Botticelli’s Madonna with the Pomegranate (Madonna with Child) does not present the reader with that specific panel — and any included reproductions are merely attempts to hide the absence of the panel. At best, any form of reference to the panel will refer the reader to the other textual sources which also reference it. In conjunction with understanding “historical facts” as consisting of (affirmative) statements which are themselves dependent upon other historical interpretations, we can see how it is that this initial nonreferentiality is shifted from the name to the objects themselves while retaining the emptiness of the referential act. My “historical fact” that Jarves sold a Botticelli is not, in the end, wrong — he did indeed sell a Botticelli. That the Botticelli that he sold no longer exists (for reasons directly opposed to the reasons that the Giorgione no longer exists) does not in any way deflect this “historical fact.”

                * * *

Surely the reader will realize that the status of this once-existing panel and its relation to the presently existing panel at Yale was merely a foil for a much broader issue — the issue of what it is that an art historian does. The strategy of the historian is to find in the “objects” of the “past” precedents that are useful for arguing for a specific view of the present. These objects are brought to the present only by being presented as evidence. In order for the rhetoric to be effective, however, these precedents have to remain as part of the past. To illustrate them as contemporary constructions would be to degrade their usefulness. What Norman Bryson describes as the “super-Hirschean indignation” (AC 77) of art historians in the face of postcontemporaneous readings of works of art could be considered as symptomatic of a dis-ease in the presence of works of art. The attempt to “explain” conclusively works of art would be, then, an attempt to control them. The implication of this indignation is that “significance” is thought of being fixed at the time of execution of the work. “Significance” could be understood as an additional ingredient in the varnish that makes up the finish — another alchemical formulation. This placing of significance at the time of execution might help to explain the tremendous amount of anxiety that the restoration of such canon-fodder as Michelangelo’s Sistine Chapel or Leonardo’s Last Supper can occasion — the fear being that the “significance” of the work will be stripped along with one of the layers of discolored, and therefore ironically no longer transparent, varnish. [54]

   A discourse that purports to deal with an artifact engages, more often than not, previous exegetical accounts of the work rather than the works themselves. To return to an idea from the beginning of this essay, as Michael Baxandall has noted, the history of art history is ultimately the history of writings about art — art which only is presumed to remain the same object throughout its history.

   The combined effect is that historians, or critics, cannot speak of any work that they might find in front of them presently. It must be distanced by its placement in the museum, the archive, or the library — all of which denote the storage of artifacts no longer current. As such, as notational in the rhetorical strategy of writing history and thus “insignificant” by virtue of its place within the rhetoric, an existent work of art poses a threat to the argument as presented by the historian. Part and parcel of the notion of canonical works is that they are not exhausted by any one — or any series — of interpretations. As Kermode puts it, “[i]nterpretations may be regarded not as modern increments, but rather as discoveries of original meanings hitherto hidden” (BR 75). It would be hard to prove this wealth, if not excess, of significance of which the work is considered capable but, in light of the number and history of interpretations of “canonical” works, it might be even more difficult to disprove this potential. Like Burckhardt’s Renaissance, it appears to be a self-fulfilling prophecy. The source of these significances, however, is not necessarily the work itself, although the object is offered as (unproblematic) source. The continued existence of the work as a presence, then, is a threat to any reading of it that claims any sort of authority. The reception of a work, the possible readings after the fact, might be to a certain extent limited by a “master-reading,” but the materiality of the work — unaffected aside from certain extenuating circumstances to be discussed — may return to contradict the reading offered. In this sense it could become embarrassingly unruly to the point of countering the textual arguments made “on its behalf.”

                * * *

As a preliminary conclusion for this essay, I should offer some sort of simplified graph of the role that the “fact” (and by now I will use this term to refer to all those empty references that the historian can use, references to absent texts, absent documents, absent objects) plays in the writing of history.

   If we think of the art historian working in the library or the archive, we would see her working on material objects of various sorts — objects photographs, illustrations, as well as archival documents and facsimiles those things usually called “primary sources.” We would also see her working with other sorts of material objects — although these are less often considered as material objects — the secondary sources of existing books, articles, photographs, and illustrations. In other words, at this point she is conducting research. If we have tracked her progress, we would see that she began in the present and worked her way backwards through time, attempting to access farther and farther into the past of her subject while leaving its present state.

   During the course of this process, we would also see her conducting a rather strange ritual — transcribing notes, phrases, descriptions onto index cards. This process, seemingly innocuous, is the process of creating the referential facts from which she will later weave her narrative. [55] At a certain point (and here I must confess that I am idealizing the activity somewhat), having reached her destination in the archive, she returns the books and objects to their rightful place in the archive or library, gathers her note cards, and returns to her study.

   She began in the present, researching backwards, and now in writing she writes from the most distant toward the present where she began — thus “prescribing for beginnings what is in reality a point of arrival” (86). These index cards are now the raw material with which she works. She orders these cards, links them together with the help of conjunctions and some narrative substance, and produces a text which contains the content of these index cards as newly fashioned referential facts and “historical facts” which might later be recycled as referential facts for future historians. As referential facts, they now have a life and meaning of their own, separate from the material objects which we might think of as their source. The significance of these facts is made in the text, not in the objects themselves.

   The story might have ended here, but this historian’s text is published and it is taken up into the preexisting discourse concerning her subject. In the reception of this text by other historians, these referential facts meet up with other referential facts which appear to be similar in content (for instance, Madonna with the Pomegranate (Madonna with Child)). At this point, these referential facts meet up with the material object to which this name refers. This panel, for whatever reason, is thought significant. The significance of this panel, however, comes from the discourse that surrounds it. It might receive more attention, it might even be submitted to the attention of a conservator and thus be directly affected, or it might become the centerpiece of a blockbuster exhibition. None of this is a result of the significance of the panel itself, however, as any significance it might be granted comes from the discourse which props it up. [56]

                * * *

It is vital to assert one clarification in this formulation. As I have already indicated, for illustrations of my argument I have used for the most part those forms of art history which are most susceptible to these strategies — specifically attribution studies. I have already noted that the act of attribution is not limited to straightforward exercises in connoisseurship, and that the same action of distance occurs (if not more forcefully) in what we might think of as traditional art history (which we might typify as monographs which purport to deal exclusively with the “objects” of art history in past contexts).

   In recent years, art history has undergone its own form of the “linguistic turn” in the formation of what has come to be called the “new art history” [57] and, even more recently, a trend towards (for lack of a better phrase) a “newer art history” — of which one of the best indicators is the proliferation of graduate programs with various permutations of visual studies and cultural studies as names. [58] If the “linguistic turn” in historiography can only be pinpointed as having occurred in the years between the establishment of the journals History and Theory and New Literary History, [59] its parallel in art history can be also only described as happening over a range of time. Beginning with the Winter 1982 issue of The Art Journal and, more importantly, echoed in an editorial in The Art Bulletin in March of 1986, the discipline announced to itself that it was in crisis. [60] The influence of this announcement might have been negligible (self-proclaimed crises are usually easily solved) if it had not been followed by a continuing series of self-referential essays in The Art Bulletin. [61] This development was probably also aided and influenced by the institutionalization of poststructuralist-friendly criticism in contemporary art since the late 1970s as exemplified by the journal October, [62] and the “invasion” of the subject of art history by literary theorists in search of new material in the early to mid 1980s. [63]

   It might be thought by some of the participants of these new and newer art histories that the subterfuges hidden in the use of “historical facts” may have been superseded by the self-reflexivity that they practice. To look at Kermode’s Forms of Attention as an example in seeing if the applicability of my analysis extends to such self-referential projects, I am led to the inevitable conclusion that — as far as the use of “facts” goes — nothing much has changed. This claim is not a condemnation of either traditional or more recent, theory-informed practices. It is merely to face the inevitability that even a text wishing to illustrate the construction of history needs to utilize the same strategies of more traditional historical accounts. Examples of constructions of the “past” from past historical texts are inevitably placed within “past” historical contexts. In their present-ing they are distanced. At the same time that Kermode is describing what equates to an opportunism (if unconscious) on the part of the nineteenth-century supporters of Botticelli, he cannot question too closely his own opportunism in using the recovery of Botticelli to help illuminate debates about the role of the canon in 1985. His placing of Botticelli firrnly in the canon is based upon the “merits” of these panels as masterworks, as being worthy of consideration. He does not consider the possibility that his own valuation of Botticelli (or Shakespeare, the other example he uses) as worthwhile might be only because they are useful in defending the legitimacy of a certain history of art based upon the logic of the canon. His argument is performed as a defense of the notion of canon in the face of the admission of the contingent construction of that notion. To extend this argument even closer to home, neither can I question too closely my use of Kermode in making an argument about the contradictions of writing history.

University of Rochester

                * * *


1 Yale University Art Gallery, Handbook of the Collections (New Haven, 1992), p. 135.

2 Carl L. Becker, “What are Historical Facts?” in Detachment and the Writing of History: Essays and Letters of Carl L. Becker, ed. Phil L. Snyder (Ithaca, 1958), pp. 41-64; hereafter cited in text as WH.

3 See Nancy F. Partner, “Historicity in the Age of Reality-Fictions,” in A New Philosophy of History (Chicago, 1995), pp. 21-39. See esp. pp. 21-22.

4 Similarly, Michael Baxandall’s claim that art historians “do not explain pictures: [they] explain remarks about pictures,” can be read to say that art historians do not study art, they study art history. To make my intentions as clear as possible, I ask that this last sentence be read in the strictest possible sense (Michael Baxandall, Patterns of Intention: On the Historical Explanation of Pictures [New Haven, 1985], p. 1) .

5 Frank Ankersmit, “The Reality Effect in the Writing of History: The Dynamics of Historiographical Topology,” in History and Tropology: The Rise and Fall of Metaphor (Berkeley, 1994), p. 145; hereafter cited in text as RE.

6 Roland Barthes, “The Reality Effect,” in French Literary Theory Today: A Reader, ed. Tzvetan Todorov (Cambridge, 1982), p. 17.

7 Barthes, it is true, does use an example of a historical text of Michelet’s; however, he treats this text purely as a literary artifact. While considering a historical text as anything other than a literary artifact would be wrong, I am working under the assumption that to treat it as only a literary artifact would be a mistake. My argument, as well as my discipline, is predicted on there being a difference between “pure” fiction and history, no matter how constructed history turns out to be.

8 Dale H. Porter, The Emergence of the Past: A Theory of Historical Explanation (Chicago, 1981), p. 7; hereafter cited in text as EP.

9 Becker does refer to the historical fact as a symbol. Whether we are to take his usage as being identical with the present usage is moot. See Becker, “What are Historical Facts?” p.44.

10 Stephen Bann, “Analysing the Discourse of History,” in The Inventions of History: Essays on the Representation of the Past (Manchester, 1990), pp. 33-63, esp. p. 38, and Partner, “Historicity in the Age of Reality-Fictions,” p. 24.

11 Bruno Latour, Science in Action: How to Follow Scientists and Engineers through Society (Cambridge, Mass., 1987); hereafter cited in text as SA.

12 Although Latour’s argument is directed toward the “hard” sciences, his discussion of the rhetoric of the scientific fact is easily translated to the discourse of history,

13 Michael Holly, Past Looking (Ithaca, 1996), p. 48.

14 James Nichols, “Introduction,” in Jacob Burckhardt, Force and Freedom: Reflections on History (New York, 1943), p. 54.

15 Jacob Burckhardt, The Civilisation of the Renaissance in Italy, An Essay, tr. S. G. C. Middlemore (London, 1960), p. 53. For Holly’s analysis of Burckhardt, see her “Burckhardt and the Ideology of the Past,” History of the Human Sciences, 1, no. 1 (1988), 47-74 and “Cultural History as a Work of Art: Jacob Burckhardt and Henry Adams,” Style, 22, no. 2 (1988), 209-18.

16 Admittedly, Latour is speaking of science, but the argument is easily expanded, as I indicate in my discussion above.

17 That is, I doubt that anyone would seriously take me up on my previous offer of showing the Yale catalogue to them personally.

18 Barthes’s main example in “The Reality Effect.”

19 He uses this term as a logical entity, one that cannot necessarily be distilled out into test tubes, but which names a function that can be found in the structure of the text much like the reality effect itself (Frank Ankersmit, “Historical Representation,” in History and Tropology, pp. 97-124, esp. p. 113).

20 Frank Ankersmit, “Six Theses on Narrativist Philosophy of History,” in History and Tropology, p. 38; hereafter cited in text.

21 James Elkins, “Art History without Theory,” Critical Inquiry, 14 (1988), 354-78. This argument is a more nuanced version of Baxandall’s claim in n. 4 above. For a more materialist (in the literal sense) account, see Keith Moxey’s description of the relationship of art historians and “their” library shelves as used in the introduction to his investigation of the ideological underpinnings of (the canon of) art history (Keith Moxey, “Motivating History,” The Art Bulletin, 77, no. 4 [1995], 392-401).

22 Frank Kermode, “Botticelli Recovered,” in his Forms of Attention (Chicago, 1985), pp. 3-31; hereafter cited in text as BR.

23 See Francis Steegmuller, The Two Lives of James Jackson Jarves (New Haven, 1951).

24 See, for instance, the attribution of plate 30 (Madonna with the Pomegranate) as a Botticelli in James Jarves, Art Studies: The ‘Old Masters’ of Italy: Painting (New York, 1861), and Steegmuller, The Two Lives of James Jackson Jarves, p. 173.

25 For an example, see Russell Sturgis, Jr., Manual of the Jarves Collection in Early Italian Pictures, deposited in the Gallery of the Yale School of the Fine Arts (New Haven, 1868). That Sturgis was obviously commissioned to write this manual, and thus his “intentions” could be suspect, does not discount his “status” as “authority.” That he was chosen to be commissioned in the first place is indicative of his authority. That Berenson’s, indeed Vasari’s, intentions in attributions are suspect does not discount the very real effect that their writings have had on the discipline.

26 Charles Seymour, Jr., Early Italian Paintings in the Yale University Art Gallery (New Haven, 1970), p. xxiv; hereafter cited in text as EI.

27 Yale University Art Gallery, Rediscovered Italian Paintings: An Exhibition, March 25 through May 18, 1952, of fourteen recently cleaned paintings from the Jarves Collection (New Haven, 1952), pp. 10-11; hereafter cited in text as RI.

28 Joan W. Scott, “The Evidence of Experience,” Critical Inquiry, 17 (1991), 773-97.

29 I am working under the assumption that illustrations, such as to be found in any of the texts I am referencing, are part of the textual web. The importance of the captions, which anchor these illustrations to a particular interpretation (a more accurate term than attribution) of the work thus referenced, cannot be downplayed. The caption can be considered to work in a similar manner to the argument concerning photographic captions in Roland Barthes, “The Rhetoric of the Image,” in Image-Music-Text (New York, 1977). After all, these illustrations are (for the most part) photographs of the work itself.

30 Jarves, Art Studies, p. 130; Sturgis, Manual of the Jarves Collection of Early Italian Pictures, p.68.

31 William Rankin, “Some Early Italian Pictures in the Jarves Collection of the Yale School of Fine Arts at New Haven,” American Journal of Archaeology, 10, no. 2 (1895), 137-51, esp. 148; hereafter cited in text as SE.

32 Osvald Siren, A Descriptive Catalogue of the Pictures in the Jarves Collection belonging to Yale University (New Haven, 1916), p. 135.

33 Richard Offner, Italian Primitives at Yale University (New Haven, 1927), p.6.

34 H. P. Horne, Alessandro Filipepi, commonly called Sandro Botticelli, Painter of Florence (London, 1908), p. 118, and Raymond van Marle, The Development of the Italian Schools of Painting (The Hague, 1931), pp.238-39.

35 Bernard Berenson, Italian Pictures of the Renaissance: A List of Principal Artists and their Works with an Index of Places (Oxford, 1953), p. 104.

36 For what it is worth, present art historical research suggests it and the Yale panel are not the work of Giorgione. The Louvre Holy Family is thought to be a Sebastiano del Piombo (Laurence Gowing, Paintings in the Louvre [New York, 1987], p. 219), while the Yale Circumcision is attributed, with some reservation, to Titian (Seymour, Early Italian Paintings, p. 275).

37 Mary Knight Potter, The Art of the Louvre: Containing a Brief History of the Palace and of its Collection of Paintings, as well as Descriptions and Criticisms of Many of the Principle Pictures and Their Artists (Boston, 1905), p. 94.

38 Paul G. Konody and Maurice W. Brockwell, The Louvre: 50 Plates in Colour, ed. T. Leman Hare (New York, nd. [ca. 1900]), p. 68.

39 Saul Kripke, Naming and Necessity (Cambridge, Mass., 1980); hereafter cited in text as NN.

40 On the author-effect, see Michel Foucault, “What is an Author?” in Textual Strategies: Perspectives in Post-structuralist Criticism, ed. J. Harari (London, 1980). On implied author, see Genette who suggests that the term is redundant in that all authors are only implied (Gerard Genette, “Implied Author, Implied Reader,” in Narrative Discourse Revisited [Ithaca, 1988]).

41 I am, of course, not suggesting that such a scene ever transpired — merely that this is the logic of the naming process. I am also not suggesting that the work of Botticelli was completely unknown before his popularization, but it makes the scene easier to visualize if we bar Vasari and others at the door.

42 See Kripke’s first chapter on this seemingly contradictory combination (Kripke, Naming and Necessity, pp. 22-70).

43 With regard to the textual definition, these two terms (“a priori” and “necessary”) should only be read in the technical sense.

44 The insertion of “properly” into this scheme is owed to Judith Butler’s analysis of Slavoj Zizek’s analysis of Kripke. See Judith Butler, Bodies that Matter (New York, 1993), esp. ch. 7, “Arguing with the Real,” pp. 187-222, and p. 217 in particular; hereafter cited in text as BM. Her analysis is of ch. 3, “Che Vuoi?” in Slavoj Zizek, The Sublime Object of Ideology (London, 1989), pp. 87-130; hereafter cited in text as SO.

45 See Horne, Alessandro Filipepi, p. 118.

46 Admittedly, Zizek is writing on the use of names in political ideologies. I do not think it a stretch to consider the “names” of art history being any less ideologically motivated. His analysis leads toward the Lacanian notion of the truth-effect, which could be considered another name for the effect of the reality effect.

47 Jacques Ranciere, The Names of History: On the Poetics of Knowledge, tr. Hassan Melehey (Minneapolis, 1994), p. 2.

48 It would be useful to think of this process of attribution as an example of the countersignature as described in Jacques Derrida, “Signature, Event, Context,” in his Margins of Philosophy, tr. Alan Bass (Chicago, 1982), pp. 307-30.

49 Svetlana Alpers, Rembrandt’s Enterprise: The Studio and the Market (Chicago, 1988), p. 125.

50 See Latour, Science in Action, ch. 2, “Laboratories,” esp. pp. 79-94.

51 Using the work of Michael Taussig as a model, I would rather that this reference not be taken as only a metaphor. See esp. his Mimesis and Alterity (New York, 1993).

52 The whole tenor of this argument should strike familiar with the reader familiar with Norman Bryson and Jonathan Culler’s discussions of context. See Norman Bryson, “Art in Context,” in Studies in Historical Change, ed Ralph Cohen (Charlottesville, 1992), pp. 1842, and Jonathan Culler, Framing the Sign: Criticism and its Institutions (Norman, Okla., 1988).

53 Note 36 could be considered an example of this very practice.

54 Michael Holly has suggested to me that perhaps one should think of history writing as the final layer of varnish.

55 “In history, everything begins with the gesture of setting aside, of putting together, of transforming certain classified objects into documents. This new cultural distribution is the first task. In reality it consists in producing such documents by dint of copying, transcribing, or photographing these objects, simultaneously changing their locus and their status. This gesture consists in ‘isolating a body as in physics and ‘denaturing’ things in order to turn them into parts which will fill the lacunae inside an a priori totality. It forms the collection of ‘documents…. Far from accepting ‘data,’ this gesture forms them” (Michael de Certeau, The Writing of History [New York, 1988], pp.72-73; hereafter cited in text).

56 This formulation echoes Jacques Derrida, “The Parergon,” in The Truth in Painting, tr. Geoff Bennington and Ian McLeod (Chicago, 1987), pp.15-148.

57 For examples of the “new art history,” I refer the reader to the collection of articles that goes by that name, The New Art History, ed. A. L. Rees and Frances Borzello (London, 1986).

58 And I hope that I have the proper ironical tone in the text at this point, considering the program that I am presently affiliated with. This essay, and my educational history, is an example of this shift to these “newer art histories.” For discussions concerning this shift to “visual studies,” I refer the reader to October, 77 (1996).

59 See Richard T. Vann, “Turning Linguistic,” in A New Philosophy of History, ed. Frank Ankersmit and Hans Kellner (Chicago, 1995), pp. 40-69. See esp. pp. 56-61.

60 “The Crisis in the Discipline,” Art Journal, 49, no. 4 (1982), guest editor Henri Zerner; and Richard E. Spear, “From the Editor,” The Art Bulletin, 68, no. 1 (1986), 6. This is not to suggest that Zerner and Spear are uniquely to be blamed (or lauded) for the instigation of this crisis, but with the Art Journal and The Art Bulletin playing the roles of “the voice” of the discipline in the United States, the announcement of this crisis in these journals is particularly significant for the development of the discipline.

61 An informal survey of graduate students identified with “visual studies” shows that these series of articles are often the only essays read in The Art Bulletin. This probably is significant. The series include the “State of the Art,” the “Views and Overviews,” and the “Range of Critical Perspectives” essays. Initiated under the editorship of Richard E. Spear, the “State of the Art” articles can be found in issues 68, no. 1 (1986); 68, no. 4 (1986); 69, nos. 1 - (1987); and 70, nos. 1-2 (1988). The “Views and Overviews” articles, under the editorship of Walter Cahn, can be found in issues 71, no. 4 (1989); 72, no. 4 (1990); and 73, no. 2 (1991). The editorship of Richard Brilliant offers nothing comparable to these collections, although he does offer a consistent editorial voice — the only structural problem is that his is the only voice offered. As editor, Nancy J. Troy instituted the “Range of Critical Perspectives,” which are found in each issue since 74, no. 3 (1992).

62 In many ways October merely took the place that was held by Artforum in the 1970s. The difference between the two is the shift from being primarily image driven (or, more correctly, artwork driven) to being theory driven and textually based. As far as I know, Artforum has never had an issue not published in color, whereas October has never published in color.

63 The most visible of these “poachers,” who then became absorbed by the discipline, being Norman Bryson, with his Vision and Painting (New Haven, 1983), and Mieke Bal, with her Reading Rembrandt (Cambridge, 1991). I would also, however, consider that Frank Kermode’s Forms of Attention is another example of a literary theorist who appropriated an art historical project the nineteenth-century recovery of Botticelli to his own purposes. Although he has not been “absorbed” into the discipline as Bryson and Bal have, this text can be assumed as an art historical one (an assumption that I have already made in my use of him as a primary and secondary source).