13 Borders and Crossings
Disciplinary Demands
There are often multiple ways to understand a given topic, for the relevant information may be integrable into diverse accounts. A geneticist, a nutritionist, and an epidemiologist are apt to understand the increasing incidence of diabetes differently. Even if they draw on the same empirical data, each interprets those data in terms of different models and appeals to different background theories for support. Each highlights features that the others marginalize. Different histories of the same episode show similar diversity. One may focus on political factors, another on economic considerations, a third on the effects on domestic life. This enables them to discern different patterns in the data, and draw different inferences from them.
Can works of art afford historical understanding? Or are the commitments of art and history too disparate for a single work to satisfy the demands of both? To examine the issue, I focus in this chapter on Maya Lin’s Vietnam Memorial. As a monument, it denotes and commemorates something deemed to be historically significant. It makes reference to the past. That alone does not indicate that it affords any understanding of the events it refers to. The Washington Monument affords no understanding of the first US president. Nor, if it does afford some understanding, should we automatically conclude that it affords a historical understanding. The question is how the Vietnam Memorial refers to the episode it commemorates. Does it do so in ways that comport with the expectations of the discipline of history?
A discipline is constituted in large measure by methodological and normative framing commitments that are reflectively endorsed by the legislating members of a particular realm of epistemic ends. Natural science is committed to accepting only findings that are intersubjectively accessible, backed by empirical evidence, preferably testable, and defeasible. Although romantic poetry affords an understanding of the natural world, it does not afford a scientific understanding, for it does not respect science’s framing commitments. Mathematics demands proof. Although there is plenty of evidence for Goldbach’s conjecture, that evidence does not figure in a mathematical understanding, since instantial induction—regardless of the number of instances—does not amount to proof. History, like natural science, requires that evidence be publicly accessible, publicly assessable, and defeasible. The evidence must concern past events. A historical understanding of those events typically treats them as sequential and shows why, in the circumstances, they were to be expected (Hempel, 1965). This is far from a detailed characterization of what disciplinary understanding consists in. My point is to emphasize that disciplines are more than ways of marking off turf. They carry commitments about how a particular subject matter is to be properly understood.
One critical question is how strong a discipline’s framing commitments are. Should they be construed as inviolable requirements, as rules of thumb, or as something in between? Disciplines evolve. So presumably framing commitments cannot be utterly rigid. But how flexible they are or should be, and whether this varies from one discipline to the next, remain open questions. In this chapter I am concerned primarily with history. I shall treat its framing commitments as touchstones that establish strong presumptions but not rigid requirements. Some works that are treated as uncontroversially providing historical understanding seem not to fully conform to any plausible rigid rubric.
The question I address here is whether it is plausible to contend that works of art, functioning as such, can embody and convey the sort of understanding of a period that the discipline of history demands. Obviously works of art can serve as historical evidence. Like other artifacts, they are products of an era and can be used to shed light on what was going on during that era. Poe’s “The Raven” and Mussorgsky’s Night on Bald Mountain can, for example, serve as evidence of the nineteenth-century fascination with the macabre. My question is whether works of art can wholly or largely constitute an understanding of an episode, where an episode is a constellation of related events that take place over a limited span of time. I will suggest that the answer is yes. In particular, I will urge that some monuments, and perhaps some other works of art, are jointly artistic and historical symbols. Nor is their performing both roles a coincidence. They embody and convey a historical understanding of the episode they concern, and they do so because of their aesthetic functions.
Must Understanding Be Discursive?
Grimm (2008), Khalifa (2012), and others contend that understanding is keyed to explanation. This at least strongly suggests that those who understand a phenomenon can give or accept a verbal account of it. If so, all understanding is discursive. In that case, my contention that dance embodies understanding is incorrect. They are surely right that formulating and articulating an explanation of a phenomenon is a manifestation of understanding it. But Lipton (2009) makes a compelling case that the ability to explain is not necessary. Someone could, he maintains, display her understanding of the retrograde motion of the planets by properly manipulating an orrery.1 Moreover, I would add, someone could teach why the planets display retrograde motion by giving a nonverbal demonstration using an orrery. With the orrery, we can just see why, from the perspective of an observer on Earth, Mars appears to reverse its direction as it orbits. If we cannot correctly express our understanding in words, that seems at best a minor defect. Still, our inability to explain might just be put down to inarticulateness. We apprehend and appreciate what is going on, but words fail us. Lipton’s second example goes deeper. Galileo demonstrated that gravitational acceleration is independent of weight by showing that the alternative leads to a contradiction. We understand that this must be so—that it could not be otherwise—even if we have no idea why gravitational acceleration is independent of weight, or what an explanation of that independence would look like. In general, it is very hard to explain why a is independent of b. A third example, this one mine, is Richard Feynman’s dropping an O-ring into a glass of ice water to demonstrate that O-rings lose elasticity in the cold. This enabled scientific novices to understand why the O-ring’s failure caused the space shuttle Challenger to explode. Feynman’s point could be captured in words. Indeed, expert witnesses testifying before a congressional committee had put it into words. But the congressmen did not understand the verbal explanations. The demonstration did the trick (Feynman, 2001). These examples show that understanding need not be discursive. It is possible to understand something without being able to put one’s understanding into words, and possible to understand something without understanding a lucid verbal explanation of it in a language one speaks fluently.
Lipton does not convince everyone. Khalifa (2012) contends that the understanding Lipton’s examples provide is somehow deficient. The best understanding of the phenomena would be explanatory. He gives no reason to think so. What could supply a better understanding of the independence of an object’s weight from its rate of fall than Galileo’s demonstration that it could not be otherwise? Granted, that understanding might prompt us to wonder why it could not be otherwise. And to understand that would likely require a good deal of theorizing. So embedding the demonstration in a tenable theory would be worthwhile. But that does not show that the unembedded understanding that the demonstration provides is in any way second best.
Grimm (2008) advocates expanding our conception of explanation so that Lipton’s demonstrations count as explanations. Then to demonstrate that p is to explain that p. This seems ad hoc. Explanations are discursive; they say why something is so. Demonstrations exemplify; they show why something is so. We currently lack a workable conception of explanation that comprehends demonstration.
Whether or not demonstrations afford the best sort of understanding, whether or not they are (perhaps in an extended sense) explanations, there is little doubt that they afford some understanding of the phenomena they concern. As we have seen, understanding admits of degrees. Children, novices, and cutting-edge inquirers often have a level of understanding that falls short of our (realizable or unrealizable) ideal. That is all that we need. For our purposes here, the issues raised by Lipton’s critics are irrelevant.
Still, the examples come from science. Our question is whether the point extends to history. To answer that, we need some characterization of what is required for the distinctive sort of understanding that works of history provide. I cannot give a criterion for being a work of history, or for constituting a distinctively historical understanding. But some features stand out. A work that embodies and conveys a historical understanding is a selective account of the past. Selectivity is mandatory because a mere amalgamation of everything that is known about an episode of any complexity—the Reign of Terror, the moon landing, the ratification of the Nineteenth Amendment—would be such a motley chronicle of events that it would make no sense. Selectivity fosters coherence. The various elements that comprise a work of history should collectively constitute a coherent account. The account should be as of events structured sequentially in time. Williams (2002, 243) maintains that the account itself should be structured sequentially. If so, it will have a narrative form, with a beginning, middle, and end (White, 1965).
It is not obvious that narrative form is strictly required. Garry Wills’s Lincoln at Gettysburg (1992) is largely thematic. It weaves the Gettysburg Address into several historical narratives: a political narrative, a rhetorical narrative, and a narrative about the nineteenth-century cult of death. Each enriches and deepens the understanding that the others provide. Arguably, all three are needed to establish Wills’s thesis that the Gettysburg Address fundamentally changed America’s conception of itself as a nation. This suggests that a narrative form may be just one, albeit fairly standard, way to structure events in time so as to make historical sense of the episode they belong to. A history should be accurate (Williams, 2002). Just how accurate, and in what respects it should be accurate, is controversial. Thucydides composed the speeches he purportedly quoted in The History of the Peloponnesian War, including Pericles’s famous funeral oration. No one knows what, if anything, Pericles or the other orators actually said. Contemporary historians often use hedging verbs. They say that their protagonists would have done this, or must have known that, or were bound to have accepted the other. These locutions signal that although the historians have no specific evidence to back their claims about the protagonists, such actions and attitudes were characteristic of people relevantly like the protagonist in the period under discussion. Moreover, like the sciences, histories often make use of schematized models, such as ‘triangular trade’, ‘coup d’état’, and ‘oligarchy’. These, like scientific models, idealize, simplify, and thereby depart from truth.2 When they contribute to historical understanding, they are felicitous falsehoods. Rather than saying that historical accounts must be true, it is preferable to say that they must be true enough. Moreover, a work of history must be backed by evidence, and the chain of evidence should bottom out in primary source material—firsthand testimony, contemporary documents, artifacts of the period under investigation. To be sure, historians can draw on the work of other historians. Appeal to secondary sources is not ruled out so long as the chain of evidence, however long, terminates in acceptable primary source material. Histories, being empirical, must be defeasible. Information about the past must be capable of discrediting a historical account. I will call anything that embodies and conveys a historical understanding a work of history.
Nonnarrativity
If what is suitably accurate is true enough, documentaries such as The Sorrow and the Pity qualify as works of history. Although they are not primarily verbal, they are nonfictional cinematic representations that provide a historical understanding of the episodes they concern. I will go further and urge that static works can also provide historical understanding. Before I do so, however, I want to return to the scientific cases discussed above.
The manipulation of the orrery shows why Mars seems to change direction. It highlights consequences for terrestrial observers of the fact that the Earth travels around the sun more quickly than Mars does. Galileo’s thought experiment is a reductio. It makes manifest the inconsistency inherent in the assumption that gravitational attraction and weight are linked; that is, it shows that the assumption cannot be correct. Feynman’s demonstration shows why the O-rings could not create a seal: the material, having lost elasticity, could not expand to seal the gaps that the O-rings were supposed to seal.
In all three cases, the primary mode of reference is exemplification. The orbs on the orrery exemplify motions; the postsubmersion O-ring exemplifies inelasticity. They instantiate and thereby refer to some of their properties, making them salient, and enabling interpreters to appreciate them and their significance. As we recognized earlier, in Galileo’s thought experiment nothing actually falls. Since nothing has the property of falling, nothing exemplifies that property. Rather, object-representations (either mental representations, or publicly available descriptions or depictions) exemplify an abstract mathematical property that they share with actual falling bodies. The thought experiment exemplifies the independence of the abstract mathematical property that in material objects is a measure of weight from the abstract mathematical property that in material objects is a measure of rate of fall. It thereby exemplifies the independence of weight from gravitational acceleration. In each case, the symbols that embody and convey an understanding of a phenomenon are not primarily discursive or representational. They afford understanding via exemplification.
I have argued that in the arts, nondenoting symbols embody and convey an understanding of a topic (Elgin, 2002). It would not be surprising, then, that works of art can afford an understanding of the past. This involves more than conveying discrete bits of knowledge about the past. It is easy to see that a Gainsborough portrait conveys the knowledge that eighteenth-century British merchants wore ruffles, and that a Heda still life conveys the knowledge that in the seventeenth century the Dutch had access to lemons. These are isolated facts. They provide evidence about the past. A historical understanding, on the other hand, is an integrated account of a past episode that involves a sequence of events. So to claim that works of art embody an understanding requires saying more than that they convey separate factoids about the past or even that they convey factoids that can serve as evidence about the past. And to say that they embody a historical understanding requires saying that they do so in terms of a selective sequence of events. A documentary plainly satisfies these requirements. I suggest that a static work of art can do so as well.
Monuments are works of public art, expressly designed to commemorate what are taken to be historically important episodes. If any works of art embody and convey historical understanding, monuments are apt to do so. I will argue that Maya Lin’s Vietnam Memorial (henceforth the Memorial) does exactly that.
The Memorial is located on the Mall in Washington, DC, a long parkland dotted with monuments that collectively constitute a patriotic gloss on US history.3 The Mall is not flat. At the Memorial’s location there is a hill. But the Memorial is not on the hill. Rather, it is gouged out of the hill. It is a gash in the land, an open wound. That wound in the physical land is a metaphor for a wound in the country, the political land. The Memorial is made of black stone. It consists of two conjoined triangles that share an edge and open out toward the Mall in a V shape with a 125° angle. One triangle points toward the Washington Monument; the other, toward the Lincoln Memorial. The Vietnam Memorial thus locates the episode it commemorates with respect to the fundamental ideals on which America was founded. It alludes to those ideals.
Etched in the stone are the names of the 58,272 US military personnel who were killed or declared missing in action during the Vietnam War. The names are all the same size and are in the same font. Rank is not included. Every one deserves to be remembered. Each is represented as being as important as any other. The recording of the names thus exemplifies the ideal that all are created equal, an ideal articulated in the Declaration of Independence, operationalized in the Constitution, and reinforced in Lincoln’s Gettysburg Address. The stone is highly polished; viewers see their own reflections as they look at the names. This invites the living to identify with the dead and to reflect on their own attitudes toward the Memorial and what it commemorates. The order of the names is the order in which their bearers died or went missing. The shape of the Memorial displays the course of the war, starting small in 1959, ascending to the apex in mid-1968, then descending to 1975—from a trickle to a flood to a trickle. The configuration thus constitutes a narrative arc, displaying the trajectory of the war in terms of its American casualties. Collectively the names represent the war’s cost, not in money, but the irrecoverable cost in lives.
The Memorial also symbolizes American ambivalence about the war. Among the ideals that were ensconced at the founding of the country are freedom of speech, freedom of assembly, and the right to petition the government for redress of grievances. The Memorial’s pointing across the Mall to the Washington Monument and the Lincoln Memorial can be read as alluding to the patriotism of those who exercised these rights by assembling on the Mall to protest against the war. Then there is the wound. Is it simply a mark of the loss of lives? Is it an honorably acquired war wound? Or is it a wound on the honor of the country that professes ideals that it does not live up to? Did the war instantiate the ideals or betray them? The legacy of the war is controversial. The ambiguity of the gash expresses the nation’s ambivalence about the war.
It seems obvious that the Memorial conveys an understanding of the war. But does it convey a historical understanding? Let’s look back at the touchstones mentioned above. The Memorial is structured sequentially in time. The names are listed, not in alphabetical order, or by rank or unit or age or state, but in the order in which their bearers died or were declared missing in action. So you can read the course of the war by reading the names in order. The names could have been put in columns of the same height. Instead, there is a quite literal narrative arc, peaking at the apex of the triangles, displaying the escalation and de-escalation of the war. The work is grounded in primary sources—the Department of Defense casualty lists and documents pertaining to those missing in action. The work is defeasible. It would be discredited if, for example, the list was irresponsibly augmented or truncated (‘irresponsibly’, because a few erroneous inclusions or exclusions would presumably not discredit the work) (Kvanvig, 2003).
The work is plainly selective. Much is omitted. There is no recognition of, for example, the geopolitical context of the conflict, the Machiavellian machinations of politicians, the perspective of the Vietnamese people. But all works of history are incomplete. There is simply too much information about any historical episode to incorporate it all into a coherent, comprehensible account. The historian has to solve complex problems of selection—what to include and what to omit, at what level of precision to present the material, from what orientation to address the issue, what themes to emphasize. She must, that is, decide what features of the episode her account will exemplify. The Memorial omits any explicit recognition of the reasons for the war. It tells its story by counting the cost. And it measures the cost in lives according to the metric ‘all are created equal’. The Memorial does not and does not purport to convey a comprehensive understanding of the war. No single work of history could do that. Still, you will enrich your understanding of this period in US history, if you properly interpret it and incorporate what it conveys.
We might doubt that an expert on the war could gain an understanding from the Memorial. He would presumably already grasp the insights and orientation that the Memorial presents. This is not so obvious. What I have called the narrative arc of the piece might make manifest aspects of the escalation and de-escalation that he had overlooked. If so, the Memorial would function for the historian in much the way that a work written by another historian would. It would serve as a secondary source. Still, it is perhaps unlikely that the understanding of a scholar who specialized in the war would be much enhanced by the work. If the Memorial is a work of history, it is a work of popular history, functioning not primarily to deepen the understanding of experts, but rather to deepen the understanding of ordinary people.4
Arguably, then, the Memorial affords historical understanding. It bears at least one interpretation that satisfies plausible requirements for being a work of history. A separate question, though, is whether it does so by functioning as a work of art. Or does it simply play two distinct, mutually independent roles?
Aesthetic Functions
The history of aesthetics is littered with failed attempts to articulate necessary and sufficient conditions for being, or functioning as, a work of art. At best we have clear cases, and what Goodman called symptoms of the aesthetic—that is, defeasible but presumptive indications that an item is functioning aesthetically. The Memorial displays Goodman’s symptoms. Its doing so underwrites its capacity to provide historical understanding. That being so, then at least by Goodmanian standards, it is functioning as a work of history by functioning as a work of art. The symptoms are these:
(1) Syntactic density, “where the finest differences in certain respects constitute a difference between the symbols” (Goodman, 1978, 67–68). Had the stone been less reflective, or the orientation toward the Mall been different, or the Memorial been placed atop the hill, it would have been a different symbol.
(2) Semantic density, “where symbols are provided for things distinguished by the finest differences in certain respects” (Goodman, 1978, 68). The names could have been given with ranks, units, and so on. Fonts could have been different for military personnel of different ranks. That is hardly unprecedented. It is what is done in the war memorials in Canterbury Cathedral. The choice to omit such markers was a choice from among a dense range of alternatives.
(3) Relative repleteness, “where comparatively many aspects of a symbol are significant” (Goodman, 1978, 68). As we have seen, a vast number of features of the Memorial symbolize: its position, orientation, shape, configuration, the order of names, the font, the polish of the stone, and so forth.
(4) Exemplification, “where a symbol, whether or not it denotes, symbolizes by serving as a sample of properties it literally or metaphorically possesses” (Goodman, 1978, 68). The number of casualties, the rate of casualties, the names of the casualties, the relation of the Memorial to other monuments, and the shape of the Memorial are literally exemplified; the ideals, realized or betrayed, the wound in the body politic, and the ambivalence about the war are metaphorically exemplified.
(5) Complex and indirect reference, “where a symbol performs several integrated and interacting referential functions, some direct, some mediated through other symbols” (Goodman, 1978, 68). The triangles pointing toward the Lincoln Memorial and the Washington Monument make direct reference to those monuments. Via that reference, the Memorial makes indirect reference to the ideals that the two monuments stand for. It indirectly refers to their metaphorical exemplification of ‘all are created equal’, and in turn refers back to the exemplification of that ideal in the names on the Memorial itself. It refers to the exercise of the constitutional rights by the antiwar protesters and both makes reference to and calls into question the nature of the wound symbolized by its shape and position. And so forth. This makes it plausible that it is by functioning as a complex aesthetic symbol that the Memorial conveys an understanding of history.
One might, however, wonder whether the Vietnam Memorial is a special case. In some respects the answer may be ‘yes’. The obvious narrative arc is distinctive. But as I suggested earlier, narrativity may be optional. Then the relevant requirement is that a work of history be “a selective account of the past, structured sequentially in time, that is designed to make sense of it” (Williams, 2002). Regardless of what we say about how representative the Vietnam Memorial is, I do not mean to suggest that every memorial or monument functions as a work of history. A work can commemorate an event without affording any historical understanding of it. But the criteria for a history can be realized at different levels of generality or abstractness. So a suitably accurate memorial that commemorates wars or battles, or the reigns of kings or the tenures of poets laureate in a temporal sequence, might function as a work of history, even though it lacks the specificity of Lin’s Memorial.
Moreover, a work of art that is not expressly designed to commemorate a particular historical episode might plausibly function as a work of history. Consider, for example, The Disasters of War, a series of prints Goya made in response to the Peninsular War (1807–1814) and its aftermath. They focus on the horrors of war: humiliation, degradation, rape, torture, mutilation, desecration, destitution, death. They concern not the causes of war, but the effects—in particular, the effects on the innocent: women, children, noncombatants, priests. The prints open out into an indefinite space, suggesting that the mayhem flowing from war is random. It could occur anywhere to anyone. There is no differentiation between friend and foe. Atrocities are perpetrated by both sides. Nothing in the prints points to hope, charity, or even rationality. War is monstrous. It strips people of reason and empathy and turns them into savages. Many of the prints can be correlated with specific events in the Peninsular War. Interpreted as a commentary on that war, they function in much the way the Vietnam Memorial does. But the series might also be interpreted more generally. It is a template or schema into which you can insert any war you please: the Trojan War, the Anglo-Boer War, the Chaco War, the Vietnam War. It makes no difference. The atrocities Goya depicts will be found there too.
Does The Disasters of War, interpreted as a general comment on war, convey a historical understanding? Maybe. If it is true that wars generally instantiate the depicted atrocities, then the series is a schema. Each instance affords a historical understanding of a particular war. And if that is so, then it is reasonable to say that the schema figures in a historical understanding of war as such. According to Hempel (1965), lawlike generalizations about human conduct figure in or are tacitly presupposed by explanations in history. Treated as a schema, The Disasters of War is akin to those generalizations. If Hempel’s generalizations figure in an understanding of history, it is plausible that Goya’s prints do too.
I have been assuming that a work of history can be purely descriptive. It can present a cogent account of what happened. It need not provide a causal explanation of why it happened. If this is right, then Hempel’s generalizations need be no more than statements of robust Humean regularities. They may afford a basis for prediction, but need not contribute to explanations of how it came to pass that events played out in the way they did. So although Lin’s Memorial provides no explanation of the course of the Vietnam War, and Goya’s prints provide no explanation of why wars lead to the commission of atrocities, that is not by itself an argument against treating them as affording a historical understanding. Arguably they function like Lipton’s orrery or Feynman’s O-ring.
Maybe Not
Even so, we might want to resist the claim that a work of art, functioning as such, affords genuinely historical understanding. It is not implausible to think that aesthetic standards are too subjective to underwrite the epistemic standing history demands. Earlier I argued that aesthetic standards are not wholly subjective. As we have seen, in discussions about art, reasons are adduced and disputed. Still, the range of reasons aesthetic judgments can appeal to may seem too open ended to satisfy the disciplinary demands of history. Thus far, my discussion has appealed to loosely characterized touchstones. Perhaps we should tighten them, excluding symbols whose function turns on their displaying objectionable symptoms of the aesthetic.
One worry is that works of art often admit of multiple acceptable interpretations. Written histories strive for and often achieve univocality. Different histories may provide different interpretations of a given event. But each individual work is apparently supposed to provide only one. We can accommodate this worry by saying that it is only under an interpretation that a work of art functions as a work of history. This is a trivial concession, since it is only under an interpretation that a paper in the Journal of American History functions as a work of history. Only interpreted symbols are about something.
Perhaps the point is that unlike the journal article, the work of art lacks a standard interpretation. Every competent historian knows how to read the journal article, and they all pretty much agree about what it says. The same cannot be said for a work of art. Connoisseurs, such as Greenberg and Fry, may permanently disagree. If this is the worry, we could insist that only under a relevant, accessible, nonidiosyncratic interpretation can a symbol function as a work of history. Except perhaps for the bit about the antiwar protests, my interpretation of the Vietnam Memorial satisfies this requirement. It is a standard, relevant, accessible, and nonidiosyncratic interpretation of the monument (see Lin, 2000). It is intersubjectively accessible and assessable. Reasons can be given for it and for the interpretation of the war that it provides. So the restriction would not exclude works of art that have standard interpretations from functioning as works of history under those interpretations. In any case, the restriction is ill advised. An account of the siege of Stalingrad written in an obscure code would be nonstandard, but assuming it was duly researched it would be a work of history. The issue is not whether a work has a standard interpretation, but whether it has an intersubjectively accessible and assessable interpretation. Insofar as the code could be learned, an otherwise historically creditable account of a past episode written in code would qualify as a work of history.
Maybe the worry is not semantic, but syntactic. Should we perhaps deny that works of art afford historical understanding because they are not verbal? There is, as far as I can see, no reason to privilege the verbal over other modes of symbolization. And if we do, we exclude not just prints, monuments, and documentary films, but also diagrams, charts, and maps. To restrict historical understanding to that which is captured in a language would be costly. We would have to deny, for example, that Minard’s map affords an historical understanding of the fate of Napoleon’s army in the Russian invasion. Semantic density seems both innocuous and unavoidable. Syntactic density is sometimes welcome.
I suggested earlier that the historian’s solution to the problem of selection is a matter of deciding what aspects of an episode are to be exemplified. He chooses events and event descriptions that highlight the features of the episode that he considers central to the sort of understanding he seeks to provide. This determines the items he includes, the vocabulary he uses to describe them, and the orientation he takes toward them. Describing the construction of the first transcontinental railroad, Howard Zinn says this: The Central Pacific Railroad
spent $200,000 on bribes in Washington to get 9 million acres of free land and $24 million in bonds, and paid $79 million, an over payment of $36 million, to a construction company which was really its own. The construction was done by three thousand Irish and ten thousand Chinese, over a period of four years, working for one or two dollars a day. (Zinn, 1980, 254)
This description configures the facts in such a way that they exemplify the plutocrats’ venality, corruption, collusion, and callous indifference both to workers, who were members of despised minorities, and to ordinary citizens, who might reasonably object to their legislators being bribed, their land being given away, and bonds being sold to them at artificially inflated prices. It is hardly laudatory, but that does not discredit Zinn’s description. The passage exemplifies a pattern to be found in behavior of the plutocrats behind the construction of the first transcontinental railroad. This is not the only pattern. A different account might highlight the technological achievements, the logistical difficulties, the economic benefits the railroad supplied, or even cast the protagonists in a less unfavorable light. Then different features of the construction of the railroad would be exemplified.
Some histories describe seemingly insignificant or irrelevant items in order to exemplify something of importance. Early in Lincoln at Gettysburg, Wills describes the layout of the Gettysburg cemetery. The Union dead are buried according to the states they came from, with no recognition of differences in importance either of the state or of the soldiers. “By considerable ingenuity in grading the cemetery’s incline, and by arranging the graves in great curving ranks, [William Saunders, the cemetery’s architect] avoided preferential treatment of states or inequality in the ranks of the fallen” (1992, 22). The book’s central theme is that the Gettysburg Address changed America’s self-conception from a union of states to a union of free and equal citizens. The cemetery design and Lincoln’s speech, Wills suggests (1992, 30), are mutually reinforcing because they exemplify the same values.
Exemplification is thus critical to historical understanding. The historian juxtaposes events, agents, actions, and standard conditions so that collectively they exemplify a pattern or trajectory in the course of events. An understanding of an episode must make sense of it. To do that, it must highlight some features and ignore or downplay others.
Written works of history typically deploy a variety of modes of reference. They denote the events and agents they concern. They exemplify features and patterns in the play of events. They refer indirectly to events, actions, and authorities that are not explicitly mentioned. They cite their sources and tacitly allude to whatever makes those sources credible. They may also refer indirectly to other histories that support or clash with the position they are taking. This can be done by a variety of means, including the selection of events deemed worthy of mention, the order of exposition, even word choice. Some indirect reference is unavoidable. A history of the Bogside Massacre must say where it took place. If it locates the event in Derry, it thereby implicitly slants toward the Irish; if in Londonderry, toward the British. Either way, it indirectly refers to a history of animosity. Each name is freighted with significance, making partisan reference to Irish history. No neutral name is available. Complex and indirect reference too seem unavoidable.
Goodman’s remaining symptom is relative repleteness. A symbol is relatively replete to the extent that it symbolizes along multiple dimensions. Goodman illustrates the difference between repleteness and attenuation by comparing a Hokusai drawing of Mount Fuji with an electrocardiogram (EKG). Even if we imagine the same configuration of ink on paper playing both roles, there is an enormous difference in the ways they symbolize. Only two features matter in the EKG—the shape of the wave and the frequency with which it repeats. In the drawing, any of a vast number of features might matter: the size, shape, color, and intensity of the line; the weave, texture, and material of the paper; differences, however subtle, in shading, from one part of the picture to the next; and so on. The picture is inexhaustible. There is always more to be found.
One might think the same holds for the EKG. In principle, there may also be subtle, medically significant differences in wave patterns that cardiologists have previously overlooked. So it too bears repeated inspection; but only up to a point. The number of dimensions along which it symbolizes is fixed in advance. The cardiologist does and should overlook the fact that the line gets paler as the test continues. That is an indication that the printer’s toner is running low. It is not her problem. Moreover, the machine has an antecedently determined level of precision. Past a certain point, further wobbles in the curve are insignificant. They might be manifestations of background vibrations or noise in the system or fine-grained irregularities in the paper the graph is printed on. They should be disregarded. Perhaps one could glean a bit more information than we currently do from a given EKG, but cardiologists rightly think that the information the printout provides is exhaustible. If they want to learn more, they move on to different tests.
What should we say about works of history? Written histories are, I suspect, more replete than EKGs. But they are apt to be less replete than works of art. Although they symbolize along several dimensions, arguably at least, the number of dimensions is limited and intersubjectively agreed upon in advance. That a history includes or omits the names of the founding members of the Royal Society may be significant. Even the order in which the names are listed may be significant. But it is doubtful that the omission of maiden names of the founding members’ maternal grandmothers is significant. If we are loath to consider works of art to be works of history, perhaps the reason lies here. When interpretation is too open-ended, intersubjective agreement is sacrificed. Scholars become permanently at odds as to what a particular work conveys about the past.
I have argued that a fundamental difference between scientific and aesthetic understanding lies in the fact that science puts a premium on intersubjective agreement; aesthetics does not. To attain such agreement, science limits precision and constrains repleteness. If only a few antecedently recognized dimensions of a symbol matter, and only up to an antecedently recognized point, agreement among the knowledgeable is readily achieved. Scientists build on one another’s findings. It is therefore important that they can tell exactly what those findings are and why and to what extent they are to be trusted. They sacrifice repleteness and precision to achieve agreement. Because artists do not build on one another’s works in the same way, they have no incentive to make such sacrifices. That connoisseurs never agree about whether the weave in the paper of the Hokusai drawing or the direction of Le Compotier’s brush strokes contributes to its aesthetic function makes no difference. Artists who are influenced by Hokusai or Cézanne can decide for themselves whether to take that feature of the work to be significant. Historians fall somewhere in between. They do not build on their predecessors in the way that empirical scientists do. So agreement about what exactly a given historian accomplished is not mandatory. But it is not irrelevant either. What has been done sets the stage for what comes next. So perhaps a reluctance to treat works like the Vietnam Memorial or The Disasters of War as works of history stems from a well-founded desire to achieve a measure of accord. In that case, we should say that although such works embody and convey something close to a historical understanding, that understanding remains beyond the boundaries of history proper.
The question is where to draw the line. We could decide that because of their potentially unbounded repleteness and precision, works of art do not afford genuinely historical understanding. Or we could decide that repleteness and precision need not inhibit and may even enhance their capacity to afford historical understanding. The decision turns not on the sort of information the items convey or on the accuracy or reliability of that information. It turns on how historians want to configure their discipline—what framing commitments they want to endorse, what restrictions they will place on the accounts it sanctions, and so on. And that turns on how they want to use those accounts. If historians want to be able to build on previous histories in the way that scientists build on previous science, they should limit precision and repleteness. If the growth of historical understanding takes a different path, then perhaps not. The decision depends on the relation of scholars to one another, not on their relation to their object of study.
Notes
1. Both Grimm (2008) and Lipton (2009) require that the agent who understands is not acting by rote. Her capacity to perform correctly must be stable across a range of counterfactual circumstances.
2. I am grateful to Kenneth Walden for this point.
3. I am indebted to Ajume Wingo for this interpretation of the Mall.
4. This is not to say that all works of art that afford historical understanding do so only or primarily for the inexpert. Some works—for example, those with religious themes—convey understanding only to those with considerable expertise.