It is unusual, to be sure, to offer a set of reflections such as these after the conclusion to a book, all the more so when they deal with the method putatively used throughout the book. Of course, it is almost as unusual—and perhaps a bit self-indulgent—for a historian to take the time to lay out his or her guiding theoretical principles, primitive as they may be.
And yet I take seriously Samuel Moyn’s admonition that we all too often outsource this task to professional philosophers of history.1 As in other fields, this has generally been the case in Jewish history, though there are a number of notable exceptions, such as Salo W. Baron in The Contemporary Relevance of History (1986) and Moshe Rosman in How Jewish Is Jewish History? (2007).2 I, for my part, am not a philosopher, nor is this book intended as a full-blown philosophy of history. But it does seem responsible to take up Moyn’s challenge to try to articulate how I go about the work of historical interpretation.
From the advent of my formal professional training, I have been drawn to two conflicting sensibilities: a strong belief, nurtured by the riches of the archive, that the historian must seek to reconstruct the past meticulously in the quest to approach truth; and, in tension with that belief, a persistent pull toward various forms of constructivist accounts of historical method focused on the recognition that we are unable to do what Ranke aspired to do as a historian: “extinguish myself in order that the things could speak [for themselves].”3 The tension between these two sensibilities became clear to me upon first encounter with Roland Barthes’s 1967 essay “The Discourse of History.” It was a jolt to a young graduate student to have to grapple with Barthes’s suggestion that “the narration of past events” was not different “in some indubitably distinctive feature, from imaginary narration, as we find it in the epic, the novel, and the drama.”4 Barthes’s identification of narrativity as the shared and open border between history and fiction produced in me less a belief in the unavoidable indeterminacy of historical interpretation than a newfound sensitivity to the process of crafting historia rerum gestarum, the narration of the events of the past.
One of the most significant elucidators of Barthes’s position was Hayden White, whose monumental Metahistory expanded our understanding of the nature of modern historical writing by categorizing works according to differing modes of emplotment, argument, and ideology. With respect to the last category, ideology, White argued that nineteenth-century historians were ideological not so much by virtue of their political or polemical stances but to the extent that they came to regard the form of the historical narrative as “a content or essence.” In doing so, they conflated historia rerum gestarum, the narration of the past, and res gestae, what actually happened.5 Roland Barthes and then White flipped this somewhat unwitting conflation on its head by asserting that historical narrative bore strong similarities to and, in fact, assumed the properties of fictional narrative. In doing so, they challenged the underlying logic of a naive historical “realism,” as well as its closely related source of methodological validation, objectivity.6
I was intrigued and compelled by this kind of sharp, semiotically inspired critique of my emerging craft. Over time, though, I came to realize that it was hard to follow or accept fully their analysis of history as akin, in form, to fiction—in no small part because I clung to a shred (or more) of the realist credo that guides most practicing historians.7 In forging my own view of the matter, I was stimulated and challenged both by White and by his forceful critic, Carlo Ginzburg, to whom reference was made in Chapter 3. The polemic between them at the 1990 conference Nazism and the “Final Solution” and then in the ensuing volume, Probing the Limits, carved out a nuanced historiographical space between the poles of an unvarnished realism on one hand and wholesale surrender to the literary, imagined, and fictive qualities of historiography on the other.8
In seeking to understand that position more fully, we might profitably turn back to some of the most incisive thinkers about history from the past century. One recalls, for example, the Italian philosopher Benedetto Croce, who insisted, “Where there is no narrative, there is no history.”9 Closely aligned with Croce was the Oxford philosopher and historian R. B. Collingwood, mentioned in the Introduction, who translated Croce and was in correspondence with him. Both men affirmed a simple and intuitive claim: the historian stands at the center of historiographical production. For Croce, this principle was best expressed in his 1912 aphorism “All history is contemporary history.” The historian’s angle of observation onto the past is shaped by present-day preoccupations that animate and inform his or her work.10
Meanwhile, Collingwood understood that “the only possible knowledge of the past is mediate or inferential or indirect, never empirical.” His particular reformulation of this commonplace notion, in the posthumously published The Idea of History (1946), was the theory of “re-enactment.” The historian, Collingwood wrote, must not merely seek to comprehend an earlier text in exact philological terms; rather, he must understand how and why the author of that text chose to solve the key problems in it in the way he did. This places on the historian the obligation of “re-thinking for himself the thought of his author”—or “re-enact[ing] the past in his own mind.”11
This kind of mental construction, keenly attentive but not altogether subordinate to the primary source, is a fair description of what the historian does. But an objection can be quickly raised: if we turn to history for empirical validation of the past in order to understand the present better, but recognize that our knowledge is simply a construct of the individual historian’s mind, are we not consigned to an incapacitating relativism? Not necessarily, Collingwood responds; with respect to reenactment, we need not assume that “because it is subjective, [it] cannot be objective.” After all, someone who engages in “an act of knowing” can both do that and be aware that he or she is performing that act.12 Moreover, the fact that one reenacts the past does not mean that one invents it out of whole cloth. The quality of one’s historical labors relies on the credible use of sources, analytical rigor, and a compelling narrative, all regulated by the consensual norms of the profession. Those criteria cannot guarantee Truth, but they do differentiate between more and less convincing versions of the past—and help propel us to what Collingwood wisely calls “relative truth.”13