Collingwood, in his reaction against ‘scissors-and-paste history’, against the view of history as a mere compilation of facts, comes perilously near to treating history as something spun out of the human brain, and leads back to the conclusion referred to by Sir George Clark in the passage which I quoted earlier, that ‘there is no “objective” historical truth’. In place of the theory that history has no meaning, we are offered here the theory of an infinity of meanings, none any more right than any other – which comes to much the same thing. The second theory is surely as untenable as the first. It does not follow that, because a mountain appears to take on different shapes from different angles of vision, it has objectively either no shape at all or an infinity of shapes. It does not follow that, because interpretation plays a necessary part in establishing the facts of history, and because no existing interpretation is wholly objective, one interpretation is as good as another, and the facts of history are in principle not amenable to objective interpretation.
E.H. Carr, What is History? Footnote 11. Realism and anti-realism in the philosophy of history
Before explicating my attempt to transcend the realist/anti-realist divide in the philosophy of history or historical theory, it is worthwhile to sketch in broad terms what this divide exactly consists of, and why both sides run into certain difficulties with respect to their conceptions of history – difficulties I aim to resolve with my alternative account. The epistemic dimensions of historiographical practices brings up the question of the status of historical knowledge, specifically whether normativity and aesthetics does not preclude the viability of representing the past in an objective way, and if it does not – as many practicing historians believe – how this can be justified in philosophical terms. This point is captured in what Mark Bevir at times refers to as theory choice in the context of historical theory, denoting the way in which one can adjudicate between rival historical accounts or narratives, the objects with which philosophy of history is traditionally concerned.Footnote 2 This in turn ‘raises the question of what criteria should guide the comparison’:Footnote 3 either it is done on epistemic grounds, as maintained by realists like BevirFootnote 4 and Chris LorenzFootnote 5 ; or on normative and aesthetic grounds, as maintained by anti-realists like Hayden WhiteFootnote 6 and Frank Ankersmit.Footnote 7
Beginning with anti-realism, that of White and Ankersmit are taken as the paradigmatic forms because they represent some of the most influential accounts of it as also acknowledged by their realist opponents, whose ideal-types will be the already mentioned Bevir and Lorenz for the same reason.Footnote 8 White and Ankersmit's anti-realism are not entirely alike, particularly as it concerns their respective normative views that underlie their philosophical ones. White is a self-described LukácsianFootnote 9 who believes that all social theory, irrespective of field or discipline, is inherently ideologicalFootnote 10 in nature by virtue of being value-laden.Footnote 11 In his seminal Metahistory, in which the historical imagination (ideologies) of the great nineteenth century historians and philosophers of history is investigated, it is posited that the narratives they constructed were products of certain modes of emplotment, explanation and ideology.Footnote 12 These are in turn rooted in a deeper level of historical consciousness, shaped by certain types of tropological prefiguration, perceived by White to be rhetorical, poetic (normative) strategies: Metaphor, Synecdoche, Metonymy, and Irony.Footnote 13 Hence he concludes that ‘there are no apodictically certain theoretical grounds on which one can legitimately claim an authority for any one of the modes over the others as being more “realistic”’, for ‘the best grounds for choosing one perspective on history rather than another are ultimately aesthetic or moral rather than epistemological’. This reduces theory choice to a mere normative preference: ‘the demand for the scientization of history represents only the statement of a preference for a specific modality of historical conceptualization, the grounds of which are either moral or aesthetic, but the epistemological justification of which still remains to be established’.Footnote 14 While there have been some changes in White's conception of historical theory since Metahistory – particularly as it relates to the structuralist argument that all historical writing has to conform to a strict set of normative strategies and categories of tropological prefiguring – he has remained a committed anti-realist of the kind typified in his affirmation of being a Lukácsian.Footnote 15
Ankersmit was influenced by White's early work and also adopted an anti-realist stance, emphasizing in particular the aesthetic dimensions of historical narratives or representations.Footnote 16 Describing himself as a ‘postmodernist’ opposed to realist historiography with its outdated modernist (objectivist) pretensions in the early 1990s,Footnote 17 he later on attempted to take a middle position that nonetheless left his anti-realist conception of theory choice and the determinative role played therein by aesthetics unchanged.Footnote 18 In his early Narrative Logic he sought to uncover the nature of historical narratives – what he terms ‘narratios’ – by perceiving them as consisting of ‘narrative substances’. These denote a set of historical events as presented by historians – so-called ‘theses on the past’ – such as ‘Cold War’, ‘Industrial Revolution’, and ‘the Renaissance’.Footnote 19 These narrative substances and the narratios they constitute do not refer to reality as for example concepts like ‘this chair’ do. Instead they are self-referential, existing on a separate ontological domain (linguistic in kind) from the irretrievable past, meaning that ‘the objects in the past so often mentioned by historians, objects such as intellectual, social or political movements and even nations or social groups, have no status in the past itself independent of the narratio: they spring from and are justified by the narratio’.Footnote 20
The criteria for distinguishing narratios are aesthetic in nature, for by severing them from the ontological domain of reality – a notion of the past ‘as it really was’ – there is only an aesthetic preference for this rather than that narrative that remains open to us. Ankersmit's own aesthetic preference is for the narrative that ‘takes most risks and is most courageous’.Footnote 21 It is hard to miss the presence of an elitist crypto-normativity in this conception, which restricts proper appreciation of historical narratives to those who possess a certain kind of cultivated aesthetic disposition. This has been pointed out in particular in relation to Ankersmit's later work on the sublime historical experience.Footnote 22 As with White, there have been some changes in Ankersmit's views since Narrative Logic. Already mentioned was the softening of his self-proclaimed ‘postmodernism’ in Historical Representation, wherein he sought to find a middle ground between the realist and anti-realists positions. However, he remained committed to the notion that historical representations exist on a separate ontological plane, which precludes their description as real representations of the past in any meaningful sense. Instead the aesthetic rather than cognitive value of a historical narrative is determinative in theory choice, with such aesthetic criteria as ‘scope’ and ‘boldness’ identified for distinguishing between competing historical representations.Footnote 23
*
In order to understand the development of post-foundational realism in historical theory it is useful to briefly sketch its philosophical ancestor, in opposition to which it arose. Foundationalist realism in historical theory was premised on the belief that a monistic scientific model could be developed to capture the true nature of historical reality. For those influenced by Rankean historism this meant a way of writing so as to represent history wie es eigentlich gewesen (how it actually happened).Footnote 24 Hempelian ‘critical philosophy of history’ was a particularly influential expression of this kind of realism. It was dominant in historical theory in the first decades of the postwar era, and still maintains a lingering influence in some corners among adherents of foundationalist realism and practicing historians, who tend to be unconcerned with the philosophy of their discipline. It attempted to furnish the scientific methods allowing for the writing of objective histories, such as the later much maligned (both in the natural sciences and humanities) covering law model.Footnote 25 With the steady critique of positivism and the foundationalist premises it was based on, realist historical theorists began to question it as well, eventually coming to reject it in favor of a post-foundational alternative. They came to embrace a conditional conception of ontology and epistemology, which views all knowledge as being socially mediated, yet still capable of yielding conditions allowing for a notion of objectivity to subsist by locating it in communities. ‘To put it crudely, ideas, words, and language are not mirrors which copy the “real” or “objective” world but rather tools with which we cope with “our” world’, as Cornel West describes it.Footnote 26
The reorienting of the locus of knowledge as emanating from communities – or to speak with the late Wittgenstein Lebensformen Footnote 27 – brings an analysis of these to the forefront in uncovering the nature of our shared epistemological and ontological beliefs, which has been picked up and extended in realist historical theory.Footnote 28 Yet it remains contentious how exactly such an investigation should proceed, and whether or not it is even possible to maintain a realist position by finding such basic elements of our being in the world to base it on. In order to elide the relativist threat, post-foundational realists have to maintain a conception of realism that manages to stake a viable claim to being intersubjectively shared across otherwise disparate communities. They do so by referring to common aspects of our being in the world that exude certain general features of knowledge such as self, object, object permanence and causality – so-called conceptual primitives.Footnote 29 These arise out of the common physiology we share when we engage in practices in otherwise radically varied forms. As the influential post-foundational historical theorist Bevir argues, we all require such basic goods as food and water, and we all interact with our environment in a realist sense – by way of conceptual primitives – so as to fulfill such needs.Footnote 30 It is this that provides the necessary elements of what he calls an anthropological epistemology. Bevir and Lorenz's post-foundational realism is therefore rooted in a primitive or intuitionist conception of the basic aspects of human cognition that allows for a realist understanding of ontology and epistemology to get off the ground.Footnote 31 Aside from such commonalities there are also differences between them. Bevir is greatly influenced by W. V. O. Quine and Donald Davidson, and presents a comprehensive post-foundational realist position in the philosophy of history.Footnote 32 Lorenz relies heavily on Putnam and provides a less substantive historical theory, his work in the field being limited to several polemical interventions in defense of realism.Footnote 33 My primary focus in taking on the post-foundational realist approach will therefore be Bevir, though Lorenz offers important insights as it relates to historiographical practices such as those classified as the Historikerstreit, which I draw on extensively in the final section.
As noted, Bevir and Lorenz have to propose viable criteria for theory choice capable of withstanding the problem of normativity if they are to defend against anti-realist critiques. The source of these criteria has already been mentioned, namely the nature of being in the world, but Bevir and Lorenz differ in where they are to be found. Bevir argues that it is possible to derive out of the shared aspects of our web of beliefs that emanate from our being in the world certain criteria of judgement that are intersubjective and hence can serve as objective criteria for theory choice.Footnote 34 These are rooted in the realist assumptions we all share and come to the fore in everyday life, such as the belief in the existence of objects.Footnote 35 At this stage we still only have a thin conception of objectivity, but Bevir argues that by further unpacking these assumptions we can arrive at one that is sufficiently thick to allow us to adjudicate between rival histories.Footnote 36 In order to accomplish this and develop such a substantive anthropological epistemology, he has to flesh out what he includes in his conception of being in the world, adding to it the natural ability of identifying the best belief by means of certain substantialized criteria of judgement.Footnote 37 As is elaborated in the following section, it is at this point, when the move is made from what can be described as a thin to a thick conception of objectivity, that I depart from Bevir's approach. What results from this move is the following list of criteria: accuracy, comprehensiveness, consistency, progressiveness, fruitfulness, and openness. These allow for theory choice among rival historical accounts on intersubjective grounds, thereby saving an objective conception of historical knowledge which is irreducible to normativity.Footnote 38 Bevir's methodological pluralism logically follows from this, for in finding the criteria of theory choice in our shared web of beliefs rather than a specific method, he has effectively dissipated the latter's value in understanding historiographical practices. It is not the means by which historical narratives are constructed that matter, but rather their existence as such, for it is only that which is susceptible to theory choice by means of the proposed criteria.Footnote 39
Lorenz's account is as noted less philosophically elaborate but comes to the same overall answer to the question of theory choice and the role played therein by normativity. Relying on Putnam's concept of internal realism, which posits that the truth conditions of theories and statements are relative to conceptual schemes, Lorenz argues that the anti-realist challenge poses no problem for the post-foundational realist. It is aimed at the wrong target: a positivist conception of knowledge which denies internal realism, instead positing that truth conditions require a direct access to reality (from referent to referred). In fact, by missing its target, the anti-realism of White and Ankersmit should be seen as merely inverted forms of foundationalist logic.Footnote 40 Rather, it is possible to find appropriate criteria for theory choice within the intersubjectively shared conceptual framework of historians. This must be done by expanding our purview of justification in historiographical practices to include the normative dimension alongside the factual or epistemic, while moving our perception of the latter to a post-foundational rather than foundational conception of it – dependent on our conceptual frameworks and hence subject to contestation rather than being absolute.Footnote 41 When this is done, it is possible to demarcate certain ‘truth-tracking’ criteria which are immanently present in historiographical practices, for the claims that are made thereby are truth-claims referring to an extra-textual reality (‘the real past’).Footnote 42 Lorenz does not spell out what all these criteria are, but does mention a few examples which are similar to Bevir's, such as scope, explanatory power, and comprehensiveness. As noted, these are also grounded in an intersubjective aspect of our being the world, though for him this is specifically located in the intersubjective controls imposed by the community of historians.Footnote 43 What is missing in Lorenz's account is an explication of the origins of these intersubjectively imposed standards of theory choice, and a substantive philosophical defense of it. Instead, the philosophical aspect of theory choice is conflated with the sociological, an error also committed by anti-realists to which I return in the final section regarding the Historikerstreit.
2. First-order realist and second-order normative/aesthetic dimensions of historiographical practices and ‘normal history’
Having clarified the realist and anti-realist positions, I now turn to explicating my proposed alternative philosophy of history that transcends the divide separating them, which hinges on the concept of ‘normal history’. The positing of it was inspired by Thomas Kuhn's concept of normal science, delineating the ‘body of accepted theory’ in a given scientific community and the epistemological and ontological commitments involved in its establishment as such.Footnote 44 Richard Rorty similarly spoke of normal discourse, meant to delineate ‘any discourse (scientific, political, theological, or whatever) which embodies agreed-upon criteria for reaching agreement’.Footnote 45 The concept's meaning and purpose is to describe a certain stability in beliefs inhering in a given community over time, premised on intersubjectively shared epistemological and ontological commitments. I use the term normal history to denote the same in the context of historiography. What I call ‘normal history’ is the product of the first-order realist as well as the second-order normative beliefs of historians, embodied in their dispositions and by extension the community of historians or the historian's field that is made up of them. Unlike Ankersmit's ‘narrative substances’ it is therefore not posited as a purely linguistic entity, but is also institutional and social in kind.Footnote 46 It denotes the status quo of historical knowledge at any given moment in time, on any given topic, as is established in contemporary historiography – the ‘body of accepted historiography’ as it can also be called. It does not refer to ordinary conceptions of history as they exist in a given community, what I term folk histories. Rather, normal history is officially sanctioned as such and promulgated by the historian's field, specifically the institutions that disseminate the knowledge it procures and produces via historiographical practices (e.g. history lessons in schools, journals, articles and conferences).Footnote 47 Its existence is therefore based on and requires a more or less autonomous historian's field, one that is unimpeded by other fields in society (most importantly the political, but also the economic), which act as distorting influences interfering with its otherwise stable epistemic foundations inhering in its first-order dimension, as I enumerate below.Footnote 48
In terms of normal history's composition, the first and second-order beliefs inhering in historiographical practices are homologously reflected in it. It is therefore possible to distinguish between the two dimensions of historiographical practices by delineating how the first and second-order aspects of them operate within the broader economy of practices of the historian's field. This in turn offers insights into the epistemic and normative/aesthetic dimensions of historiographical practices; how they relate to one another and, most importantly for the purposes of my alternative philosophy of history, whether it is possible to delineate certain intersubjectively shared epistemic criteria for theory choice within them.
To begin with elaborating what I have described as the first-order realist dimension of normal history, it comprises what Bevir terms the ‘generally accepted facts’ allowing for the comparison of rival webs of beliefs,Footnote 49 and Ankersmit describes as the ‘singular statements’ out of which narrative substances are made up.Footnote 50 They are the historical facts on any given topic that can be found in compendiums such as Cambridge histories and standard textbooks, and are generally shared by the community of historians irrespective of their second-order normative views. It denotes, to use Quine's phrasing,Footnote 51 the backlog of accepted history, what any practicing historian writing on any given topic must study and include in their own work. There is then a general consensus among historians – derived from their being in the world – legitimizing the epistemic status of these historical facts qua historical facts. Carr describes the acknowledgement of historical facts of this kind as the necessary but not sufficient condition for engaging in a historiographical practice, which, when not properly adhered to, excludes one from being recognized as a member of the historian's field.Footnote 52 This is what makes it lexically prior to second-order normative beliefs. An example of this discussed in more detail below is Holocaust revisionism, which violates the first-order epistemic standards inhering in the historian's field and thereby leads to exclusion from it.Footnote 53 However, these historical facts comprising the first-order realist contours of normal history are not value-free and absolute, granting them the epistemological and ontological status of atomic, foundational facts as is the case with Ankersmit's singular statements. What Bevir describes as the lingering modernism,Footnote 54 and Lorenz as the inverted positivism of White and Ankersmit is based on such a foundationalist conception of historical facts.Footnote 55 They require them to possess this epistemic status in order to viably denote realism, leading to White's rejection of it on these grounds alone,Footnote 56 whereas Ankersmit, who as noted does believe they have this status on the level of singular statements, believes it is negated on the level of the narrative.Footnote 57
However, such a positivist conception of facts is elided by moving to a post-foundationalist realism, which places them in webs of belief and thereby grants them the epistemic status of being conditional rather than foundational facts. Our apprehension of them is therefore fallible, and they are inherently enmeshed in the narratives in which they are employed, incapable of being ontologically severed into their separate domains as required by White and Ankersmit's inverse positivism.Footnote 58 ‘A fact is not given; it is a piece of evidence nearly everyone in a given community either accepts or perhaps has good warrant for accepting given the other intersubjective beliefs of that community’, Bevir observes.Footnote 59 The criteria of its acceptance as a fact are formed by ‘the other things we believe’ by virtue of our being in the world, such as the existence of objects and the distinction between past and present in relation to our conceptual categories.Footnote 60 The thin epistemic standards for accepting historical facts is not imperiled by the normative/aesthetic dimension of historiographical practices exactly because it remains attached to the thin realism emanating from our intersubjectively shared being in the world. A historical fact like ‘the Battle of Hastings was fought in 1066’ only makes sense in relation to prior categories, theories and languages elucidating the concepts of ‘battle’, ‘year’, ‘Hastings’ – all of which are ultimately grounded in the thin epistemic standards underlying our web of beliefs.Footnote 61
But how are historical facts established as such by virtue of intersubjectively shared epistemic standards, and what exactly do they consist of? Because our apprehension of them is fallible, they are amenable to change, with some having entered normal history by being included in a historiographical work, only to later fall out of favor and be excluded again.Footnote 62 This can happen for any number of reasons, such as lack of interest and perceived relevance, which are guided by second-order normative/aesthetic concerns; but it is its first-order epistemic validity that is primary in its consecration as historical fact. The finding of evidence per the standards of first-order realist beliefs questioning its validity dethrones its epistemic status as such, as famously demonstrated by one of the first of such acts in historiographical practice, Lorenzo Valla's demonstration in 1440 that the Donation of Constantine, which granted the Vatican certain territories, was a forgery, for it included terms not in use in Constantine's time.Footnote 63 This evidential aspect relying on our first-order beliefs for its validity is the only constant for historical facts to be sanctioned as such. As for what historical facts consist of, Paul Ricœur in my view correctly observes that they are reducible to testimonies, which form ‘the fundamental transition structure between memory and history’.Footnote 64
What exactly is the nature of such testimonies? Here we enter somewhat murky waters that I shall attempt to illuminate. Bevir talks of historical accounts being made out of ‘relics of the past’,Footnote 65 but fails to explicate what the nature of these relics are; whether they are limited to written documents, and of what kind (e.g. diaries, statistical reports, narratives, notes); whether it also includes objects such as archeological sites or buildings; and whether photographs, audio and video that come into being and proliferate the closer we get to the present time can be counted among them. Ricœur is more helpful in this regard. As noted, for him historical narratives are primarily constructed out of testimonies, by which he means witness accounts of any kind (e.g. diaries, notes, letters), but also other documentary evidence and clues drawn from a myriad of sources providing information about the past, both in and outside archives: ‘the debris coming from archeological excavations and other such vestiges, but in a more striking way kinds of information as diverse and mercurial as price curves, parish registers, wills, databases of statistics’.Footnote 66 Closer to the present, this also includes ‘fingerprints, photographic evidence, and today samples of DNA – that biological signature of the living – [which] “testify” through their muteness’.Footnote 67 However, Ricœur maintains the primacy of testimonies as written witness accounts in the construction of historical accounts,Footnote 68 and I follow him in this.
He does so because highly diffuse nature of documentary evidence – what can be termed indirect testimonies – serves to illuminate the past only in relation to our understanding of it derived from direct testimonies, the witness accounts that ‘fill in’ the space in which they obtain comprehensible meaning and take shape as a means of corroboration and verification. To clarify, take our knowledge of Ancient Greek poleis around 300 BCE derived from direct testimonies such as the writings of Aristotle and Plato. This allows historians and scholars in related fields to reconstruct the general contours of the modes of living in that time and place, such as dietary habits, marital relations, class divisions, legal forms, and so on.Footnote 69 Together with indirect testimonies of all kinds such as statistics regarding trade volumes, the amount of arable land and the population level that could sustain, the normal history concerning Ancient Greece around that time is established. It is due to the direct testimonies padded by the indirect that we can position within historical space a polis that is discovered in an archeological site somewhere in the Greek archipelago, and dated back to the period around 300 BCE by means of indirect testimonies such as radiocarbon testing applied to the excavated organic materials. It is exactly because such direct testimonies are lacking in non-literate societies that our knowledge of them is so rudimentary as to not contain any normal history of them, being limited to guesswork drawn from objects and materials found through archaeological fieldwork. As is noted below, the same lacuna affects our (lack of) knowledge of Ancient Greece and subaltern classes.Footnote 70 Ricœur is therefore right to place the primary locus of historical knowledge on direct testimonies, for ‘everything starts, not from the archives, but from testimony, and that, whatever may be our lack of confidence in principle in such testimony, we have nothing better than testimony, in the final analysis, to assure ourselves that something did happen in the past, which someone attests having witnessed in person, and that the principal, and at times our only, recourse, when we lack other types of documentation, remains the confrontation among testimonies’.Footnote 71 From now on any reference to testimonies includes both the direct and indirect variants of it unless specified otherwise.
Carlo Ginzburg offers a useful analogy of the historian as judge to clarify how the first-order realist dimension of normal history comprised of conditional historical facts is constructed out of testimonies by means of intersubjectively shared epistemic standards. Just as is the case with respect to witness testimonies in a courtroom, there are means by which the historian can reasonably distinguish between the epistemic validity of the claims it includes. Ginzburg's case-study to demonstrate this is the Holocaust,Footnote 72 which has been a common point of reference for those wishing to save a conception of realism given its normatively sensitive nature. It was also the subject of debate in the Historikerstreit, discussed in detail in the following section. Were there indeed gas chambers in concentration camps used to kill Jews, gypsies, communists and other ‘undesirables’ by the Nazis? For many, including Ricœur and Ginzburg, this question represents the absolute zero point in the debate between realism and anti-realism in the philosophy of history.Footnote 73 It also affected anti-realists, leading White to propose the inchoate concept of ‘modernist events’ that are incapable of being historicized due in part to our overly emotional involvement in them – with the Holocaust being the primary example of such an event.Footnote 74 And yet for all historians qua historians the answer to this question is a resounding yes. Raul Hilberg, Pierre Vidal-Naquet and Yehuda Bauer are just some of the many historians who have meticulously documented the evidence to make it so, amassing countless indirect and direct testimonies from a vast multiplicity of perspectives (both among victims and perpetrators, prisoners and executioners) and documents (e.g. architectural plans, deportation statistics).Footnote 75 The exclusion of Holocaust revisionists as engaging in pseudo-history is premised on their denial of the reasonable viability of the first-order claims contained in these testimonies, and even of their very existence. In his essay A Paper Eichmann: Anatomy of a Lie,Footnote 76 Vidal-Naquet shows how the revisionists proceed, and why they are to be excluded from the historian's field and thereby participating in establishing normal history:
The principles of revisionist method can in fact be summarized as follows: 1. Any direct testimony contributed by a Jew is either a lie or a fantasy. 2. Any testimony or document prior to the Liberation is a forgery or is not acknowledged or is treated as a ‘rumor.’ (…) 3. Any document, in general, with firsthand information concerning the methods of the Nazis is a forgery or has been tampered with. (…) 4. Any Nazi document bearing direct testimony is taken at face value if it is written in coded language, but unacknowledged (or under-interpreted) if it is written plainly (…) 5. Any Nazi testimony after the end of the war – in trials either in the East or in the West, in Warsaw or Cologne, Jerusalem or Nuremberg, in 1945 or 1963, is considered as having been obtained under torture or by intimidation (…) 6. A vast pseudo technical arsenal is mobilized to demonstrate the material impossibility of mass gassings.Footnote 77
Contrast this with the handling of the first-order aspects of the same class of testimonies relating to the Holocaust by the conservative and left-wing sides in the Historikerstreit concerning the legacy of Nazism. Both accepted their validity by virtue of their intersubjectively shared first-order realist beliefs and the epistemic standards emanating from them, which they, and by extension the historian's field, embodied. Our thin realist intuitions makes it irrational to maintain a dismissive attitude toward reasonably verifiable first-order truth-claims contained in direct testimonies relating for example the occurrence of events. Any notion of history, including our own past, is in large part based on the same class of testimonies. Maintaining a radical skepticism toward testimonies – particularly those that are direct – is therefore self-defeating, for we are committed to a view of the world that includes epistemic standards for designating their validity by virtue of our human practice and being in the world.Footnote 78 Bevir refers to this as a weak intentionalism, which builds on Davidson's charity principle in forwarding the claim that unless the historian has good reason to doubt the veracity of testimonies, they should assume good faith, which again is how we engage with the testimonies of others in our daily practice (and is institutionally consecrated in a courtroom setting).Footnote 79 The first-order truth-claims contained in testimonies then constitutes the intersubjectively shared, objective kernel of normal history, with the epistemic criteria for its consecration as such, emanating from our being in the world and inscribed into the nomos of the field, constituting the necessary but not sufficient condition which historiographical practices must meet in order to be considered as such.
This approach to delineating the epistemic nature of historiographical practices is close to Ernest Sosa's conception of virtue epistemology, which helps elucidate it in more philosophical terms. For Sosa beliefs constitute epistemic performances or practices that are correct by virtue of ‘a competence exercised in its appropriate conditions’.Footnote 80 This gives rise to the question of how beliefs derived from testimony can be considered correct or, as he terms it, epistemically safe, for unlike other forms of knowledge they are not reliant on the epistemic competence of the individual.Footnote 81 It is this gap that is deemed unbridgeable by anti-realists, who therefore give up on the idea of an epistemically relevant analysis of testimonies, especially when included in a narrative.Footnote 82 Sosa however offers a solution to this problem by locating the epistemic competence not solely in the individual, but also in the group, which, acting together in unison, ensures that the belief in question meets the intersubjective requirements for being considered epistemically safe: ‘The correctness of one's belief is still attributable in part to a competence seated in oneself individually, but the credit that one earns will then be partial at best. The quarterback's pass derives from his competence, but its great success, its being a touchdown pass manifests more fully the team's competence. Similarly for one's testimony-derived belief’.Footnote 83 The game metaphor is especially apt, for if the players are not adhering to the same basic rules they are, as noted, simply excluded, as in the case of Holocaust revisionists. Neither White nor Ankersmit can properly account for why narratives like the Holocaust revisionist's are not sanctioned as ‘historical’ on rationally compelling, intersubjectively shared epistemic grounds, rather than the normative/aesthetic.
The crucial epistemic dimension for theory choice then shifts to the treatment of testimony in them, and as I elaborate in the following section, it is there where shared epistemic dimensions in historiographical practices can be located. These accrue specifically from a distinctive set of historiographical practices that provide a superior handling of testimonies per the intersubjective standards of our shared first-order realist beliefs. The same first-order beliefs that afford the status of historical fact to information contained in testimonies therefore also yield criteria for theory choice on the level of historiographical practices and the narratives they produce based on the historian's handling of testimonies – specifically (1) The scope of testimonies, denoting the variety of perspectives included and sought out; (2) The depth of testimonies, denoting the degree to which they are analyzed on their own terms and the information they contain is taken seriously rather than dismissed. Per these criteria Holocaust revisionists fail miserably for being both scope and depth-constraining, whereas historians working in such areas as feminist history, subaltern and post-colonial studies typically do very well given the scope and depth-enhancing nature of their work, unearthing and bringing into normal history the perspectives of hitherto dismissed or simply unknown perspectives through their presence in testimonies. Unfortunately there are many lacunae in this regard given the dearth of such testimonies, especially the further back in history one goes, for the obvious reason that subaltern actors often did not have the resources or ability to create and leave behind testimonial accounts.
The thin realism inscribed in the historian field's nomos and homologously reflected in first-order historiographical practices offers the means by which additional epistemic virtues and dimensions inhering in these practices can be identified. It thereby reorients our view to the nature of historiographical practices – as engaged in by historians who thereby constitute the historian's field – in uncovering their epistemic value, rather than severing ourselves off from them and appealing to a substantialized set of general criteria as attempted by the overly thick realisms of Bevir and Lorenz. For White the question of scope or depth of testimonies in historical narratives is irrelevant for theory choice, as it remains reducible to the moral and aesthetic.Footnote 84 As noted, such a radical skepticism is self-undermining in daily practice, where we do rely on a reality principle inhering in testimonies for our existence. Memory, which defines our self-understanding and identity, is a product of testimonies; from our family, friends, and ourselves, and we do not take up a radically skeptic attitude toward it in our being in the world.Footnote 85 This is the thin realism that Bevir employs to counter the radical skepticism of White.Footnote 86 However, Ankersmit is right to point out the unviability of theory choice in accordance with thick criteria of realism as outlined by Lorenz and Bevir, given that these are crypto-normative,Footnote 87 brought to the forefront by the essentially contested history of the legacy of National Socialism as represented by the Historikerstreit. Lorenz and Bevir's post-foundational realism is not mortally wounded by this however, as long as the criteria they enumerate are weakened to become regulative epistemic virtues, indicating a greater likelihood of a greater degree of objectivity by virtue of the standards of first-order realist beliefs, those that are epistemically safe.Footnote 88 Thus reconceptualized, such criteria as comprehensiveness, accuracy and fruitfulness – functioning as regulative ideals – can be attached to the identified thin criteria emanating from our first-order realist beliefs with respect to the elements out of which historical narratives are made – testimonies. The extension of their scope and depth affords historians a way to epistemically discern between rival accounts on the basis of first-order realist criteria, thereby showing that these indeed serve as epistemic advantages in historiographical practices that pursue and adhere to them as regulative ideals and virtues.
3. Transcending the realism/anti-realism divide: The case study of the Historikerstreit
This is exemplified clearly by the Historikerstreit, a historiographical struggle that arose in the late 1980s over the legacy of Nazi Germany. Two general sides in the struggle can be discerned, one being on the political left, among whom Jürgen Habermas was most prominent; and the other on the right, represented most influentially by Ernst Nolte and Andreas Hillgruber. As Lorenz notes, most commentators consider the Historikerstreit to have been won by the left-wing, Habermasian side.Footnote 89 But why was this the case? Was it merely, as anti-realists like White and Ankersmit claim, a matter of normative/aesthetic preference, or does the alternative thin-realist account I have sketched provide a better answer? The first-order strategies of the Nolte and Hillgruber sides were explicitly scope and depth-constraining rather than enhancing in nature, as Lorenz outlines:
Hillgruber thereby flatly denies that his choice in favor of the perspective of the ‘Wehrmacht’ conceals a normative choice. He presents this choice as one dictated by historical reality itself. The historian of the Eastern front, according to him, is confronted with the following alternatives: to choose to write history from the perspective of Hitler, or from the perspective of the Russians, or from the perspective of the inmates of the concentration camps, or from the perspective of the German civilian population and the German army protecting it. The first three perspectives do not match reality, according to Hillgruber, because the German population did not identify itself with one of these parties. Therefore the perspective of the German army remains as the only ‘realistic’ point of view for the historian. (…) His formulation of the factual historical problem evidently hinges on the separation in his frame of description between 1) Hitler on the one side and the German army and civilian population on the other, and 2) the German army and population on the one side and the inmates of the concentration camps on the other. Apparently the latter – mainly Jews, gypsies, Communists, and socialists – are not ‘real’ Germans in Hillgruber's view, because they are not identified as such by either the majority of the German population then – this is an undisputed historical fact – or the (present-day) German historian in the 1980s – this is his normative choice.Footnote 90
Lorenz is right to refer to Hillgruber's choice as normative, but does not identify the crucial epistemic dimensions of this choice, i.e., of testimonial scope and depth-constraining historiographical practices. By pursuing first-order realist strategies that indicate, by the intersubjectively shared standards of our being in the world, a lower degree of objectivity by excluding crucial testimonies, in particular of those historical actors that had the status of subaltern, the scope criterion is transgressed. At the same time the depth criterion is selectively applied to those historical actors least likely to produce objective accounts of their crimes, namely those who perpetrated them. The same holds true for Nolte, who emphasized the genuine ‘fear of Bolshevism’ among many Germans in his historical narratives so as to relativize the uniqueness of the crimes the Nazis perpetrated. He thereby sought to alleviate them and Germans more generally from a great burden of responsibility, which was instead shifted to the objects of their fear, who also happened to be their victims: Bolsheviks and Jews.Footnote 91 The historiographical practices pursued by those opposing these narratives rightly pointed to such epistemic limitations, as well as the second-order normative beliefs that motivated their construction as such. They ultimately managed to successfully defend a view of normal history which rejects such testimonial scope and depth-constraining accounts.Footnote 92
This success was not guaranteed however, and cannot be reduced solely to the ideological nature of these accounts, thereby conflating the sociological and philosophical aspects of theory choice. Nolte and Hillgruber accept the (implicit and explicit) rules of the historiographical game, and deny the role played by second-order normative beliefs in guiding their practices, which as Lorenz observes they tend to represent as being purely factual in a foundationalist sense,Footnote 93 appealing exclusively to the first-order realist beliefs that underlie the historian's field. Rather, it is the epistemic defects of these accounts and the advantages of those of the left-wing, coming to the fore in their handling of testimonies – scope and depth-constraining and enhancing, respectively – that sealed the deal for the general community of historians.Footnote 94 This shows that these do in fact indicate epistemic advantages and quite important ones at that – if the case of Holocaust revisionism had not already indicated this. What Lorenz and Bevir fail to observe is exactly this. These epistemic aspects inhere in practices, and come to the fore in relation to the first-order realist beliefs inscribed in the historian field's nomos, homologously reflected in the first-order strategies pursued by historians. The realist standards by which historians judge accounts and engage in theory choice are therefore intersubjectively shared, and indicate the presence of epistemic virtues such as accuracy and fruitfulness by the standards they explicate.
Now that the rational kernel of historiographical practices and how it comes to constitute normal history is epistemically positioned, it is possible to analyze on post-foundational realist grounds, by means of a research programme of the kind pursued by empirically minded historical theorists like Herman Paul,Footnote 95 the development of these practices over time with respect to instantiations of normal history. Such a hindsight look at the success or failure of historical accounts allows for the epistemic dimension of theory choice to be maintained, while the moral and aesthetic are mapped alongside it. White and Ankersmit are perfectly right when they say that the criteria for theory choice lack a rationally compelling force, often being trumped by moral and aesthetic concerns in the construction of normal history.Footnote 96 It would be self-defeating to deny this, as Lorenz and Bevir tend to do, by referring to instances of (partial) success such as the Holocaust, when there are so many others of blatant failure such as the French Revolution.Footnote 97 However, such epistemic criteria as the significance of testimonial scope and depth, rooted in the intersubjectively shared being in the world of historians, do function as regulative ideals in historiographical practices, and exclude or disadvantage those who constrain rather than enhance them. Theory choice as it has been and is actually engaged in by the community of historians in the construction of normal history is thereby explained on epistemic rather than normative/aesthetic grounds, which resolves the problems that the anti-realist and thick realist positions run into in cases like the Holocaust and the Historikerstreit. Whereas White and Ankersmit's positions can be characterized as being overly pessimistic and claiming too little with respect to realism, Bevir and Lorenz's suffer from the obverse and is overly optimistic, claiming too much. The alternative thin-realist philosophy of history I have explicated thus navigates a middle path between the Scylla of objectivism and the Charybdis of relativism.
