Values and Historiographic Knowledge:

The Historian as A Judge

Aviezer Tucker (Trinity College)

Around the turn of the nineteenth century, historians like F. A. Wolf and Niebuhr, and following them Ranke adopted new communal methodologies that are often called "scientific" or "critical" historiography. By "communal methods" I mean:

With the aid of these critical methodologies and the theories upon which they are based, historians have been producing historiographic knowledge. This knowledge, such as it is, is the subject matter of the philosophy of historiography. The examination of the epistemology of historiography entails, as I show next, the basic metaphysical problems of historiography and a plausible empirical research program that is relevant for the examination of these epistemic and metaphysic questions about historiography.

Consensus on beliefs in historiography is founded on consensus on cognitive values. The community that agrees on cognitive values is usually larger and lasts longer than the community that reaches consensus on beliefs. As we see next, this is also the case with the cognitive values of historiography. The identification of the cognitive values of historiography is initially a historical inquiry, a study of the historical development of the cognitive values of historians, as they emerged at the turn of the nineteenth century, spread, became dominant, and provided a foundation for communal consensus that had been absent previously in historiography. These new cognitive values are not exclusively historiographic, nor did they emerge first in historiography, nor ended their expansion through inter-disciplinary migration with historiography.

The only cognitive values to ever win the support of a heterogeneous and uncoerced historiographic community proved themselves to be conducive to the acquisition of knowledge on such an extensive scale, that virtually all modern heterogenous uncoerced societies accepted the cognitive values of critical historiography, if not in all the realms where they can be applied, at least in most of them, discounting areas that are "compartmentalized," artificially isolated from the spread of the cognitive values of other disciplines. The cognitive values that historians rely on are shared also by detectives, judges, philologists, textual critics in biblical and classical studies, and evolutionary biologists.

Historians that are called sometimes "scientific" or "critical" share a set of cognitive values that allow them to reach agreement not just on the interpretation of evidence, but also on deciding what to admit as evidence in the first place. Shared cognitive values both help to define the community of historians, and are assumed in the current inquiry as we attempt to discover the historical emergence of the community of historians who share these cognitive values. Thus, an aspect of the philosophy of historiography is a part of historiography itself: The philosophy of historiography must rely on the cognitive values and methods of historiography to discover the conditions of historiographic knowledge, the self-same cognitive values and methods that it assumes. But the philosophy of historiography does not beg the question because the cognitive values it shares with historiography are also shared by other disciplines. The best explanation for the interdisciplinary consensus on these cognitive values is their conduciveness to knowledge.

The community of historians is distinguishable by its subject matter, namely the study of the human past, and by its cognitive values. It is possible to discover the cognitive values of critical historiography by making two comparisons: First, between the cognitive values of the uncoerced, large and heterogeneous community of historians that emerged in the first half of the nineteenth century, and the cognitive values underlaid by the historiographies that were practiced earlier. Second, it is possible to compare the cognitive values of the established, post-Ranke, community of historians with those of people who write about the human past, but are not recognized as members of this otherwise heterogenous community or are even ostracized, following judgments of cognitive values. For example, historians routinely reject historiographies written by Holocaust deniers, the memoirs that politicians write in mid-career, and conspiracy theories. Historians do not reject them as bad historiographies, as they would if they criticize the work of a colleague in a book review, but as no historiographies at all that are ignored. The difference in cognitive values underlies this social distinction.

Before the rise of critical historiography, traditionalist historiography accepted uncritically the authority of ancient texts, in the European context, of the Christian scriptures, classical Greek and Roman historiography, and various chronologies and historiographies written during what came to be known later as the Middle Ages. Traditionalist historiography further accepted traditional oral interpretation of these authorities. The Latin tradition is derived etymologically from the verb to hand over, hand down or transfer. Similarly, the hebrew masoret is derived from the verb limsor, to hand over or transfer. Respect of tradition as a cognitive value assumes that what was handed over in tradition was precisely preserved, authentic, through continuous transmissions, as a coin maintains its shape and consistency through numerous transactions. The authority of tradition can be sustained only as long as faith in an unbroken, uncorrupted and precise chain that connects a present text, or tale, or piece of information, or institution to a legitimate origin is preserved. If this faith is lost, so will the authority of tradition and the beliefs and institutions it legitimizes. In other words, traditional historiography ends with the development of a suspicious approach to the evidence of other people, a suspicious approach that probably emerged first in courts of law facing the ubiquitous presence of false witnesses. The rudimentary solution for this problem, the presentation of at least two independent cross referenced witnesses, also emerged first in the courts. Since much of the traditional, ancien regime, social order in Europe was based on uncritical reverence of the historiographic authorities, the political and ecclesiastic powers coerced (by punishment) and cajoled (by employing court or church historians) consensus on historiographic issues as journalists have been treated by more recent authoritarian societies. Before the present became the dominant time-frame of humanity and with it journalism became the relevant profession for influencing popular consciousness, the dominant temporal reference for locating the place of the person in the universe was the past. Consequently, historians were important in influencing popular perceptions of the past and their consequent behavior in the present.

Critical biblical scholars, classicists and historians were not the first to note discrepancies between texts. For example, Jewish medieval biblical exegesis consisted to a large extent of explanations of discrepancies in the bible by invoking "hidden meanings." Real inconsistencies were considered impossible. Apparent inconsistencies were considered an economic method for the texts to add a hidden, at times mystical, meaning beyond the literal meaning of the text. Similarly, later, historians attempted in the seventeenth century to interpret the scriptures so that they are consistent with the geographical, anthropological and historical discoveries of their era. For example, much work was devoted to presenting the chronologies of Chinese history as consistent with the biblical chronology of world history. All this work on presenting diverse texts as consistent is based on the cognitive value of faith and trust in the authority of tradition. No text was subjected to criticism or to doubt; such a possibility was not even rejected, it was not considered in the first place. Therefore, any apparent inconsistency implies that the readers misunderstand something that they need to discover.

Traditionalist cognitive values were dealt their first blow in Europe with the Reformation. Martin Luther attacked the traditional, homogenous by coercion, interpretation of the scriptures. Though Luther expected the text to speak for itself, as it were, to have a literal meaning, it soon became apparent that the text spoke in different voices to different readers. In other words, once the coercive element in the Christian tradition was weakened, the tradition began to fragment, as all traditions do naturally. Once there emerged several competing traditions, all claiming to be authentic, the authority of any tradition, of traditionalist cognitive values themselves, began to weaken.

The scientific revolution of the seventeenth century dealt a second blow to the authority of ancient texts. If Aristotle could be shown to have been mistaken and be corrected by Galileo and Newton, what makes Thucydides and Tacitus more reliable than Aristotle?! Even more so, the biblical account of history could come under scrutiny. The geographical and anthropological discoveries of the seventeenth century further questioned the scope and truth of the ancient texts.

Facing this assault on traditionalist cognitive values, the faithful could only recommend sacrificium intellectus, not a very convincing solution once tradition has already been subjected to widespread doubts. Through the seventeenth and eighteenth centuries, traditionalist cognitive values continuously lost ground and were replaced by skepticism about history. Instead of valuing all ancient texts and traditions as true, they were considered highly unreliable. The rationalist, most notably Spinoza's, examination of the bible presented it as a historical document of a fairly low probability. Others suggested that it is impossible to know anything about all but most recent history. For example, the Jesuit Hardouin maintained that the whole of ancient historiography was fabricated at the 13th century by monks. Enlightenment historiography apart of being ahistorical, was as incredulously skeptic as its competitors and predecessors were believers. The Reformation and the Enlightenment did much to doubt, unhinge, and shatter the certainty of traditionalist cognitive values. But rationalism and the enlightenment, philosophers like Descartes and Bayle, could offer only skepticism of any knowledge of the past in its stead.

Critical historiography emerged out of this crisis of historiographic cognitive values to overcome the dismal choice between blind faith in tradition and skepticism. Critical historiography accepted from the enlightenment its cognitive values of suspicious, distrustful approach to tradition and its naturalism, rejecting any supernatural explanations or descriptions of historical events. Critical historiography added to these enlightenment cognitive values respect for primary sources, cross referenced, as reliable evidence that can generate knowledge of history. These values were used, if not theorized upon, for centuries in the courts of law. The first attempt to apply them for knowledge of the non-immediate past used the foremost traditional texts of the Western Civilization, the judeo-christian scriptures, as evidence for knowledge rather than as knowledge simpliciter. Once successfully established there, the cognitive values and theories they generated migrated to the study of the Greek and Roman classics and from there to philology, historiography and eventually evolutionary biology. The conduciveness of these cognitive values to the generation of knowledge across the disciplines explains their amazing interdisciplinary expansion and the generation of consensus of cognitive values in heterogenous, no longer coerced, intellectual communities, and eventually to the emergence of new consensus of beliefs about the past. Once scholars were able to adopt critical cognitive values to the holy scriptures and survive it politically, exporting their critical values to the less politically charged field of historiography was smooth and quite risk-free, at least in those political units that have already tolerated biblical criticism.

The significance of objective interpersonally recognized evidence in the cognitive values historiography imported from biblical criticism, classical philology and the natural sciences, led to valuing the achievement of inter-subjective consensus, as a proof of the "objectivity" of historiographic knowledge. "Objectivity" was adopted as a worthy goal, though its meaning has been disputed ever since. I shall not discuss in this book the meaning of historiographic objectivity because the term has been overused and unclearly defined in its historiographic context. Instead, I merely concentrate on intersubjective agreement in a heterogeneous, uncoerced and large community.

The overwhelming significance of Leopold Ranke is in establishing a universal community of historians that shares cognitive values. Many commentators noted that Ranke did not invent the critical cognitive values, theories and methods that won him acclaim as the founder of scientific historiography, though he was as good and brilliant a practitioner of them as any historian that came before him. Historiographic criticism, criticism of received tradition and forged charters and naturalism against religious miracles were present already during the renaissance. "Critical historical research... preceded Ranke. Already in De re diplomatica (1681), the French historian Jean Mabillon had developed rules and criteria for judging sources and determining their authenticity. There had developed in the eighteenth century at the university of Göttingen a school of historians, including Johann Christoph Gatterer, August Ludwig Schlözer, Arnold Hermann Heeren, and Johann von Müller, who combined the critical methods of érudits like Mabillon with the concern of the philosophic historians of the eighteenth century..." (Iggers & Moltke, 1973, xxvi-xxvii). Yet, all these attempts and precursors were insufficient for the creation of a consensus on cognitive values in a heterogenous uncoerced and large historiographic community. The founder of this community was Ranke. By the time Ranke came on the scene, biblical criticism and historical philology had already been established and their cognitive values were already accepted and even entrenched. In this sense, Ranke sawed a land that had already been plowed by biblical criticism. Ranke's position in the Prussian state and at the University of Berlin allowed him to found the community by training the next generation of historians. "Most of the great German historians of the nineteenth century, among them Heinrich von Sybel, Wilhelm von Giesebrecht, Georg Waitz, and the Swiss Jacob Burckhardt, passed through Ranke's seminar in which he applied the critical method to historical writings. These students in turn trained a generation of scholars who dominated German university chairs of history well into the twentieth century. Ranke thus became in fact the father of modern historical science in Germany." (Iggers & Moltke 1973, xxvi). Ranke's significance was in importing the cognitive values, theories and methods that he had studied in academic departments of theology and philology to historiography. He then used his hierarchic position in the Prussian academic bureaucracy and Royal patronage to establish these values, theories and methods as dominant within the historical profession, in defining the historiographic community. Ranke is not the first person to be considered an innovator or inventor because of bureaucratic position rather than novelty of thought. For example, traditional Chinese historiography considers the bureaucrat who supervised the first state printing enterprise as the inventor of printing, though he did not invent the technology. Al Gore's claim in the 2000 US presidential elections to be the inventor of the internet may be in this tradition....

Ranke is the founding legislator of the historiographic community, the one who instituted the cognitive values that define the community. Ranke's cozy relations with the Prussian royal family are criticized often as indicators of a bias in favor of reactionary politics in his writings. Actually, Ranke's royal patronage protected the establishment and institutionalization of the critical cognitive values of historiography against their two main competitors, lingering traditionalism and the new popular nationalist-mythical approach to history that considered historical knowledge to be what united the nation, without distinction of class. Royal absolutist hostility to reactionary religion and popular nationalism facilitated the institutionalization of the cognitive values of modern historiography.

Ranke was of course a product of his time and place. As many historians noted (Krieger 1977, Iggers & Moltke, 1973), Ranke's books reflect the influence of some of the values of his environment, such as Ranke's protestant religious world view, his adherence to older models of universal historiography, his moderate monarchism, etc. It is impossible for any historian to avoid making such value judgements. But, had Ranke's historiography been marked only by these non-cognitive values, it would have been forgotten. The part of Ranke's historiography that became universal and influenced countless formal and informal pupils is the one where he established and followed universal critical cognitive values and used critical theories and methods to gain knowledge of the past. The peculiar values of a mid-19th century protestant loyal subject of imperial Prussia had no lasting effects on members of the heterogeneous historiographic community. They accepted Ranke's universal cognitive values and rejected all the irrelevant but inevitable value judgements that historians always make along the way, and often left lying there on the side of the road for future generations of readers.

Since Ranke, critical historians have been distinguishing legitimate historiography from illegitimate or fictive "historiography." Illegitimate historiography includes for example, "revisionist" historiography of the Holocaust; nineteenth century nationalist historiographies that "discovered" ancient national sagas; the changing Bolshevik historiographies of their revolution, and some of the conspiracy theories of the murder of J. F. Kennedy, Rabin etc.

Historians distinguish legitimate historiography from the illegitimate variety according to the cognitive values of the historiographic community. Illegitimate historiography usually uses one of many therapeutic cognitive values to choose "correct" historiographic statements. Therapeutic cognitive values value historiographies according to their effect on the psychological well-being of its intended audience. Some of the main therapeutic cognitive values in historiography include: The denial of all historical guilt (eg. Holocaust denying); denial of inferiority and promotion of self-aggrandizement, (eg. national myths), and the elimination of a sense of alienation and absurdity from one's relations with history, through eg. conspiracy theories. Therapeutic historiography does not care for the cognitive values of critical historiography because conduciveness to knowledge is unimportant from a therapeutic perspective. Some knowledge of history may have therapeutic value, for example a member of a racially discriminated community may feel better and become more confident if she learns of those achievements of her forefathers that clearly refute dominant racial stereotypes. But the critical cognitive values of historiography are not particularly conducive to therapeutic effects. They may just as well generate the opposite. For example, a member of a downtrodden community may wish to believe that those better off designed a global conspiracy against his people. Critical historiography may discover that nobody had any plan against his people or even noticed them. Their misery had no reason or sense and nobody benefitted from it.

The distinction between critical and therapeutic historiography is indicated by the difference in the communities that accept them. Historiography founded on critical cognitive values is accepted by a large heterogenous and uncoerced community. Historiography founded on therapeutic values is accepted by a particular homogenous community that is clearly identifiable: Holocaust denying is popular among Neo-Nazis, particular national historical myths are promoted by particular nationalists and denied by their rivals of different nationalities, faith in a "Jewish conspiracy" is promoted by losers in the transition to modernity and by political enemies of the state of Israel, etc. The clear exclusive identification of communities united by historical pain with therapeutic historiography is a clear indication that they do not reflect knowledge of history. Yet, there will always be a market for therapeutic historiography because people and their institutions will always pay to promote, read or hear therapeutic accounts of their past. It simply make people feel better.

The inconsistency between therapeutic and critical cognitive values manifests itself in historical conflicts between critical and therapeutic historiographies. For example, during the nineteenth century various forged "ancient" documents surfaced but were exposed despite their therapeutic value for nationalist causes. For example, the poems of the "Scottish Homer," Osian, who counted Goethe as one of his admirers was exposed in the early nineteenth century. In the Czech lands, Tomas G. Masaryk participated in exposing similar "ancient" Czech poems, during the Czech national struggle for self-determination against the Austro-Hungarian Empire. The universal value of the cognitive values of historiography is further strengthened by the fact that Masaryk was also the foremost leader of the Czech national movement and later became the first president of Czechoslovakia.

Legitimate historians, like Masaryk, accept a hierarchy of values, according to which their critical cognitive historiographic values take precedence over the therapeutic cognitive values and needs of their political/ national/ class/ ethnic/ gender/ racial/ religious/ and/ or/ other/ group(s). We may want to believe that a group with which we identify has always been virtuous and faultless; and that whatever blemishes we find in our group are the product of the evil that was done to us unjustly by some other political/ national/ class/ ethnic/ gender/ racial/ religious/ and/ or/ other/ group(s). But if this involves overriding the critical cognitive values of the historiographic community, this is exactly what the legitimate historiographic community should not let us believe in.

Legitimate historiography is marked by the precedence of critical cognitive values over other values, not by the absence of other values. Works of legitimate historiography are affected by the non-cognitive values, affiliations and political biases of its authors, this is the reason for the differences between historiographic works that cover similar subjects. Yet, their agreements are generated by the identical cognitive values that they all share. One these are satisfied, there is plenty of space for personal expression of value judgements. For example, William Dray showed how the sympathies of historians of the American Civil War affected their dating of the beginning of the causal chain that led to the war; this does not violate historiographic critical cognitive values.

The historical formation of modern historiography as a respectable, independent and scientific discipline has been fraught with struggles with ideological or therapeutic fabrications of history, from the struggles against the pseudo-medieval epics in the nineteenth century and the struggle against whig historiography that fabricated an ancient English constitution of liberty, through the struggle with Bolshevik historiographic fabrications, to the contemporary struggle against revisionist historiography of the Holocaust. These struggles helped to define the historiographic community and distinguish it from authors whose writings refer to the past but deny the tenets of critical historiography.

The difference between therapeutic legitimate and illegitimate historiography can be illustrated by comparing Holocaust denying revisionist fabrications of the past, and its "contextualization" by Ernst Nolte. The therapeutic-political purposes of both are similar: to eliminate Nazi or German guilt for the crimes committed by Nazis, Germans and their allies during the Second World War, and to dissociate Nazism, or German nationalism, or radical nationalism from crimes against humanity and facilitate resurgence of Nazism, or radical nationalism, or German nationalism. The revisionists simply deny history and then fantasize an historiographic account of the second World War that does not include crimes against humanity. Nolte, by contrast, does not deny relevant evidence and its accepted interpretation concerning the events of the Holocaust. His theoretical and explanatory interpretation of the holocaust varies from that of other historians, but not in such a way that violates the basic values of the historiographic community. Nolte constructs a comparative theoretical model that subsumes Nazism under a more general model of twentieth century totalitarianism that connects it with other atrocious regimes to the therapeutic affect that "we" (Nazis, Germans etc.) were not the only ones to commit horrendous crimes, so relative to other regimes (the Soviets) "we" were not particularly guilty and evil, so two wrongs make a right. Then, Nolte looks for those aspects of the Third Reich that were not evil, a matter of emphasis (the Nazi regime also supervised the construction of autobahns). Finally he suggests a theory that holds that one of the causes for Nazism was Bolshevism, ie "we" are not to be held fully responsible for the evil that "we" committed, because it was determined to an extent by the evils of others.

Nolte's arguments resemble those of a defence attorney who would argue for leniency for a client after the court comes with a verdict of guilty as charged! He claims that the defendant is from a bad neighborhood where everybody became a criminal, that he also did good deeds for the community, and that he was provoked by the crimes of others. There is nothing in Nolte's claims to make them inferior epistemically or methodologically to mainstream historiography, irrespective of what we may feel about the pragmatic uses he intends for them. The arguments of Nolte's opponents, such as Christian Meier and Jürgen Kocka for the uniqueness of the holocaust and the peculiarities of National Socialism resemble those of the prosecution that argues for the incomparable severity of the crime, for the absence of extenuating circumstances, and the unprecedented suffering of the victim. Yet, what is interesting in the Historikerstreit is not the predictable disagreements on the interpretation of recent history that have obvious implications on contemporary political discourse, but rather that both sides were able to agree on so much and remain within the bounds of a united historiographic community.

A simple thought experiment that may clarify the distinction between disagreements within the historiographic community of cognitive values and between this community and those who eschew its values is whether one side could win a libel suit against the other in a democratic court. Legal and historiographic reasoning are quite similar, as the cognitive values of historiography emerged originally in the courts of law. What would stand in court is likely to be accepted by the historiographic community and vice versa. The criteria of a libel case are significant because reasonable doubt is not relevant as it is in a criminal prosecution. The reference to the UK is relevant because in the USA it is necessary to prove both that the allegation was wrong and that the libelous party knew it was wrong before libelling. The consciousness clause is not required for example in the UK.

Skeptical philosophy of historiography denies the exclusivity of the critical cognitive values of the only existing heterogenous, uncoerced and large historiographic community and their primacy in relation to non-cognitive values. Instead, they would endorse value pluralism, where different and inconsistent historiographies be allowed to follow different cognitive values that would not necessarily take precedent over non-cognitive values. If skeptics interpret historiography as composed largely of narratives, they would claim one narratives is as good as another, and therefore there is no substantial difference between historiography and fiction.

I am not interested in examining the normative aspect of the skeptical position, the prescriptive recommendation that historiography should be written without preference to any particular set of values, cognitive or other. As a descriptive scheme, skepticism just does not fit the history and sociology of historiography. The existence of an uncoerced heterogenous community of historians must be a complete mystery for the skeptics.

For Hayden White "historical narratives... are verbal fictions, the contents of which are as much invented as found and the forms of which have more in connection with their counterparts in literature than they have with those in the sciences." (White 1978, 82). Had White been right, the sociological structure of historiography makes no sense at all. Had White been right, historiography should have been as fragmented as literature or art. White concludes that the choice between competing historiographic narratives should be undertaken on aesthetic grounds. The conclusion of historiographic aestheticism, as its adherents admit, is the denial of the possibility of proving that the Holocaust took place (White) or that the victims of Stalin's trials were innocent (Spitzer). Yet, clearly, historians do think they can prove that there was a Holocaust and that Stalin and his minions were suffering from paranoid delusions. The skeptic would find this consensus puzzling. Normatively, I guess that even the most radical skeptic would prefer to recognize the primacy of the cognitive values of critical historiography if he is ever brought as an accused before a court of law for a crime he did not commit.

The skeptical interpretation of historiography is founded to a large extent on pre-critical historiography and substantial philosophy of history (white). Nancy Partner (1995) drew an interesting correlation between the mixing of fictional and evidence-based accounts in classical and medieval historiography and contemporary American "TV docudramas." But she brought no examples from contemporary critical historiography to prove her claims about the limited extent to which historiography is founded on critical cognitive values. True, critical cognitive values and the ensuing critical historiography did not dominate historiography over night. They emerged first in Germany at the beginning of the nineteenth century and travelled westward during that century. Literary, non-critical, historiography, for example of Macaulay and Michelet had a wide reading public and consequently a market niche during the nineteenth century even if it did not adhere to the new Rankean standards. Substantive philosophy of history has never adhered to any critical standards either. But neither represent historiography as a communal enterprise for the generation of knowledge of history. This evidential base is useless for drawing any conclusions about critical historiography.

Once historians adopted the epistemic value of objectivity, the achievement of inter-subjective consensus became a significant goal, as a proof of the objectivity of historiographic knowledge. Therefore, the establishment of historiographic consensus, like scientific consensus, is a sure sign of the presence of historiographic knowledge, whereas the discovery of disagreement and even absence of communication indicate underdetermination. I reject the mystification of assigning agreement among historians to an alleged "craft," "artistic sense," "brew master's nose" or acquired through experience and age (British style) "commonsense." All these terms are equally vacuous in this context. Historians decide in communities of epistemic values and common methodologies. Academic A meaningful research program in the philosophy of historiography should examine then how and why do historians reach consensus on some issues, but are unable to agree or even communicate on others. The empirical aspect of this research would be a historical and sociological examination of unity and diversity in the historiographic community that shares the above epistemic values, into schools and sub-schools (or factions in alternative terminologies). The sociological structure of historiography reflects theoretical and methodological agreements and disagreements. Historiographic consensus must be based on a methodological core, founded on confirmed theories, with accepted evidence and acknowledged standards of theory choice that create historiographic scientific knowledge. The interpretive, critical, textual-archival methods that Ranke assembled, institutionalized, practiced successfully, and practically taught are the theoretical core that generates consensus. The significance of Ranke is not in what he thought of his enterprise, the misleading quote on the task of historiography "only to show what actually happened [wie es eigentlich gewesen]." or "It merely wants to show how, essentially, things happened." that everyone mentions in relation to Ranke, but in what he actually achieved, the establishment of critical historiography and its core methodologies. Some of these core consensus-generating theories may be shared with other disciplines that confirmed them independently, eg the rules of evidence in jurisprudence and most significantly, biblical criticism and historical philology. Though Ranke was probably the first to assemble a group of core theories and methods that defined critical historiography, these theories have been constantly refined and supplanted or even surpassed during the subsequent historical development of historiography. For example, textual criticism can be conducted today by applying algorithms in computer programs.

Ranke's methods were particularly appropriate for dealing with textual evidence of political history. Consequently Ranke's successors criticized him for being in the position of a person who looks for lost keys under a street lamp because that is where the light is. Actually, this kind of behavior is rational and prevalent in science. To look for their keys elsewhere, historians had to find other sources of light, other methods, and that is precisely what they did, while maintaining the defining communal epistemic values that Ranke introduced. Historiography became specialized and each of the sub-fields such as economic historiography or cultural historiography developed its own methodologies and theories.

Philosophers who were innocent of historiography looked in vain for the elusive theoretical background of historiography in the social science and commonsense generalizations. That theoretical background lay all the time "under their noses," but they looked in all the wrong places, probably because they assumed that theoretical historiography must be something like Marxism or Newtonian physics. Some philosophers of historiography found it difficult to identify local historiographic methodologies that are routine and habitual and are neither discussed nor articulated, and yet are passed on from one generation of historians to the other. At least one philosopher was able actually to invent a formal language to describe the critical method of philologists and historians, but W. W. Greg's 1927 book seems to have been ignored.

Contemporary epistemology arguably began with Quine's naturalized epistemology. Quine suggested that how we actually arrive at our beliefs is relevant for questions of how we should arrive at them, descriptive questions about belief acquisition are relevant for normative ones. Description is better than rational reconstruction because it is better to discover how science in fact developed than to "fabricate a fictitious structure." Epistemology should study the relation between input and output, evidence and theory. A Quine inspired research program in the philosophy historiography would examine, then, the relations between historical evidence and historiography.

An analogous and quite successful research program is present in the philosophy of Law. Philosophers of law examine the relations between the class of legal reasons (laws, cases, regulations, statutes, constitutional provisions, etc.), the facts about events and actions that are presented in court, and the outcomes of the legal process, verdicts (that include the judge's decision, order, and opinion). A related question is why do judges agree on some verdicts (easy cases) and tend to differ on other cases (hard cases)? There are three main approaches to this question in the philosophy of law: Determinism, Underdeterminism, and Indeterminism.

Legal Determinism holds that "the law rules"; judges deduce verdicts uniformly and regularly from the class of legal reasons and the facts they are presented with in trials. Legal Indeterminism is the opposite position, the law and the facts of the case carry no influence on the decisions of judges. Judicial decisions can still be regular and even predictable, but the regularity does not result from the law, but from external factors such as political pressures ("all law is politics"), class interests, etc. This position is associated with the school of Critical Legal Studies.

Legal Underdeterminism is a moderate position between determinism and indeterminism. It holds that taken as an aggregate, the decisions of judges are constrained but not determined by laws and facts. Judges choose among options within limits. Legal underdeterminism is associated with the school of Legal Realism that attempts to form a descriptive theory of the relations between input and output of adjudication.

Historians are compared more often to detectives than to judges. But historians do much more than detective work: they must be able to convince through arguments other professionals and the general public of their conclusions and they must consider the arguments of other "litigants" as it were. There is an interpretive element in historiography as in the judicial process, not of normative laws but of theoretical approaches associated with schools. A historiographic conclusion may be "appealed" and historians like judges do not limit themselves to judgements of facts, but make judgements of value and present opinions that are not directly relevant to the verdict. The similarity breaks on the legal criterion of "reasonable doubt" that is far less strict in historiography because historians do not usually determine the fate of their subjects.

It is interesting to note the high correlation in history between constitutional systems that implement the separation of powers and protect the relative independence of the judiciary, and the presence of critical historiography. Regimes that fear an independent judiciary attempt to control judgements about history as well. Further, there is an internal connection between historiography and the judicial process, because they are founded on common values and methodologies that can either appear simultaneously in the legal profession and among historians in a certain socio-historical milieu or can be copied from one community to another.

Another advantage for modeling the philosophy of historiography after the philosophy of law is rendering the narrative debate superfluous. The important issue is not whether historiography is a narrative (it is obvious to me that some parts of it are and other are not), but what is the relation between historiography, narrativist and non-narrativist, and its evidence. A significant research program in the philosophy of historiography should examine the relations between historical input (evidence, chiefly primary sources) and historiographic output, written accounts of the past. Analogously, there can be three approaches to the philosophy of historiography:

Determinist philosophy of historiography would claim that historians deduce from the class of historical reasons that includes evidence, and historiographic theories and methods a single historiographic "output." Historiographic determinists could recognize that different sub-fields of historiography have different "inputs," different types of theories, methods and evidence, because what is considered evidence is partly determined by the theories themselves. But they would claim that all historiographic outputs are consistent, creating together a jigsaw puzzle picture of the past. This position may be associated with the naive anti-philosophical positions that historians who defend historiographic lore espouse.

Historiographic indeterminists would claim that historical epistemic values, evidence and historiographic critical-scientific methodologies do not determine the content of historiography. Instead, whatever consistency and regularity we find in historiographic judgements result from political, ideological or sociological-historical values that groups of historians share. This approach can be related to the old sociology of knowledge as well as to relativism and deconstruction.

Historiographic underdeterminism would claim that some historiographic accounts are determined by the evidence and historiographic methodologies, while others are underdetermined, ie historians are constrained by their epistemic values, they are forced to choose among a finite range of possible interpretations.

This distinction between determined and underdetermined parts of historiography should not be confused with a distinction between chronicle facts and explanations. Some chronicle facts are underdetermined, for example, whether Czechoslovakia's Foreign Minister Jan Masaryk was assassinated or committed suicide in 1948, or how popular was the Communist takeover of Czechoslovakia in February of that year; while some explanations seem determined, for example, that the global economic recession that began in 1929 was a contributing cause to the rise of Nazism, or that the British and French position immediately before and during the 1938 Munich conference contributed to a rise in the popularity of the Czechoslovak Communist Party. It is a common philosophic mistake to assume that historians agree only on chronicle facts, and disagree about generalization and laws. The logical distinctions between particular and general statements; facts and laws do not correspond to an alleged historiographic knowledge of particular facts and the social-science provision of general laws. Historians agree on many theories, explanations and causes, while they also disagree on some of the chronicle "facts."

This distinction resembles the distinction in the philosophy of law between "easy cases" where any sane independent judge would rule identically and honest lawyers would advise their clients to settle and avoid unnecessary litigation, and difficult cases that are litigated, in the US context, before the Supreme Court. As in the case of realist philosophy of law, underdetermination leads to an empirical research program, examining the extent, kinds, and reasons for underdetermination in historiography.

Each of these possible epistemic philosophies of historiography is likely to support a different metaphysical interpretation of historiography. Several recent books considered the realism-anti-realism debate, that has been dominant in the philosophy of science, in the context of philosophy of historiography. To be clear, the philosophy of historiography is not interested in the reality of the past, this metaphysical dispute cannot be enlightened by a an examination of historiography. Rather, the philosophy of historiography examines the metaphysical assumptions of historiography. Metaphysics asks whether Abraham Lincoln existed, philosophy of historiography asks whether biographies of Lincoln assume that they describe the life of the late president, or just one possible interpretation of evidence that has, or does not have, equally valid alternatives. Historiographic determinists are likely to suggest that historiography assumes that its account of the past is a fairly accurate description of those segments and aspects of the past it chooses to deal with. These segments and aspects of history existed independently of the historians who described them. Historiographic indeterminists are likely to deny that historiography makes any metaphysical assumptions about the existence of the past, independent of its historiographic representation. Historiographic underdeterminists would claim that parts of historiography assume the independent existence of the past they describe, while other assume with varying degrees of certainty, the reliability of their account of that past.

I suggest to approach these epistemic and metaphysical problems empirically by examining: Where, how and why to historians agree on some parts of historiography, disagree on others, and are unable to communicate on history in some other cases? Mapping in which parts of historiography historians established a consensus and on which questions they disagree and where they cannot communicate at all can be established by a sensitive and sophisticated empirical study of historiography. "How" and "why" are interpretive questions that require using philosophical theories that relate to analogous discussions in contemporary philosophies of science and law and epistemology. The history and sociology of historiography should confirm or falsify the philosophical interpretation of historiography. Philosophers of historiography should examine the relations of historians in their institutional formation with historiography, how semantic differences are reflected in the sociological structure of the discipline (pragmatics).

The absence of consensus in some areas of historiography indicates at the very least that the core theories that define the historiographic community are insufficient to generate a single determinate interpretation of history. The theoretical basis of historiography is incomplete, and the additional theories that particular historiographic schools add to the core are controversial and gravitate towards the ad hoc pole. I do not think that long term disagreements result from different use of evidence (though different theories recognize different evidence), because following the recognition of a disagreement, historians tend to proceed with presentation of evidence that would then be shared. If disagreement persists beyond the presentation of evidence, it indicates deeper reasons for failure to reach agreement.

Historiographic core theories constrain the possible range of interpretations, but historiography on the whole, I argue, is underdetermined. Underdetermination by itself does not imply an anti-realist metaphysical interpretation. As Bunzl indicated, objective disagreements (as distinct of subjective disagreements when historians think they disagree though in fact they fail to communicate and understand each other) still assume a realist ontology, that there is a fact of the matter even if historians cannot agree on it.

Some of the theoretical foundations of historiographic schools or factions are sufficiently radically different to prevent communication among historians who work within different theoretical frameworks. The sociological-historical manifestation of this is the fragmentation of historiography among schools. Historiographic schools are associated with theoretical perspectives like those of Annales, Marx, history of mentalities etc. Unlike in science, where the evidential facts are idealized to fit the theory, in historiography theories (as in law according to Dworkin) are interpreted to fit particular cases, thus historians of historiography may notice a genealogic multiplication of theories and schools, down to the "atomic" individual level. If philosophers of historiography ignore the important distinction between use and mention in analysing historiographic statements (between the mentioning of identical symbols such as `the fall of the Maya Empire' and use of these symbols to convey different meanings by historians from different schools), they fail to detect the breakdown of communication between schools and end up with a confused representation of historiography. But rigorous study of what appears like debates among historians from different schools may establish the absence of communication.

The philosophic project may be to understand why is historiography so fragmented? Why do historians adapt some of their theories to the evidence rather than the other way round, causing increasing theoretic fragmentation and atomization? I explored elsewhere the possibility that the uniqueness (in a particular, philosophic, sense of unique) of some historical events and the complexity of some historical processes may have to do with the underdetermination of historiography.