A Scientific History: Problems, Possibilities, and Probabilities (Spring 1990)

[I submitted this paper as part of the requirements for a University of Florida grad student seminar in the Anthropology Department titled Seminar in Contemporary Theory (ANT 6286) taught by Marvin Harris in the spring of 1990. It expressed my frustration with the discipline of history. I remember how truly shocked Professor Harris and the anthropology grad students were when I would complain about how unscientific a discipline history was.]

When Thucydides combined two basic tools‑‑literary narrative and critical, objective analysis‑‑in his The History of the Peloponnesian War, he became the world’s first “scientific historian” and established the standards of the discipline of history for the next two and a half millennia. However, by so doing, Thucydides created a potential problem for history as a category of knowledge. Was history “art” or “science,” part of the “humanities” or the “social sciences?” Perhaps this epistemological difficulty presented no problem in ancient Greece and, ceteris paribus, historians would have continued contentedly to ply their trade in peace. But, while the eternal verities of Western art were being lovingly preserved, somebody, fortunately or unfortunately, let science out of Pandora’s box. ¹

Can history be both traditional art and modern science? The question addressed is not whether a scientific or nomothetic approach to historical problems is possible, for clearly such an approach is possible. Rather this essay analyzes the ups and downs of attempts at scientific history in the United States, the present state of scientific history, and whether a sustained effort at scientific history is possible within the academic environment now existing in the United States.

Most American historians have a very short‑sighted view of their own profession, concerned as they are mostly with works written within the past twenty years. Historians, in reaction to the present proliferation of “quantitative” and “social science” history, call today for a return to a more “traditional” history. They ignore that history became a professional academic discipline in the late nineteenth century United States, undistinguished from its sister social sciences, committed to the goals and principles of the natural sciences. Most historiographical reviews trace the origins of this American version of scientific history to the dominance of German training in critical methods among many early academics and the influx of Darwinian evolutionism (Grob, 1:6‑8). Higham finds, however, that “scientific history had already taken root in America outside of academic circles” by the mid‑nineteenth century, reflecting an independent but parallel “turn in American culture from romanticism to realism” (92).

Indeed, perhaps reflecting the general insular nature of the discipline in this country, American historians borrowed very selectively from their European counterparts. “Unlike their contemporary colleagues in England, France, and Germany, the Americans made not a single, sustained effort to discuss the nature of historical knowledge” (Higham, 98‑100). Always pragmatic and conservative, American historians rejected almost to a man the whole European notion of a “philosophy of history.”

Never strictly practicing the scientific method, this “scientific history” nevertheless rose well above Thucydidean critical, objective analysis. A scientific spirit ruled‑‑”impersonal, collaborative, secular, impatient of mystery, and relentlessly concerned with the relation of things to one another instead of their relation to a realm of ultimate meaning.” Although never abandoning narration as their basic communicative tool, they rejected “mere narrative” in favor of institutional, evolutionary analyses (Higham, 94‑97).

The dominance of this first “school” of history in the early twentieth century predictably led to the first intradisciplinary battles. The attack came from two directions: those who favored a return to “history as art” and those who favored a new kind of “science.” The “history as art” advocates, backed by the opinions of many laymen and amateur historians, proclaimed the “scientific” history “unreadable” and assailed “German pedantry.” Such criticisms, however, had relatively little impact on the profession (Higham, 104). The more important assault came from those historians who, reacting both to the rapid changes occurring in modern society and to the seeming rejection by most social scientists of diachronic studies, rejected the natural sciences as their inspiration and the supposed moral neutrality of older historians. They turned instead to a presentist, reform‑oriented, “progressive” history more in tune with the rapid advances being made in the social sciences (Grob, 1:9‑10; Higham, 111). “Political history” became the dominant form of expression.

Neither of these attacks represented a true rejection of “scientific” history. As Higham points out, the “real problem in early twentieth century America was not one of emancipating history from science, but rather the reverse: preventing science from repudiating history” (108). The “progressive” concern to catchup with the social sciences dominated much of American historical thought for the remainder of the twentieth century. Although history seemed to fall further and further behind, in many ways the discipline paralleled the ups and downs of the more turbulent social sciences. The “art” critics merely wanted to distinguish between “science as a method of investigation and art as a method of presentation” (Higham, 105). The “progressive” historians maintained an evolutionist, impersonal outlook, emphasized social and economic forces over institutions, searched for the “laws of history,” and remained distrustful of any “philosophy of history” (Higham, 110‑116).

However, during the 1930s and 1940s, the discipline underwent distinct changes. Diminished faith in “progressive” values in the face of world disorder led some historians to turn towards European “philosophy of history,” especially the works of Wilhelm Dilthey and Benedetto Croce. More generally, historians began affirming the uniqueness of history–neither an art nor a science but sui generis–and rejecting the utter subordination of history to the social sciences and the concomitant subservience of the past to the present (Higham, 117‑135). But far from reject­ing the social sciences, this new disciplinary “self‑respect” actually led to calls for even closer collaboration in the 1950s, albeit on a more equal footing (Higham, 137). Historians believed that the discipline could synthesize art and science employing “the historian’s supreme technical virtuosity” to fuse “the new method of social and psychological analysis with his traditional storytelling function” (Hughes, 77). History became more eclectic and pluralistic, borrowing randomly from philosophy, literature, European history, as well as all of the social sciences. “Intellectual history” became the dominant form in the 1940s and 1950s.

In the 1960s and 1970s, a new movement, led both by histor­ians and historically‑oriented social scientists, gave rise to a plethora of “new” histories (the “new economic history,” the “new political history,” the “new urban history,” etc.), culmi­nating in the all-encompassing “new social history.” Responding to the influences of the Annales school in France, advances in techno­logy, and general political turmoil, these historians believed that history could be more than national accounts of national politics or the elite biographies. They were determined to take scientific methods and theories and apply them to the analysis of the whole of historical society from the bottom‑up. History became behavioral, analytical, quantitative, and computerized.

By the 1980s, reaction set in against these “new” histories. The discipline was in the “utter turmoil” with no clear sense of direction. Traditional as well as many early “quantitative” historians repudiated statistical analysis as a proper historical tool. Marxist historians became disenchanted with the failure of the numbers to support their theories. Historians of all stripes charged that the “new” histories led to fragmentation of the discipline into specialized subfields and failed to deliver a new synthetic narrative along the lines of “progressive” history. Historians increasingly turned to post‑modernism and anti‑science cultural anthropology in the search for solutions to the mess created by the “new” histories.

So what did this century of “scientific history” accomplish? The answer is: not much. All this discussion of the major trends in the American historical discipline ignores the great bulk of traditional historians who continued to ply their trade in blissful ignorance or flat out rejection of the social sciences throughout the twentieth century. The strength of the narrative tradition, disciplinary defensiveness, the mathematical and scientific non‑aptitude of the great majority of historians, as well as the fragmented nature of the social sciences, prohibited the development of a truly “scientific” history. No social science theory came out of the historical discipline and few historians even bothered to rigorously test existing theory. Historians continued to use social science theory much as they originally used Darwinian evolutionism, merely as a descriptive, organizational tool.

The extent to which history has become “scientific” after a hundred years as a professional discipline can be readily shown by comparing and contrasting two recently published, highly respected attempts to apply social science methods to the study of the early American South: Bertram Wyatt‑Brown’s Southern Honor: Ethics and Behavior in the Old South and Darrett B. and Anita H. Rutman’s A Place in Time: Middlesex County, Virginia, 1650‑1750. Both Wyatt‑Brown and the Rutmans count themselves among the historians who follow the social sciences, but both believe that an historian should not wear his social science “on his sleeve.” Both Wyatt‑Brown and the Rutmans restrict their use of theory to “common sense” applications (Wyatt‑Brown, xiii; Rutman and Rutman, 25‑27). For the most part, they relegate their social science to the footnotes and simply let it “inform” their narrative. An historian, no matter how much he aspires to be a social scientist, is first and foremost an historian. In the end, all depend heavily on intuitive “historical imagination” (Rutman and Rutman, 13).

Both, however, deviate from traditional forms of historical narrative in order to incorporate a degree of explicit analysis. Wyatt‑Brown comes up with the “‘braided narrative,'” interweaving argument into the “plot” (Preface xiv). The Rutmans place themselves squarely in the historical narrative as “visitors,” in essence cultural anthropologists, to the community who choose “representative” examples of historical behavior that can highlight the essence of the community’s social structure.

Southern Honor is an excellent example of the direction that history is going in the 1990s. Deemphasizing material, infrastructural, and evolutionary explanations of historical change, Wyatt‑Brown stresses the “continuity of human ethical principles” and superstructure over time (Preface xi). “Southern honor” descended directly and virtually unchanged from a prehistoric Indo‑European ethical code. The ethic entered the New World primarily through the heavy migration of Celtic peoples (Irish, Scottish, Scotch‑Irish, and Welsh) to the British colonies in the 17th and 18th centuries.

Wyatt‑Brown draws heavily upon Julian Pitt‑Rivers’s essay in the International Encyclopedia of the Social Sciences which defines “honor” as a complex of substantive and normative attributes, “simultaneously” emic and etic, mental and behavioral (Wyatt‑Brown, xiii, 502n10, 509n38; Pitt‑Rivers, 503). Honor is “both internal to the individual and external to him–a matter of his feelings, his behavior, and the treatment that he receives” (503). Pitt‑Rivers variously refers to honor as a state of individual conscience, civic virtue, royal favor, military prowess, social dignity, economic advantage, courage, state of grace, inheritance, personal autonomy, symbolic representation, sacredness, social status, steadfastness, beneficience, patronage, code duello, fidelity to individuals, “universal principles of social action,” and “mirror to an age” (503‑510). Similarly, Wyatt‑Brown defines honor as everything from “the cluster of ethical rules, most readily found in soci­eties of small communities by which judgments of behavior are ratified by community consensus” (Preface xv) to “a vital code” that “gave meaning to life” (114) to “a state of grace linking mind, body, blood, hand, voice, head, eyes, and even genitalia” (49).

Although Wyatt‑Brown’s overall goal is to link southern mind and behavior, as a student of his notes, overall he concerns himself more with southern “myths” than realities (Morris, 3). As Wyatt-Brown himself so aptly states, “History shorn of all myths and emblems is, happily, impossible” (Wyatt-Brown, 113 [emphasis added]). He makes no attempt to connect his code of honor to material conditions, beyond a couple of allusions to “rough living.” He does not try to analyze change over space or time in the notion of honor.

Honor, as Wyatt‑Brown uses the concept, can not be analyzed scientifically. To dissect honor into its emic and etic, mental and behavioral, components is to destroy the concept. To capture the “whole,” Wyatt‑Brown draws heavily upon great literature from Homer and Tacitus to modern southern writers like John Kennedy Toole and Walker Percy. Southern Honor rests most solidly on the essence of pre‑modern American captured in the fictional works of Nathaniel Hawthorne and William Faulkner. “Through metaphor and felicity of language, the novelist’s imagination can recreate the way people once thought and acted, so ordering matters toward an ethical veracity that the historian could never achieve” (Preface xi).

The Rutmans take a quite different approach to history. Eschewing such overarching concepts as honor, they employ “community” as a value‑free, analytic framework to study the networks of relations within Middlesex County (25). Where Wyatt‑Brown draws upon the “soft,” interpretive science (some anthropologists would say “anti-science”) of Clifford Geertz, Mary Douglas, and Victor Turner, the Rutmans lean more to the “hard,” materialist science of Marvin Harris. The Rutmans are very careful to distinguish between observer and observed, emic and etic persepctives (without using those terms), and mental and behavioral activity. Although they present “vignettes” to capture “some part of the mentality of the time and place” (34), the emphasis is on strict etic behavioral analysis. Overall, compared to Wyatt‑Brown, the Rutmans have a narrower focus, ask less broad questions, and are more careful in their generalizations.

More importantly, Wyatt-Brown and the Rutmans take totally different points of view on testing the generalizations they do make. Both authors justify their general conclusions by drawing examples of certain prevalent behavioral activities. But whereas the Rutmans’ examples reflect a thorough statistical analysis of all available records, Wyatt‑Brown shows a great deal of incon­sistency in his choice of representative cases. Lacking any rigorous method of determining bias in literary sources, Wyatt­-Brown chooses those examples which support his prejudged conclu­sions while ignoring or dismissing evidence which might tend to refute. The few instances where Wyatt‑Brown does raise the potential for bias in his data make the lack of such a method even more apparent. He accepts at face value evidence from family papers reporting the “darker” side of southern family life (upon which he builds his concept of the pre‑modern, patriarchal southern household), but calls “biased” evidence from the same sources suggesting the presence of true affection in child rearing and marital relations (146‑7, 224).

More potentially verifiable behavioral generalizations drawn from the same literary sources, which seemingly play a crucial role in his desire to connect southern “myth” with southern “reality,” likewise go unverified. For example, Wyatt­-Brown believes that the prevalence of intermarriage among cousins proves the “intense Southern desire to increase the bonds of family reliance and to enhance both career and marital success by concentrating wealth,” but admits unfortunately he has no data to confirm this prevalence (217). Usually qualified with “probably” or “we may speculate,” Wyatt‑Brown leaves to future historians the task of testing and confirming these behavioral hypotheses (Preface xv).

In contrast to qualified behavioral statements, etic mental generalizations, whether based on “believable” behavioral data or not, usually appear in Southern Honor as simple facts with no need for verification. Wyatt‑Brown regularly delves deep into the Freudian psyche of southern men and women to learn their true subconscious or unconscious motivations‑‑for example, the “paternal egocentrism” which drove fathers perpetually to seek western lands or the repressed anger which drove mothers to seek “largely unconscious revenge” against a patriarchal society by sending their boys off to be killed in war (172‑174, 198).

Overall, the weight of so much selective and suppositional data makes discerning fact from fiction in Southern Honor indeed difficult. Faulkner serves the argument as well as evidence from family papers or court records. Like other popular Freudian and Gramscian theories, “southern honor” is basically a circular, non‑verifiable theory which can be neither proved nor disproved, so rejection or failure to reject any of the more testable hypotheses would have negligible effect on Wyatt‑Brown’s central theme. Purely an idiographic and idealist “work of art,” Southern Honor can thus make no contribution towards the development of a “scientific history.”

A Place in Time, on the other hand, represents an outstanding example of what Peter Laslett calls “social structural history” which can contribute to the social sciences by providing “social scientists with comparative instances and with a continuum in which they can do their work” (439‑440). But even “social structural history” does not move very far towards making history “scientific.” Towards that goal, Laslett finds most important the historical work being carried out by social scientists themselves (436). Of this work, the “new economic history” undoubtedly represents the “sharpest contrast to traditional historical pursuits” and the greatest effort towards explicitly developing a truly scientific history (Laslett, 438‑439; North, 468‑474).

The differences between the “new economic history,” dominated by economists, and the other “new” histories, dominated by historians, is patently obvious. Only the “new economic history” rigorously tests hypotheses drawn from a coherent corpus of theory (basically neo‑classical economic theory) in order to develop better theory. The “new economic historians,” while not necessarily dismissing narrative, emphasize more the heavily statistical monograph. Likewise, although not abandoning traditional historical chronological and regional divisions, these “cliometricians” feel little compulsion to force their analyses into such narrow rigid frameworks. ²

The nature of the relationship between the “new economic history” and the historical discipline reflects the near impossibility of history becoming truly scientific. Historians have heavily criticized almost every aspect of this “quasi history,” although Laslett shows that many of the criticisms might be directed at traditional history as well (438‑39). The findings of the “new economic historians,” apart from a few like Gavin Wright who have attempted to bridge the disciplinary gap and have made themselves amenable to historians, go unincor­porated in the general historical literature. History departments across the nation are altogether abandoning the subdiscipline of economic history.

Overall the mid‑1970s controversy over the publication of Fogel and Engerman’s Time on the Cross, a “new economic history” of antebellum southern slavery, brought to a head many of the problems with history as a social science. Firstly, the scientific analysis of the whole of society, like the analysis of southern slavery, meant dealing with masses of data, particularly quantitative data, with which historians were still not attuned either by ability or inclination. And the social scientists who attacked Fogel and Engerman confirmed for historians, already suspicious of “quantitative history,” that numbers can lie. Second, historians began again to question the purpose of history. If scientific history meant the potential for coming up with politically disagreeable answers (e.g., finding that slaves prospered under slavery in the light of highly sensitive race relations), with actually allowing the historical data to deter­mine the historian’s conclusions rather than finding data to fit pre‑conclusions, then perhaps scientific history was not the way to do history. The furor essentially killed any serious discus­sion of how historians and social scientists were to proceed past problem areas in the “scientific” analysis of history.

Before Time on the Cross, Robert Fogel believed that humanism could be made subservient to the social sciences. But, after all the furor, he admitted that he and Engerman had been naive. “We have come to recognize that history is, and very likely will remain, primarily a humanistic discipline. We now believe that the issue raised by historical quantifiers is not whether history can be transformed into social science but the realm of usefulness of social‑science methods in a humanistic discipline” (Fogel 341‑342).

Although perhaps a little melodramatic, Fogel’s remarks accurately capture the nature of the historical discipline in the twentieth century. Historians’ perpetual trumpeting of the social sciences resulted in little more than “a century of continuous failure” (Hexter 1971,108). History has always basically been a synthetic art with the goal of creating a national culture for the purpose of shaping modern society. Present historical debates center, not around theory or method, but simply over political disputes that even the best “scientific” research can not resolve.

Historians approach their historical world much as they would approach the present, with an intuitive, “common sense” knowledge of the way the world works, which they then try to communicate through the device of the historical narrative (Hexter 1968,371‑72). History as a discipline has always been idiographic and adamantly atheoretical, concerned with the unique and particular. (Hexter traces this disciplinary essence to the imperfect nature of historical sources and the catholicity of history.) Indeed, historians even resist attempts to codify historiographic principles beyond “Avoid plagiarism.” To claim universality is to seriously risk being labelled “ahistorical.” Historians are eclectic, pragmatic, and utilitarian, borrowing or not borrowing from the social sciences depending on the purpose for the history (entertainment, education, propaganda, etc.), the audience, the ability and inclination of the historian, the quality and quantity of historical data, and a host of other factors. They are not interested in the whole body of knowledge of the social sciences for they do not aspire to be social scientists.

Historians’ questions are usually too broad in scope to be testable. Historians also simply have less control over their data than modern natural scientists or even social scientists. There are only certain types of hypotheses that can be tested scientifically, and these are often not the most interesting. When his interpretation of Jacksonian democracy came under attack from behavioral, quantitative, “scientific” research, Arthur Schlesinger, Jr. aptly replied: “As a humanist, I am bound to reply that almost all important questions are important precisely because they are not susceptible to quantification” (33).

The disciplinary structure of modern academia also inhibits the growth of “scientific” history by discouraging true interdisciplinary research. Historians of all stripes defend the uniqueness of history and even the most “scientific” of historians refute efforts to turn history into a social science (Berkhofer 4‑5). Laslett suggests that all forms of “scientific” history are not an alternative to “traditional historical writing which should, and ultimately will, replace it” (440). Darrett Rutman, a true believer in a “scientific history,” thinks that the discipline can become scientific strictly in house, proceeding “inductively rather than deductively, upward from small observations and small theories…toward and eventually to middle‑ and higher‑level theories,” like an inchworm (22). The message is clear: we will work with you, but only in our own way, for our own purposes.

Nevertheless, refusing to drop out of the modern societal race to become more “scientific,” history struggles to keep up with the social sciences, always lagging behind, randomly incorporating those methods and theories which suit the particular needs of individual historians. As the social sciences themselves have fragmented, historians have had more and more choices, leading to greater fragmentation among historians nominally practicing “social science history.” The rise of the post‑modernists has been a godsend for the last holdouts, the most anti‑scientific of historians. For now all historians can claim to find support for their basically intuitive statements from some branch of the social sciences, even if “social science” no longer means anything.

I still believe, as naive as it may seem after this eulogy, that history can be scientific even within the present system of strong disciplinary barriers, although the process will require the strengthening of interdisciplinary, paradigmatic networks. Many historical questions demand scientific analysis and many scientific questions demand historical analysis. The question is will the historical discipline lead the way or wait until social scientists find the time and interest.

Undoubtedly there will always be methodological problems with a scientific history: codification, measurement error, high levels of unexplained variance, poor theory, ecological fallacy, etc. (Smith, 141‑148). But the success of certain social scientists, like “new economic historians” Gavin Wright and Howard Kunreuther, in showing that historical analysis can indeed help develop better middle‑level theory which can find acceptance among historically-oriented social scientists and scientifically-oriented historians, gives hope. ³

A scientific history will be realized by bringing together social structural histories, like A Place in Time, with the more theoretical efforts of social scientists, such as the “new economic history,” within a scientific paradigm like cultural materialism. Historians will still be needed to synthesize and disseminate the findings of these scientific monographs. There will still be plenty of room in academia for traditional historical studies and synchronic social science studies. Traditional history will still maintain the core of the discipline with scientific historians manning the fringes, crossing interdisciplinary boundaries to become full‑fledged members in the pursuit of the scientific analysis of man. Although this examination of the history of the discipline over the past hundred years makes the possibility of this transfor­mation happening anytime in the near future appear highly unlikely, the cause justifies the effort.

Works Cited

Berkhofer, Robert F., Jr. A Behavioral Approach to Historical Analysis. New York: Free Press, 1969.

Grob, Gerald N. and George Athan Billias, eds., Interpretations of American History: Patterns and Perspectives. 5th ed. 2 vols. New York: Free Press, 1987.

Harris, Marvin. “Anthropology: Ships That Crash in the Night.” Unpublished manuscript.

Hexter, J. H. “The Rhetoric of Historiography.”International Encyclopedia of the Social Sciences. David L. Sill, ed. New York: Macmillan, 1968. 6: 368‑394.

‑‑‑. Doing History. Bloomington: Indiana UP, 1971.

Higham, John, Leonard Krieger, and Felix Gilbert. History. Englewood Cliff, N. J.: Prentice‑Hall, 1965.

Hughes, H. Stuart. History as Art and as Science: Twin Vistas on the Past. New York: Harper & Row, 1964.

Laslett, Peter. “History and the Social Sciences.”International Encyclopedia of the Social Sciences. David L. Sill, ed. New York: Macmillan, 1968. 6: 434‑440.

McGuire, Robert and Robert Higgs. “Cotton, Corn, and Risk in the Nineteenth Century: Another View.” Explorations in Economic History 14 (1977): 167‑182.

Morris, Christopher. “The Southern White Community in Life and Mind: A Critical Review of Recent Literature.” Canadian Review of American Studies [forthcoming].

North, Douglass C. “Economic History.” International Encyclopedia of the Social Sciences. David L. Sill, ed. NewYork: Macmillan, 1968. 6: 468‑474.

Pitt‑Rivers, Julian. “Honor.” International Encyclopedia ofthe Social Sciences. David L. Sill, ed. New York: Macmillan, 1968. 6: 503-511.

Rutman, Darrett B. “The ‘New Social History’ in America.” Unpublished manuscript. [Forthcoming in Novaya i NoveishayaIstoriya in Russian.]

Rutman, Darrett B. and Anita H. Rutman. A Place in Time: Middlesex County, Virginia, 1650‑1750. New York: W. W. Norton, 1984.

Schlesinger, Arthur, Jr. “The Humanist Looks at Empirical Social Research.” Quantification in American History: Theory and Research. Robert P. Swierenga, ed. New York: Atheneum, 1970. 30-35.

Smith, Daniel Scott. “A Mean and Random Past: The Implications of Variance for History.” Historical Methods 17 (1984): 141‑148.

Wright, Gavin and Howard Kunreuther. “Cotton, Corn, and Risk in the Nineteenth Century: A Reply.” Explorations in Economic History 14 (1977): 183‑195.

Wyatt‑Brown, Bertram. Southern Honor: Ethics and Behavior in the Old South. Oxford: Oxford UP, 1982.