The Fact/Narrative Distinction and Student Examinations in History

IN THE SUMMER of 2006, Jeb Bush, governor of Florida and brother of the president, signed an Omnibus Education Bill barring historical interpretation in his state’s public schools.[1] “The history of the United States … shall be viewed as factual, not as constructed,” read the final version (the draft, written by state senator Mike Fasano, would have mandated that history teaching “not follow the revisionist or postmodernist viewpoints of relative truth”).[2] Predictably, professional historians across the country cringed at such language. Even the most epistemologically conservative members of our profession, those who snicker at the mention of Jacques Derrida’s name, have a hard time agreeing that history consists solely of facts.

When pressed to explain our reactions to measures like the Florida bill, historians inevitably give some variation of the same response. History, we say, is no more the mere discovery and transmission of facts than cooking is the discovery and transmission of ingredients. While it is true that every history is composed of facts, there is something more to history, something that makes it more than just the sum of its factual claims. That something has a technical name in our profession: we call it narrative.

A narrative, we dutifully explain, is not, as the word may suggest, simply a story (although it may be that). Rather, it is a structure for organizing factual claims. It is the spine of every historical work—the recipe that directs the combination of the ingredients, the blueprint that regulates the placement of bricks. It is what tells us which facts ought to be included, which excluded, and how they ought to be related to one another. It tells us which facts are significant and which can be safely ignored or mentioned only as examples. The following slogan captures the point: facts tell us about the past, narratives tell us what the past is about.

As many historians have noted, it is impossible to write coherent and legible history without relying on narratives.[3] One cannot simply teach facts about the past without indicating their significance and relation to each other, without giving a sense of what they are about. No executive fiat, such as the one Florida has recently issued, can change that. Whether the history of the United States is told as the story of expanding freedoms or, say, as the story of the transformation of an agrarian nation into an industrial one, some narrative will inevitably govern the selection of facts to be included and guide their presentation and relation to each other.[4]

That is not to say, however, that laws like Florida’s have no effect. By legislating the impossible (that history be exclusively factual), they push us toward the most familiar and traditional of our narratives—the ones that have been around so long that they seem as if they are natural facts rather than artful constructions. Four paragraphs before the Florida bill decrees that history ought to involve solely facts and no “construction,” for example, it mandates that Florida schools teach students how the principles of the Declaration of Independence “form the philosophical foundation of our government.”[5] We are so familiar with narratives that explain the character of our nation in terms of the Declaration of Independence and the Constitution that they may seem like unquestionable facts to us. When we hear a historian offer an alternative narrative, though—one that emphasizes the proslavery aspects of both documents, perhaps—we are quicker to recognize it as an interpretation, perhaps a “biased” one. But both are narrative interpretations, of course.

All of this is known to practicing historians. The academic subfield that takes such issues as its primary object of inquiry is a small one, but the fact/narrative distinction can be said, I think, to be part of the practical knowledge that every historian will pick up on the job even if she has not been formally instructed in it. Simply put, the centrality of narratives to history-writing is obvious to anyone who has ever had to write history, just as the importance of heat is to anyone who has ever worked as a chef.

But although the fact/narrative distinction is instinctively grasped by all practicing historians, we could, I contend, do a better job of passing it onto our students. In particular, I will argue, the examinations most often given to history students in college-level courses do not test a student’s grasp of historical narratives as directly as they might.

Consider two students who are asked to identify the Boston Tea Party on an examination. One student places it in December of 1770, another places it in December of 1776. Both have given the wrong date by three years (it actually took place in December of 1773), and to an outside observer it may seem that both should have the same amount of points docked. But to a U.S. historian, those two errors are not equivalent. The first may be an acceptable slip, whereas the second demonstrates a major failure to understand the importance of the Boston Tea Party. Why? Because the student should at least be expected to know that the Declaration of Independence was signed on July 4, 1776, and anyone who thinks the Boston Tea Party happened after the signing of the Declaration of Independence has clearly failed to grasp a great deal about the revolutionary period. Both students have got the fact wrong (by the same margin), but the second student has done something worse—she has got an important narrative wrong as well.

Now consider a third student, who correctly writes that the Boston Tea Party occurred in December of 1773. She has got the date right, clearly, but has she got the narrative right? If she did not write anything else, then we cannot know. Maybe, when she was studying for the exam, she was able to remember when the Boston Tea Party occurred because she knew that it happened after the Townshend Acts, before the Intolerable Acts, and certainly before fighting broke out at Lexington and Concord. But maybe something much simpler was going on. Maybe she just has a good memory for dates.

That third student illustrates the problem with many exam questions. As instructors, it is important for us to distinguish between students who can repeat dates and those who actually understand narratives. And yet, our tests often render the two indistinguishable. A student with a knack for committing facts to memory but who is completely oblivious to narratives will do as well as the student who has a deeper understanding of the period. And because we reward the two equally, students may be encouraged to favor the first strategy over the second on the grounds that it seems easier (although historians know that it is not). It should not come as an enormous surprise when the students who pass through our history classes go on to write legislation insisting that history consists of facts and nothing more. For most students, that view of history is compatible with a perfectly serviceable strategy for getting through history classes.

The complaint that we do not test directly for narrative understanding should not be confused with the claim that we do not test for it at all. As the second student from the above example demonstrates, it is entirely possible for students to reveal a failure to understand narratives even on the most seemingly fact-based questions. More importantly, few instructors ask students to fill in dates only. The standard undergraduate examination in history consists of a set of terms to be identified (what, where, when, significance) and an essay. Not only will a command of narratives aid students enormously in the task of memorizing a series of facts about the past, but it will also, we hope, be demonstrated in the essay. Nevertheless, it is often the case that a student with a good memory but poor comprehension can perform as well on such examinations as a student who truly understands the narratives. Any student who writes down enough of what the teacher says should be able to repeat the correct portion of the lecture with enough detail on the exam.

The problem of how memorization (facts) can be sorted from comprehension (narratives) is not a novel problem in pedagogy. Mathematicians deal with it all of the time. Many low-level math problems can be solved by a student who has memorized answers in advance. For example, a student who has memorized the multiplication tables up to 12 will be able to solve unerringly most basic multiplication problems without having the slightest clue as to how multiplication works. But, of course, math instructors have found a simple solution. They ask questions that cannot be answered by someone who has merely memorized a few answers. A student who has only committed the multiplication table up to 12 to memory will not be able to say what 12 × 13 is, but a student who understands how multiplication works will have no difficulty (adding 12 to the product of 12 × 12, for example, will produce the right answer). Math instructors have devised enough time-tested ways of distinguishing memorization from understanding that it would be hard to imagine a student walking out of one of their classes under the mistaken belief that mathematics consists solely in the repetition of numbers and algorithms.

We historians have much to learn from our colleagues in mathematics (and in other sciences, as well). We need to learn how to test the comprehension of narratives in isolation from the memorization of facts. Following are three examples of questions from the subject of U.S. history that we might ask in order to test directly a student’s historical comprehension. In these examples, we assume that the claim the student is asked to scrutinize is an unfamiliar one.

Example 1

“Those who labour in the earth are the chosen people of God, if ever he had a chosen people, whose breasts he has made his peculiar deposit for substantial and genuine virtue.” Who is more likely to have said that, Thomas Jefferson or Alexander Hamilton? Why?

Here we imagine that the student has not read Jefferson’s Notes on the State of Virginia (from which the quotation is taken) but has been instructed in the ideological differences between the two revolutionaries. Unlike many similar exam questions, in which the student is asked to recall where she read the quotation in question, the above question asks the student to deal with a quotation that she has not read. Sheer memorization will not help her; she will have to reflect on what she knows about the two men and particularly upon the differences in their systems of political economy.

Example 2

The following three assertions about the Civil War may be true or false. Assess the veracity of each and give your reasons for reaching your decision.
�����1) The Confederacy manufactured 3,000 more guns than did the Union during the war.
�����2) At great personal expense, planters in the northern part of the Confederacy arranged to have their slaves removed from their plantations and sent south into the Confederate interior.
�����3) A week after the first guns fired at Fort Sumter, Lincoln publicly declared slavery to be “an affront to God and justice everywhere it occurs.”

Again, we imagine that the student has not encountered any of these assertions before. The question does not ask the student to repeat facts, but to adjudicate between true and false claims. As before, memorization alone will not do the trick. The student has to know why it is implausible that the Confederacy manufactured more guns than did the Union, why holding slaves near the front lines would have been a difficult proposition for planters, and why Lincoln, who depended on the support of slaveholding border states, would never have made such a public claim about slavery early in the war.

Example 3

In her book Women and Economics, Charlotte Perkins Gilman claimed that women, through their economic dependence on men, had become biologically inferior. Because they were forced to win their subsistence by sexually attracting men rather than earning salaries or wages, she argued, their sexual characteristics were exaggerated at the expense of other biological characteristics. Gilman believed that women passed down these traits to their daughters and were thus largely left out of the biological progress made by mankind. In what time period would you place such a book: 1775–1790, 1835–1850, 1890–1905, or 1945–1960? Why?

Students who can recognize in that summary the intellectual strains of Darwinism should be able to realize that Gilman’s book was published in the latter half of the nineteenth century or the early part of the twentieth century (it was published in 1898). Any number of other narratives—perhaps those relating to the history of feminism or to social science—might also enable the student to answer the question.

The sample questions above have a few important things in common. First, they are not easy, at least not to an outsider. They require the student to be familiar with specific narratives, and it is possible that a bright undergraduate who took a different history course that did not stress those particular narratives would be unable to answer the questions. But if questions testing narrative comprehension turn out to be more course-specific than those testing factual knowledge, that is just a consequence of the fact that there is a healthy diversity of ways to narrate the past and that not every teacher will do it in the same way.

A second common strain in the above questions is that they all ask the student to distinguish the plausible from the implausible. The student is not expected to know the answer merely from having studied it beforehand but must rather discover the correct answer by thinking about the historical period in question. Memorizing facts may help, but only to the degree that they help the student remember or compose a narrative about the past. In other words, the student can answer the questions only by thinking historically, by thinking about patterns and processes. Such questions exploit the convenient fact that, while factual claims about the past only tell us about the things they are describing, narratives can enable us to make sense of factual claims that we might never have encountered before. By asking a student to evaluate facts with which she is not already familiar, we can isolate narrative understanding from factual knowledge.

None of what I have written should be taken as a denigration of facts. Some amount of memorization will clearly be necessary for any student of history. But it would do a disservice to our students and our profession to emphasize only facts on the exams we compose. Testing directly for narrative comprehension as well as for factual mastery will allow us to more accurately assess whether our students are learning what we aim to teach them. It will also send our students the message that narratives are an important part of historical thought. And perhaps it might make them less likely to write absurd laws as well.


1.� Thanks are due to Scott Armstrong for the stimulating discussions that led to this article and to Anders Stephanson and John Immerwahr for their guidance in matters historiographal and pedagogical, respectively.
2.� For the bill and some historians’ perspectives on it, see Robert Cassanello, “Education Reform and the History Wars in Florida,” OAH Newsletter 34 (August 2006): 17 and Mary Beth Norton, “History under Construction in Florida,” New York Times, 2 July 2006.
3.� On this technical point, there has been some minor debate. Whether a list-like set of claims such as “Abraham begat Isaac; and Isaac begat Jacob; and Jacob begat Judas and his brethren; and Judas begat Phares …” should count as a historical narrative or not is the subject of some disagreement. All agree, however, that the sorts of things that historians typically write nowadays ought to count as narratives.
4.� Facts and narratives have received a great deal of scrutiny from the few historians who engage in these philosophical matters, particularly after the publication of Hayden White’s controversial Metahistory in 1973. Some controversies remain and a number of scholars, particularly Arthur Danto, have noted that statements that appear to be atomistic facts often contain narrative elements. Nevertheless, the existence of narratives as functionally distinct aspects of historical writing separate from facts remains non-controversial. One would be hard-pressed to find a practicing historian who believes that there is no narrative element in history. To my mind, the best examination of such issues is contained in F. R. Ankersmit, Narrative Logic: A Semantic Analysis of the Historian’s Language (The Hague: M. Nijhoff, 1983). Readers pursuing these issues further may wish to consult W. H. Walsh, Philosophy of History: An Introduction (New York: Harper, 1960); Hayden White, Metahistory: The Historical Imagination in Nineteenth-Century Europe (Baltimore: Johns Hopkins University Press, 1973); Arthur C. Danto, Narration and Knowledge (New York: Columbia University Press, 1985); Michael Stanford, Introduction to the Philosophy of History (Cambridge, Mass.: Blackwell, 1998); and Geoffrey Roberts, ed., The History and Narrative Reader (London: Routledge, 2001).
5.� Those principles, the bill goes on to specify, include “limited government” and the “inalienable rights of life, liberty, and property.” Florida State Congress, House Bill 7087 (2006): section 22, paragraph 2a.



By Daniel Immerwahr