How Historians Weigh Eyewitness Accounts

Every war history rests on eyewitnesses: the soldier who wrote home, the officer who filed a report, the nurse who kept a diary, the veteran who sat for an interview decades later. Historians cannot do without them. They also know that two honest people standing a few yards apart can describe the same moment in very different ways. Weighing eyewitness accounts is less about catching errors than about understanding what each witness was in a position to know.

An old problem, first named long ago

The difficulty is as old as written history. The Athenian general and historian Thucydides, writing about the war between Athens and Sparta in the fifth century BC, explained that he did not take the first account that came to hand or even trust his own impressions without testing them. He also admitted the work was hard:

My conclusions have cost me some labour from the want of coincidence between accounts of the same occurrences by different eye-witnesses...

Thucydides, History of the Peloponnesian War 1.22, trans. Richard Crawley

He went on to name two causes: imperfect memory and partiality toward one side. Twenty-four centuries later, historians would add a few more, but those two remain near the top of the list. Our article on Thucydides and the birth of war history looks at his method more closely.

The historian's checklist

There is no formula that turns testimony into certainty. Historians do, however, ask a consistent set of questions.

The last question, about method, matters most when a writer does not just tell one story but summarizes hundreds. That is the situation at the center of one of the best-documented debates in American military history.

Case study: S.L.A. Marshall and the "ratio of fire"

The man and the method

Samuel Lyman Atwood Marshall, known as "Slam," was a newspaperman who became an Army combat historian during the Second World War. He is credited with pioneering the group after-action interview: gathering the survivors of a unit soon after a fight and walking through the action together, man by man, until the sequence of events was reconstructed. According to a 1989 account in American Heritage, he first used the method on Makin Island in the Pacific in 1943, and later with units in Europe.

Even Marshall's sharpest critics credit this technique as a real contribution. It put the experience of ordinary soldiers at the center of military history.

The claim

In his 1947 book Men Against Fire, Marshall made a startling assertion. He wrote that even in the most aggressive rifle companies, only about one man in four made real use of his weapon against the enemy, and that on average the share was closer to 15 percent. He presented this as a finding drawn from interviews with hundreds of rifle companies. He later reported that the rate rose to roughly 55 percent among American soldiers in the Korean War. The idea spread widely and influenced how the Army thought about training soldiers to fire.

The critics

Questions came decades later. In 1988 Roger J. Spiller, a historian at the Army's Combat Studies Institute at Fort Leavenworth, published an article in the RUSI Journal examining where the figure came from. He reported several problems:

• Marshall's surviving notebooks and papers contained no statistical tallies that could support a precise ratio.
• John Westover, who assisted Marshall with interviews in Europe, did not recall Marshall asking soldiers whether they had fired.
• Marshall's own count of company interviews shifted over time, from about 400 in Men Against Fire to 603 in 1952 and "something over 500" in 1957.

Spiller concluded that the systematic data collection Marshall described could not be shown to have happened. In 1989 historian Fredric Smoler brought the dispute to a wide audience in American Heritage. He also reported the doubts of Harold Leinbaugh, a former rifle company commander, and noted that veteran officers he consulted did not remember widespread failure to fire.

In 2003 John Whiteclay Chambers II examined the question again in Parameters, the U.S. Army War College journal. He drew on Marshall's papers and on the recollections of Frank Brennan, who accompanied Marshall in Korea in 1953. Brennan remembered that the question of who fired came up in passing during group discussion rather than being asked of every man. Chambers concluded that the Second World War ratio was probably a guess that the surviving evidence could not support. He also found that Marshall had overstated parts of his own record in his autobiography, which bears on how far his unchecked statements can be trusted.

The defenders and the middle ground

The other side of the debate deserves a fair hearing. Westover, while doubting the interview counts, felt Marshall's general conclusions held the essence of truth. In 1989 Marshall's grandson, a newspaper columnist, pointed to material in the S.L.A. Marshall Military History Collection at the University of Texas at El Paso as support. Chambers lists several officers and writers who continued to accept Marshall's view, including the author Dave Grossman, whose later work on the psychology of killing built on it.

A thoughtful middle position came from Kelly C. Jordan in The Journal of Military History in 2002. Jordan accepted that Marshall's methods were badly flawed, yet argued that his observation of higher firing rates in Korea was probably correct. In Jordan's view, the cause was not mainly training, as Marshall believed, but the Army's postwar reorganization of rifle squads and platoons, which gave more men weapons and positions likely to be used. His title sums it up: right for the wrong reasons.

Even Spiller, who opened the critique, credited Marshall with lasting contributions to the study of soldiers in battle.

What the debate teaches

Run the Marshall case through the checklist above and the lessons are clear. Marshall was close to the soldiers, and his interviews were held soon after combat, so proximity and timing were strong. The weakness lay in method: a precise number was presented as the product of systematic counting, but the counting could not be found. Corroboration was thin, because the figure rested largely on Marshall's own word. And his purpose mattered. Chambers suggested that Marshall wrote with a journalist's sense of what would grab attention.

Notice what the critics did not claim. They did not say the soldiers Marshall interviewed were lying, or that firing in combat is easy. They questioned how one man turned many conversations into one striking statistic.

Lessons for reading any eyewitness account

• Separate the story from the summary. A witness describing their own foxhole is on firm ground. A witness estimating what a whole battalion did is making a judgment, however honest.
• Ask how numbers were made. A precise figure deserves a precise answer to the question "How was this counted?"
• Look for independent agreement. One vivid account is a starting point, not a conclusion.
• Respect the witness while testing the claim. Questioning evidence is not the same as doubting a person's character. Our article on memory and the war memoir explains why sincere memories naturally shift.
• Know which kind of source you hold. Our guide to primary and secondary sources shows why Men Against Fire can be either, depending on your question.

Why it matters

Eyewitness testimony is not a weak form of evidence. It is often the only evidence of what war was like at ground level. The work of the historian is to honor it by taking it seriously: asking who saw what, when, why and how, and then setting each account beside others. The Marshall debate shows that this process can be uncomfortable, even for a respected figure. It also shows that careful questioning, done fairly, leaves us with a clearer and more honest picture of the people who fought.