Handwriting analysis

Handwriting analysis becomes one of the most consequential and unstable forms of evidence in the Dreyfus case: first used to make Alfred Dreyfus appear guilty, then reopened by Georges Picquart’s comparison of the bordereau with letters by Charles Ferdinand Walsin Esterhazy. The novel follows not a steady advance toward certainty but a succession of interpretations, in which visible marks on paper are made to bear the weight of prior suspicions, institutional authority, and competing claims of expertise.

The accused made legible

Picquart had taught topography at the École Supérieure de Guerre, where Dreyfus was among his students. He remembered Dreyfus as thin, pale, solemn, and myopic, but could not recall their first meeting; familiarity had accumulated over weeks of looking down from the lecture podium at eighty faces. His impression of Dreyfus was already colored by class and antisemitism. Dreyfus was a married man of means among bachelors who struggled financially, and Picquart resented the family’s connection to German-occupied Alsace and the source of its money. When Dreyfus challenged a low mark, he asked Picquart directly whether being Jewish explained it. Picquart denied allowing prejudice to affect his judgment, but Dreyfus pointed out that the word Picquart used in his reply suggested the possibility remained.

That earlier encounter gives the later handwriting comparisons a personal history. When an officer proposed that the traitor might be one of Picquart’s trainees, Picquart reopened Dreyfus’s file, which he had previously skimmed and treated as having eliminated Dreyfus as a suspect. Looking again at the letters, he saw small writing, a rightward slope, and spacing that seemed similar to the bordereau. As he and Colonel Boucher walked to Charles-Arthur Gonse’s office, suspicion gathered force through association: Dreyfus’s family in Germany, his ambition, his perceived arrogance and cleverness. By the time Picquart reached the senior officers, he had nearly persuaded himself that these traits explained the handwriting. The comparison was not made in a vacuum; his own account shows how an apparent visual clue became entangled with an existing stereotype.

At the meeting, Picquart laid Dreyfus’s letters on Gonse’s desk. The officers gathered around and successively pointed to letter forms and spacing, including the capital s, j, small m, and r. Their remarks conveyed growing conviction, although several acknowledged that they were not experts. Jean Sandherr responded as though Dreyfus’s supposed guilt had become obvious all along, while Colonel Fabre invoked his earlier unfavorable assessment of Dreyfus’s character. The group’s convergence thus rested on more than handwriting: prior judgments about the accused were treated as confirmation of what the documents supposedly showed.

Expert judgment and the arrest

The handwriting evidence was contested even before Dreyfus’s arrest. At a meeting with Auguste Mercier, Armand du Paty de Clam reported that Alphonse Bertillon, head of the identification branch of the Préfecture de Police, believed the bordereau contained strong elements of Dreyfus’s handwriting and that its differences were deliberate. Mercier noted that another examiner from the Banque de France had concluded that the writing was not Dreyfus’s. The disagreement did not prevent the process from moving forward: Mercier had already prepared a warrant and asked whether Dreyfus had returned to Paris.

Picquart’s subsequent investigation supplied a different comparison. After obtaining permission to use letters by Esterhazy, he examined their neat, regular, well-spaced writing and initially thought it resembled Dreyfus’s correspondence. The similarity recalled the bordereau, which had been retrieved from Maximilian von Schwartzkoppen’s waste-paper basket and used to convict Dreyfus of treason. The document itself was a facsimile of thirty narrow, undated, unaddressed, unsigned lines. Its contents included military information, among it a note on the hydraulic brake of the 120 and a reference to the draft Field Artillery Firing Manual of 14 March 1894. These details made the handwriting comparison consequential: the document’s authorship connected the visible script to the alleged transfer of military information.

A leading handwriting expert in Paris had sworn that the bordereau was written by Dreyfus. Picquart placed its photograph between the two letters by Esterhazy and concluded that the writing was identical. This discovery reversed the direction of the inquiry: the same resemblance that had helped implicate Dreyfus now connected the incriminating document to another officer. The novel presents the apparent visual match as a turning point, but also keeps the distinction between seeing similarity and establishing authorship in view.

A comparison that changes meaning

The accusation against Dreyfus had also been reinforced by an alleged handwriting test during his arrest. Colonel du Paty de Clam asked Dreyfus to take dictation while his wrist was supposedly injured; Felix Gribelin and others watched. Gribelin later showed Picquart the document produced during the exercise and claimed Dreyfus’s writing changed halfway through, when he realized he was trapped and tried to disguise it. The dictated text began as an ordinary letter dated Paris, 15 October 1894. Picquart said he could not see the asserted change. The episode demonstrates how the same handwriting could be read as evidence of guilt even when the alleged alteration was not apparent to the investigator examining it.

Picquart then showed photographs of Esterhazy’s letters to Bertillon, who compared them with the bordereau using transparent squared sheets, a lamp, and a magnifying glass. After examining the documents, Bertillon declared the writing identical and agreed to sign an affidavit. Picquart made the implication explicit: the letters had been written in France that summer, not by Dreyfus. Bertillon did not abandon the original accusation. Instead, he proposed that Jews had trained another person to write using Dreyfus’s system. The explanation preserved his prior conclusion by making contrary handwriting evidence part of a wider conspiracy. In the scene, the method appears unable to correct itself: an apparent match with Esterhazy does not displace the theory that Dreyfus wrote the bordereau, but is absorbed into it.

Statistical certainty in court

At Dreyfus’s court-martial, Bertillon presented a further method for reconciling expert disagreement. He acknowledged that two handwriting experts had attributed the bordereau to Dreyfus and two had identified discrepancies and concluded he had not written it. His answer was a grid drawn across the document at five-millimetre intervals. He argued that repeated words, including manoeuvres, modifications, disposal, and copy, began in nearly the same positions within the grid squares. He assigned a one-in-five chance to that coincidence in one instance, sixteen in ten thousand across the repeated examples, and odds of one hundred million to one when other analyzed words were included. From those figures he concluded that the bordereau could not have been naturally written and was forged.

The courtroom effect did not depend only on whether Bertillon’s reasoning was persuasive. Picquart observed that the judges were soldiers, impressed by diagrams, measured squares, and the language of statistics. The large numerical odds offered a graspable appearance of certainty, while Picquart himself regarded Bertillon as seeming unhinged and remained unsure that conviction was more likely than not. Du Paty, by contrast, took the judges’ apparent favorable response as a sign that the prosecution had secured its case. The contrast exposes the gap between a method’s authority in court and its reliability as proof.

The limits of the evidence

Across these episodes, handwriting analysis is neither a neutral instrument nor a settled science within the narrative. Experts disagree; conclusions shift with the identity assigned to the writer; and evidence inconsistent with an established accusation can prompt an explanation that protects the accusation rather than tests it. The documents remain tangible—the letters, the thirty-line facsimile, the dictated page, the transparent grids—but their meaning depends on who examines them and what that examiner is prepared to accept.

The account also connects technical judgment to the unequal credibility granted to people in the case. Dreyfus’s direct challenge about antisemitism is met with Picquart’s claim to impartiality, while senior officers’ suspicions and Bertillon’s assertions carry institutional force. During the degradation ceremony, a Foreign Ministry observer interpreted Dreyfus’s submission as evidence against his innocence, and Sandherr explicitly treated Jewish identity as proof of disloyalty. Such prejudice did not merely surround the handwriting dispute: it shaped the assumptions through which conduct and documents were interpreted. Handwriting analysis consequently serves as both a clue in the search for the bordereau’s author and a measure of how easily an institution can mistake its own expectations for what the evidence says.