6.1Recognition is not identification
The two texts were written for different readers, and neither for a person who would one day ask which organism was responsible.
The distinction between recognition and identification is the article's central finding. Recognition, in Thucydides's sense, is the act of a person who meets a disease and knows it from a description: the course from head to chest, the inner burning, the thirst, the marks on the extremities, the danger to those who nurse the sick, the protection of survivors. It needs no name and no cause. Identification, in the sense of the modern disputes, is the act of assigning a known organism. It needs a name and, in the material arguments, a sample. The text was built for the first act and says so. Read against it, Morens and Littman's conclusion that the account does not allow a firm identification is not a disappointment with the text but a description of it (Morens and Littman 1992). The text declined to name the cause; the dispute over Athens is about whether anything else can.
For Florence the distinction runs differently. Boccaccio's text supplies signs but was built for belief; its strongest sign, the hogs, is offered as proof that the account can be trusted, not as a way of knowing the disease. Cohn's argument, as the note gives it, treats contemporary records as evidence of what the disease was like (Cohn 2002). Read against this text, the evidence is real but shaped by the purpose: the narrator selects what will make the reader believe.
6.2What the texts mark as seen, said and believed
Neustadt and May's separation of the known, the unclear and the presumed (Neustadt and May 1986) can be applied to the two texts as a reading device. Both authors mark these differences themselves, in words. The table gives the markers in the English, with the original where the original is sharper.
| What is marked | Athens (Crawley; Greek) | Florence (Rigg; Italian) |
|---|---|---|
| Seen by the author | "I had the disease myself, and watched its operation in the case of others" (αὐτός τε νοσήσας καὶ αὐτὸς ἰδὼν) | "had not many, and I among them, observed it with their own eyes" |
| Said by others | "It was said that it had broken out in many places previously"; "it is said" (ὡς λέγεται) of the origin; the rumour of poisoned wells | "though I had had it from the lips of a credible witness" (judged not enough to write); the sign of the East, unsourced |
| Agreed or believed | "is admitted to have been otherwise unprecedentedly free from sickness" (ὡς ὡμολογεῖτο) | "it is believed without any manner of doubt" (si crede per certo) |
| Left aside | causes, "I leave to other writers, whether lay or professional"; "the varieties of particular cases which were many and peculiar" | the choice between the heavens and God; the choice between an untreatable disease and ignorant physicians |
| Declared beyond counting | nothing counted at all | the sick, "whose number could not be estimated"; the dead at the doors, "more of them than he could count" |
The table makes one contrast plain. Thucydides never counts, and never says why. Boccaccio counts once, after saying twice that counting was impossible, and marks the one count as belief. Lipsitch and colleagues' warning that ratios made during an outbreak depend on what is counted as a case and on what is missed (Lipsitch et al. 2015) applies, in this setting, to a narrator who tells the reader directly that the sick were beyond reckoning.
6.3Where translation moves the evidence
The identification debates weigh features: an organ, a day of death, a route of spread, a sign. Reading the texts in two languages shows that several of these features move between the original and the English that most readers use.
At Athens, Crawley gives death on the seventh or eighth day where the Greek has the ninth and the seventh; the stomach where the Greek has a word that names the heart; reservoirs and wells where the Greek has wells and fountains; a student where the Greek has anyone who looks. At Florence, Rigg gives precautions where the Italian has counsels, and turns an impersonal per certo into belief without any manner of doubt, with no believer named in either language. None of these changes the general picture. Each changes a detail that a reader looking for a specific disease, or for a specific public measure, would use.
The point is not that the translators erred. The evidence in the disputes is not a fixed list of symptoms but a text in more than one language, and a reading of the account as epidemiological evidence has to choose which version of its details to hold.
6.4What the modern summaries keep and drop
The two records were written for a platform that asks of every case what was known at the time and what came later, and each keeps what that question needs. The Athenian record keeps the double authority, the purpose and the boundary between testimony and identification, and widens recognition from a disease to situations. The Florentine record keeps the measures, the season and the boundary between the period's explanations and verified causes. Both drop the symptoms, the animals and, for Florence, the estimate.
The study's ledger, which reads the records and not the texts, sees what the records keep. For The Plague at Athens, the two readings of the record agree only that it names no gap in time between the event and its record and that it gives no count; whether it shows a public act at all, and of what kind, they did not settle. For Florence 1348, the two readings agree that the record shows a public act, a restriction, taken before its cause or effect was established; that it names no gap in time and gives no count; and that it names a later finding that would change how the act should be read. Both readings code Florence as giving no count, which is true of the record and not true of the text. The appendix gives the values.
That last point is a modest finding about method. A ledger built on modern summaries inherits their omissions. The record of Florence holds no count because its author did not carry the narrator's estimate into it; the ledger, faithful to the record, codes no count. A reader who used the ledger to ask whether contemporary accounts of 1348 gave figures for the dead would get the wrong answer for this case. The text has to be read.
6.5What the evidence cannot carry
The article's evidence is two texts in two languages each, two modern summaries, and the ledger's values for the summaries. It cannot carry an identification of either disease, and it does not offer one. It cannot test the material claims, which rest on samples and methods this article does not assess, nor say whether Cohn drew on this passage, nor test Carmichael's argument about a later century. It reads Crawley and Rigg as they were served online, and the Greek and Italian in one edition each; other editions may differ, and this article has not compared them.
What the evidence can carry is narrower: what these two authors wrote, in what order, with what markers of sight, report, belief and omission, and where the English moves the details.