Translator's Note
This translation of Persian Letters was made one letter at a time.
That is more than a description of method. It is the principle that shaped the edition. Rather than moving through Montesquieu's book as a single continuous block, I worked letter by letter against the French text and against John Davidson's 1892 English translation. The point was not to displace Davidson, whose version remains intelligent, readable, and often spirited. It was to return to Montesquieu with a closer ear, and to hear more clearly where Davidson illuminates the original, where he softens it, and where the textual state of his version now complicates the comparison.
The French base text used here is the Project Gutenberg transcription of Lettres persanes, Tome I and Tome II. Davidson's translation was consulted through Internet Archive witnesses of Persian Letters, Volume I and Volume II. Those French and English witnesses do not divide the work at the same point. The French text runs from Letter I to Letter LXXXVIII in Tome I, and from Letter LXXXIX to Letter CLXI in Tome II. Davidson's Volume I appears to end at Letter LXXV, with Volume II beginning at Letter LXXVI. Any comparison by physical volume therefore begins to drift at once. For that reason, this edition is aligned by letter number rather than by book division. Each letter has been treated as its own unit of translation, comparison, and editorial judgment.
That structure also made it possible to build a full per-letter corpus for the entire work. For each letter, the workspace preserves raw French excerpts, cleaned French excerpts, raw English OCR excerpts, cleaned English OCR excerpts, metadata, and comparison notes. Most readers will never need to consult that apparatus. Still, it matters that it exists, because it answers a basic editorial question: what sort of Davidson text are we actually reading? The answer is not a perfectly stable reprint. It is an OCR-derived witness that is sometimes clean, sometimes damaged, and occasionally interrupted by matter that plainly does not belong to the letter at all: footnote carryover, running headers, page numbers, printer's matter, library stamps, and even later editorial intrusion. Where recovery was secure, that debris has been removed. Where it was not, I have tried not to manufacture certainty. The cleaned English text is therefore a working editorial witness, not a diplomatic transcription.
That distinction matters because it helps explain why a new translation is worth attempting even when Davidson is at his best. He can be an excellent guide to the movement of a sentence. But he also has a recurrent tendency to elevate, regularize, and moralize. Sometimes the difference is local: a date drifts, a number shifts, a name settles into a more familiar form. Letter I offers a small but revealing example, where the French date corresponds to the twenty-fifth day and Davidson gives the twenty-sixth. Elsewhere the changes are tonal. In Letter II, a French image of the beloved as a "pillar" becomes Davidson's "monument." In Letter III, a word closer to "glory" becomes "conquest." None of these changes, taken alone, would justify a new translation. Taken together, they reveal a habit of smoothing Montesquieu's sharper edges into a more decorous nineteenth-century English.
That smoothing matters because Montesquieu's voice in Persian Letters is not stable, and is not meant to be. The book moves quickly between satire, political analysis, erotic recollection, theological dispute, travel observation, bureaucratic parody, and tragic confession. It changes pitch without warning. A translation that levels those shifts may become easier to read in one sense, but it becomes less faithful to the book's design. One aim of the present version has therefore been to resist that leveling. The goal is not literalism for its own sake, nor a stiff antiquarian English. It is a readable modern prose that keeps faith with Montesquieu's changes of pressure: the sudden turn from abstraction to mockery, from ceremony to violence, from cool observation to personal alarm.
The same balance has guided decisions about names and place-names. Older translations often preserve Latinate or French-inflected forms that now feel remote or unnecessarily opaque. Where a modern form is stable and genuinely helpful, I have generally preferred it: Qom rather than Koum, for example, or Tabriz rather than Tauris. But modernization has not been treated as a good in itself. Some older forms retain rhetorical or bibliographical value, especially when Davidson's phrasing is under direct comparison. The aim has been consistency without over-correction, and clarity without flattening the historical texture of the book.
The cleanup work behind this edition also made one further distinction impossible to ignore: textual repair is not the same thing as translation. In one pass, boundary errors had to be corrected where one letter spilled into another. In another, especially noisy letters such as XVI, LXVII, LXXV, CVIII, and CXXII had to be cleared of OCR debris and paratextual intrusion before they could be compared with any confidence. Those interventions were editorial, not translational. They were undertaken to recover a usable witness, not to bring Davidson silently into line with the new version. That limit seems worth stating plainly. An editor may remove what clearly does not belong; an editor should be more cautious when the text itself is uncertain. Throughout this project, I have tried to keep that distinction visible.
What I hope emerges from this process is a version of Persian Letters in which Montesquieu can sound a little nearer to himself in English: quicker in motion, harsher where he is harsh, more ironic where he is ironic, and less protected from his own tonal volatility. Davidson remains an important companion to that effort, and I have learned a great deal from reading him closely. But reading him closely has also made clearer why a fresh translation still matters. This book lives by its shifts of register, by its sudden rhetorical turns, and by formulations too pointed to survive habitual smoothing. If the present translation succeeds, it will not be because it has replaced Davidson. It will be because it has made both Davidson and Montesquieu easier to hear.