Photographic credit: Bibliothèque de l’Institut de France

Editorial principles

Transcription principles

The letters or handwritten texts found on this site, as well as printed texts, have, as a general rule, been transcribed with the greatest possible fidelity to the original document. Only a few changes have been made for the purposes of legibility. These changes are detailed below.

- Capital letters have been restored at the beginning of sentences as well as names of people,[1] places, and periodicals [2].

Other capital letters which were in common use in the eighteenth century have been kept, particularly in the case of subjects and personal pronouns (“Je”, “Vous”, “Votre”, etc.), names of days and months (“Mardi”, “Janvier”, “Juillet”, etc.), titles of nobility (“Comte”, “Marquis”, etc.), and certain words that may have conceptual value (“Cercle”, “Quarré”, etc.).

For everything else, decisions have been made on a case-by-case basis. Some wrote certain words, including the most common ones, in fairly distinct capitals and we have observed this practice. On the other hand, others (including Condorcet) wrote certain letters (in particular “C”, “L”, and “S”) in such a way that it is frequently impossible to tell whether they are upper or lower case. In such cases we have transcribed these letters using lower case.

- Words that were linked up in the process of writing have been separated, except in two cases: if they correspond to a spelling found in the eighteenth century (“aulieu”, “ensorte”, etc.) or if this is due to the absence of an apostrophe which we have therefore not restored. Split words (“le quel”, “long tems”, etc.) have been left as such, and likewise the graphemes “oe” and “ae” not merged into ligatures (“œ” and “æ”).

- When taking the form of a dot, a vertical or horizontal stroke, or the extension of a letter, acute and grave accents have been restored in keeping with modern rules of writing. Circumflex accents and diaereses, which can also take on various forms, have been homogenised.

- Punctuation has been restored or corrected only in exceptional cases, particularly when a text was likely to be misread. Delimiting marks signalling the end of a text (“^./^.”) have been changed into dots, and likewise for lines also drawn at the end of a text. When, at the end of a sentence, the last letter of a word was extended, a full stop has been added if missing. Dots associated with certain abbreviations have been systematically restored before superscripts (e.g. “M.^r” rather than “M^r.”). Dots used in the writing of some dates – a common practice in the eighteenth century (e.g. “8. Juillet. 1784.”) – have been kept.

- Lines drawn to underline words or a series of words, or for the purposes of postmarks or notes concerning the persons represented by some signatures have been kept.

- Undeciphered beginnings of words and ink stains have not been flagged up except when the text would become unreadable.

- Mathematical formulas have been transcribed using a mathematical font, i.e. close to italics.

- Graphic space has not been replicated. Indeed, this was frequently constrained by the space the writer had available on the page, which is different from the space available here. Trying to conform to the graphic space of a manuscript would not show the constraints encountered by the writer and, moreover, could make reading the text unpleasant. In any event, users can consult the images of the original documents if they wish to look at their graphic space.

In this respect, we have aligned and centred some mathematical formulas. The indentation of paragraphs – often highly variable – has not been replicated, instead these have been separated by line breaks. The location of salutations (“Monsieur”, “Mon cher et illustre ami”, etc.) as well as of places and dates at the beginning of letters has been standardised in keeping with eighteenth-century French usage. We have similarly followed this usage to enter letter endings, signatures, and places and dates at the end of letters. In keeping with this same practice, which happens to be similar to that of today, the titles of some documents have been centred, while addresses in letters have been arranged in several distinct lines which mainly correspond to the following elements: salutation (e.g. “À Monsieur”), name, address, and placename. Finally, the principle we have applied in transcribing crossings-out has been “diachronic” rather than “diplomatic”. In other words, we have tried to be faithful to the chronology of the writing, trying to distinguish its different stages, rather than stick to its graphic space.

Editorial interventions

Most of the time the texts are not annotated. Nevertheless, some occasional clarifications have been provided in notes, as well as references to variants found in other text versions, descriptions of the various persons who interacted with the text in writing, details of misspelled surnames or characteristics of signatures found in certain letters,[3]

or lacunae resulting from the deterioration of the paper or omissions by the writer.

The rest of the editorial interventions appear in the body of the transcripts as follows:

- Crossings-out are placed between "<" and ">" : "<erasure>".

- Additions are framed by vertical bars: “|addition|”.

- Square brackets (“[ ]”) are mainly used in the following cases:

. Foliation. It precedes the contents of each leaf. Rectos and versos are abbreviated “r” and “v” respectively and written in bold type: “[1 r]”, “[1 v]”, etc.

. Pagination (most often for printed text). It precedes the contents of each page: “[1]”, “[2]”, etc.

. Uncertain transcription: “word [?]” means that the transcription of a word is possible but not certain. On the other hand, “first word [second word ?]” marks a hesitation in transcription between two words.

. Transcription not provided: “[... ?]” means that it has not been possible to decipher a word or series of words.

. Other word corrections: “m[o]t”, “maux [= mot]”, “le mots [sic]”. This type of annotations only occurs when a misspelling could be attributed to the transcriber rather than the writer, a fortiori in a text with few or no mistakes.

[1] With the exception of signatures, which play an essential part in identifying a person since they can be compared with signatures found in other manuscripts such as parish registers or notarial acts.

[2] But not works. Let us assume, for example, that the author of a book entitled “Concerning Integral Calculus” writes “my text concerning integral calculus” – in this case it is not possible to know whether they are referring to the subject of their text (integral calculus) or the title of that text (“Concerning Integral Calculus”).

[3] In particular, information on the nature of accompanying flourishes.

Last update on Monday 8 January 2024 (10:25) by Nicolas Rieucau