Principles of Analysis and Organization

ELECTRONIC VERSIONS OF TEXTS

• In order to ensure permanence and to make it possible to produce multiple kinds of presentation of the documents, the corpus files are in XML format, and they follow the Text Encoding Initiative’s P5 Guidelines.
• The printed editions of the documents, even when somewhat unreliable, are maintained in their original form. This allows the reader to judge, among other things, which texts need new editions.
• Where the text included indications of expanded abbreviations (e.g. in italics), these have been incorporated without comment.
• Cases where the editor has supplied missing words, has suggested an alternative reading, or has indicated uncertainty are noted as changes in the texts. Not all of the editors’ emendations have been accepted, however. For example, a manuscript’s “autreiam” in a context where the third person plural is required is left as is, since it may very well be an indication of the collapse of the first person and third person plural, which is characteristic of a number of Gascon dialects.
• Changes in spacing and punctuation, particularly with abbreviations and numbers, have been made without comment.
• A few minor textual corrections have been made, primarily when a manuscript reading seems wrong for very simple reasons (“u” read as “n” for example). All such corrections are clearly indicated in the text.
• Words hyphenated at the end of a line in the print edition appear on the following line here.
• French-language headings, titles, and comments added by the editors have been omitted, but line numbering remains as in the original.
• Those passages that consist of two or more words in a language other than Gascon appear in blue in the texts. Single words in Latin are not always tagged, since they are frequently learned spellings for Gascon words. In some cases, the mix of Latin, Gascon, and Latinized Gascon made these choices difficult and somewhat arbitrary.
• A bracketed ellipsis ([...]) is inserted where the editor omitted passages (generally because they were illegible or consisted of repetitious formulas). Each such gap is explained with a pop-up “reason” attribute, although in many cases the reason is simply “unknown”.

DETAILS ON DOCUMENTS

Each text is preceded by four lines of information that allow the reader to place it historically and linguistically.

• The first line includes a reference name and number (specific to this corpus, but wherever possible shared with the Dictionnaire onomasiologique de l’ancien gascon) followed by an indication of the status of the text and the number of words contained in it. The text’s status may be:
  • “original” (the first version of a text or a copy made within the same context for one of the participants in the textual act)
  • “close” (the text appears in a document copied within 60 years of the original)
  • “copy” (the text appears in a document produced more than 60 years after the original).
• The place of origin of the original text, followed by a region label that allows sorting on the basis of a simple geographical division of the domain, and the date of the document. Place of origin is usually the town or city in which the text was written (which may or may not determine the language of the text). In some cases it is an institution like the Court of Bearn. The dates are those given by the text’s editor.
• The names and titles of those responsible for the physical document (notaries, scribes), if these are indicated within the text.
• Information on the manuscript in which the text is found: its location, its classification within whatever collection it resides, and its date.

VERB ANALYSIS

Except for texts that are very poorly edited or are clearly not in Gascon, the corpus provides an analysis for each verb form.

• Lemmas are assigned based on one of the dominant medieval spellings (see list). A hyphen following the lemma (“aver-” , e.g.) indicates that the verb serves as the auxiliary in a compound form (perfect, passive) or is the head in a causative construction.
• For the verbs “esser” and “estar”, forms that do not allow for a clear decision between the two (e.g., estat) are assigned to “esser” unless the context suggests otherwise.
• Tense/mood/aspect are indicated by the following abbreviations, which follow the Leipzig Glossing Rules. Exceptions are noted.
  • PRS.IND = Present Indicative
  • PRS.SBJV = Present Subjunctive
  • IMPF = Imperfect Indicative (for Leipzig PST.IPFV)
  • PST.SBJV = Imperfect Subjunctive
  • PRET = Preterite (for Leipzig PST.PFV)
  • FUT = Future
  • COND = Conditional
  • FUT.PST = the reflex of the Latin pluperfect, whatever its function. This form corresponds to the second conditional in Old Occitan and is generally a perfective “future in the past” in Gascon today.
  • INF = Infinitive
  • PRS.PTCP = Present Participle/Gerund (classed together): see below
  • PST.PTCP = Past Participle
  • FUT.PTCP = Future Participle; originating in the derivational suffix -ORIU, -ORIA, these forms tend over time to resemble inflectional forms of the verb in Gascon.
  • IMP = Imperative
• Person and number are indicated for tensed forms with the following labels: 1SG, 2SG, 3SG, 1PL, 2PL, 3PL.
• Gender and number agreement of participles: M.SG, M.PL, F.SG, F.PL
• Early Gascon shows some evidence of case distinctions like those of Central Occitan; these are relatively regular in documents from the Bordelais region, but they seem to be learned additions and are used inconsistently elsewhere. Based on the syntactic context a form like “deitz” (PST.PTCP of dizer ‘to say’) may be tagged “M.SG”.
• Gerunds and present participles present a bewildering array of agreement possibilities; they are labeled here based on their functions in the sentence. Where it is clear that the form does not show agreement, its gender/number tag is “0”. Note: all forms that are traditionally labeled “gerunds” today end in -s in the North, a pattern that is clearly emerging in the texts of our corpus.
• Any given verb analysis may be subject to doubt; the mark “?” indicates some level of uncertainty, while “??” suggests considerable uncertainty.