On Troubadour Manuscripts and Spelling

In Marcabru's Bel m'es quan la rana we find the words < lonja/monja/vergonha/messonja> rhymed with each other.

And in a poem of Raimon de Miravalh we find the following lines:

Senes vergonha,

Car non y a una tan conja
Non pogues estar ses vergonja

The form of the poem suggests that <vergonha, conja, vergonja> are to be taken as rhymes. It is extremely unlikely that the two tokens <vergonha> and <vergonja> were meant to be pronounced differently from one another, or that something less than a full rhyme was intended.

It seems worthwhile to use a regularized orthography for the Old Occitan of Troubadour poetry. The one I have devised is based primarily on the nòrma classica used by most writers of Modern Occitan today. It is in essence an adaptation of the nòrma classica to Lyric Occitan, on par with previous adaptations to Lengadocian, Provençal and other modern dialects. It has the effect of making Occitan look a good deal more like Catalan. 

In doing this, I part ways with the general consensus of scholars and editors of troubadour texts. 

Generally troubadour lyrics are printed in a way that more or less reflects the orthography of the later medieval manuscripts the texts are found in. Even recensions relying heavily on half a dozen or more witnesses tend to keep each emendation to the base text in the precise form found in the source manuscript. William Paden advises editors of troubadour texts:

More fundamental than metrical structure, but more delicate, is the language itself. Old Occitan was far from standardized, and the troubadour koiné accepted alternative forms in many graphic, phonological and morphological functions. Do not impose a standardized language upon your helpless poet. You must accept variations produced by the logic of the language. 
In a footnote to this paragraph, Paden expresses his (and many others') unfavorable view of previous attempts at systematized or regularized orthography for modern representation of Old Occitan lyric texts. Paden makes an error common among educated persons in modern literate societies, and a few more errors common among scholars of Old Occitan. 

The first error is to identify language with its written form. This is an easy slip for scholars studying dead languages where writing is all that we have to work with. This error is father to the assumption that changing the orthography of a language is in effect changing the language itself, and therefore respecting the attested orthography is part and parcel of respecting the text and doing right by the "hapless poet." There are extreme cases, such as Classical Chinese, where it is actually true that changing orthography is tantamount to changing the language in an important way. There are other cases (such as Ottoman Turkish, Urdu, Persian or Yiddish) where orthography itself may express information relevant to the aesthetic appreciation of a text which would not necessarily be evident to a foreign reader or even listener. But these are cases where orthography is much more "standardized" than 12th century Occitan was. Writing songs down at all seems to have been highly unusual in 12th century Occitania. Bernart de Ventadorn mentions, at the end of one song, that he is committing his work to writing because he cannot send a messenger to relay it to his beloved orally. Were it not for this, we would have no evidence that songs of this kind were written down at all in Bernart's lifetime.

A spoken language remains largely the same no matter what combination of symbols you use to represent it, just as a person is largely the same no matter what angle you photograph them from. 

The advice commonly given to singers (e.g. by Robert Taylor in Singing Early Music) is to use different pronunciations for the different manuscript spellings (e.g. <olh, uolh, uelh>). Such advice is sometimes justified by the plausible suggestion that the variants might have coexisted or singers may chosen one or another intentionally for aesthetic purposes. In recent years much hay has been made of this "diversity" present in troubadour manuscripts. But the precise details of the orthography of the chansonniers may say very little about the pronunciation of troubadour song in the 12th century. 

It is true that competing forms may alternate even in the natural speech of a single person (e.g. everyone vs. everybody.) This may be doubly true of a single singer using a performance dialect. There is even reason to believe that aesthetically motivated alternation existed for at least some features in Lyric Occitan. 

It is one thing to keep all of this in mind. It is quite another matter to imply that the troubadour texts we have, separated from the original composition of the songs by at least two human lifetimes during which they survived largely in oral transmission, could be faithful records of such diversity in any but the greatest of generalities. The very suggestion strains all credulity. 

In the specific case of <olh, uolh, uelh, uel> when such spellings co-occur in a single manuscript, the same general target sound in the performance dialect is being represented with a variety of different regional spellings. The precise realization of that target sound will have been different for different troubadours, and had different levels of consistency depending on the performer's temperament with regard to the performance dialect. But the spellings themselves do not necessarily imply anything about whether one or the other pronunciation was preferred by a given singer (on any given occasion.) And even in early and highly localizable non-lyric compositions such as the Boecis, spellings are inconsistent on this and other points. 

Medieval Occitan scribes not only did not have much use for strict orthographic consistency in vernacular writing, but they also lacked any concept of correct and incorrect spelling. 
What mattered was that what they wrote be readable and that the reader not be led to mistake a given form for something else unintended. This entails only that the orthography use symbol combinations intelligible to the reader in terms of established patterns. Like so:

R father hoo art in heaven. Hallode bee thy neym. Thy kingdum kum, thy wil be dun, on urth az it iz in heven. Give us this deigh r deighlee bred & 4give us r trespasses az wee 4give those who trespass agenst us. And leed us not in2 temptation butt deliver uss frum evil. 

No comments:

Post a Comment