From a linguistic perspective, poetry is deeply conservative. Older forms are retained in poetry long after they have passed into obsolescence, and even used for fresh compositions. We never say 'tis, but iambic meter forces many of us to rely on it. In some cases entire systems of sound change are undone for recitation, as when a Londoner rhotacizes the lyrics to a song, or a Parisian pronounces the otherwise-silent final e to make a line scan.
Something like that awareness of sound change happened in Medieval Germany. In this case, however, its effect was to internationalize the language across dialect boundaries, rather than to merely retain archaisms. From Waterman's History of the German Language (pp. 94-95):
Whatever its provenience and models, there did emerge for a period of perhaps fifty years a more or less "normalized" Middle High German literary language that enjoyed high prestige and wide currency among the poets. Its normalizing tendencies are most obvious in its rhymes, for the court poets of the Golden Age did not tolerate the easygoing assonance of an earlier time, insisting rather on pure rhymes. And to a remarkable degree they were able to use words, the final syllables of which would correspond when transposed into any of the major High German dialects.Posted by Ben Brumfield at October 17, 2006 06:08 AMAn example of this is provided by the works of a poet who stands at the very beginning of the Blützeit, Heinrich von Veldeke, a native o fthe province of Limburg in the Netherlands, whose Eneit ( = Aeneid, begun around 1170 but not complted until approximately 1190) determined the style, both in form and ocntent, that was to characterized the great court epics for the next fifty years. Though he probably wrote in his native dialect of Limburg (we have no manuscript in his own hand), slightly modified in favor of the neighboring Middle Dialects to the south and east, his choice of rhymes is such that, when converted into High German, the correspondences still hold. For example, he rhymes tît (Zeit) with wît (weit), but not with its homophone wît, which in High German yields weiss, incapable of rhyming wiht either Zeit or weit. In like manner he rhymes lîden (leiden) and snîden schneiden), but not lîden and rîden, for again — though they are a pure rhyme in Low German — the High German forms leiden and reiten would not so qualify.
OK, wot's "rhotacize" as applied to London English?
Posted by: Martha Bridegam at October 18, 2006 02:21 PMPronouncing the letter R in syllable-final position.
See the Wikipedia article on rhotic/non-rhotic accents for a nifty map.
Posted by: Ben Brumfield at October 19, 2006 07:21 PMAh. So when people in Massachusetts pahk their cahs in Hahvahd Yahd they are being non-rhotacized?
But what about the people there who call me "Mahther"? That's a displaced "r," not a missing one.
Posted by: Martha Bridegam at October 20, 2006 12:16 PMExactly.
The R in your "Mahther" case isn't so much displaced as added. Non-rhotic accents apparently fall into three groups -- ones that merely elide postvocalic R, ones that also possess "linking r" between word boundaries, and ones that also add an R to the end of all vowel-final words.
My own cousin "Mahtha" lives in a non-rhotic region of Virginia, which does not contain intrusive R. I tend to associate "Mahther" with New England.
Posted by: Ben Brumfield at October 21, 2006 07:57 PM