[gothic-l] Gothic Language Corner 11

Thu Feb 12 23:57:55 UTC 2015

As the question concerning the pronunciation of the digraph <ai> has been raised latterly, I give here a more detailed discussion, and include the digraph <au> too, as it is related to the issue. The following is based on the discussion found in Braune / Heidermanns (2004 sections 20-26).

 In older scholarship, the view that <ai> represented the diphthong /ai/, and <au> the diphthong /au/, specifically in those contexts where a diphthong existed in the Proto-language (e.g. *stainaz > Go stains, ON steinn, OHG stein, OE stan), is found almost without exception, although already Gabelentz / Loebe (1843-1846) questioned this shaky assumption. This old view was based mainly on the following arguments:

 1) The corresponding sound in the other Germanic languages appears partly as a diphthong, eg. OHG /ei/. This, however, is a non-argument, since such forms tell us only about the proto-form and nothing about Wulfilian Gothic. Such an argue is tantamount to claiming that the 'k' in Modern English 'knee' -- imagine that the pronunciation was unknown -- should be pronounced because the 'k' is in fact pronounced in all the other Germanic languages, e.g. German 'Knie,' etc., the which of course is false.

 2) The word HAILAG in the Pietroassa runic inscription is clearly a diphthong, since runic writing was by and large phonetic. But it is questionable whether the inscription is in fact West Gothic (it was found in present-day Rumania), nor is the dating certain. Even if the inscription is indeed Gothic and dates to the fourth century or even early fifth century, again this tells us nothing about the Gothic spoken in Moesia (present-day Bulgaria) by the Gothi Minores, that is, the Goths of Wulfila, since the diphthong is HAILAG could very well be an instance of dialectal difference.

 3) Some Latinized Gothic personal names show <ai> rather than <e>. Latinized versions of Germanic names, however, are notorious for being inaccurate and traditional. The Latin form of Gothic 'gut-' is consistently written 'goth-', with an inorganic 'th' and the wrong vowel. Similarly, the thematic connecting vowel in many Germanic ethnonyms, such as Langobardi (= *Langabardoz) have been apparently modelled on Gaulish and Greek names, wherein a connecting theme -o- was common, and do not reflect actual Germanic pronunciation. A further example is the common Latin spelling 'eu' for Gothic 'iu' in such names as 'Theudoricus' = *Thiudareiks. And by the end of the fourth century, spellings with <e> or <i> become in fact the norm where earlier an <ai> could appear, that is very close to the time of the Wulfilian translation. (For a discussion of Latin and Greek distortion of Germanic names, see the introduction to Schoenfeld's >Woerterbuch der altgermanischen Personen- und Voelkernamen<.)

 4 Gothic loanwords into Occitan and Ibero-Romance typically substitute the the value represented by <ai> with /a/: these languages did not contain in their phonemic inventory /ai/ but did have an /e/, and so, it is argued, the value of <ai> must have been /ai/, otherwise /e/ would have the valued adopted instead. Yet it is been pointed out that there are very few such words, and these could equally be the result of borrowings from a dialect still with /ai/, and again tells us nothing about specifically Wulfilian Gothic.

 There is in effect no real convincing argument for a diphthongal pronunciation of Wulfilian <ai> and <au>. Speaking strongly in favour of a monothongal value for these digraphs is the fact that Wulfilian Gothic is written in a script modelled on the Greek alphabet and clearly follows Greek orthographic conventions in a number of cases (e.g. <gg> = /ng/). In the Greek alphabet of the age in question, the diagraph <ai> represented a monothong, to wit an open /e/. Moreover, the diagraph <ai> is used in places in Gothic where historically there was never an /ai/ in Germanic (e.g. wair < *wiraz; haihait < *hehait; aith-thau <*eth-). The same digraph is used in adopted Latin and Greek words to represent an e-sound, not a diphthong (e.g. naubaimbair < Latin november). In fact, Wulfilian Gothic uses a different digraph to capture instances where /ai/ and /au/ still existed in the language, to wit, <aj> and <aw>: cf. Pawlus = Latin Paulus; (and from a later source, the Deed of Naples, kawtsjo = Latin cautio); tawida (/au/) verus taujan (/o/); and bajoths (/ai/) versus bai (/e/). Thus, if the word for 'stone' contained a diphthong rather than a monothong, the expected spelling would then be *stajns, and not stains.

 The survival of the old view amongst amateurs of Gothic is seemingly owing partly to the reliance on a few older works in English, such Bennet's introduction, which still tow the old line, and failure to consult more current academic research, which has seriously undermined the foundations of the old argument.

 Of course, one is free to pronounce Gothic in any ol' way one wants, but in that case, such Gothic is well on its way to becoming a kind Elvish or Klingon.

 Edmund
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/gothic-l/attachments/20150212/9bae99d8/attachment.htm>