User:Lingo Bingo Dingo/Dutch/UCLA Phonetics Lab Archive errors

From Wiktionary, the free dictionary
Jump to navigation Jump to search

UCLA Phonetics Lab Archive is an archive containing phonetic transcriptions for a variety of languages. This page is intended to provide examples why the UCLA Phonetics Lab Archive should not be used as a reference for Dutch pronunciation.

I have omitted transcriptions that I find questionable but seem to have been used as a semi-conventional notation by the transcribers, such as syllabic sonorants or codal glottal stops. Errors/inconsistencies in the use of [v] versus [ʋ] are also omitted, because these could easily be OCR errors. Likewise errors or inconsistencies regarding [ɔ] versus [o] are excluded, because it is not always clear to me whether these are simple errors or the result of a convention that conflates /oː/ and the marginal phone written as /o/ or /ʊ/ (the latter has completely merged with /ɔ/ in my lect; needless to say if this conflation were a convention it would be a very dubious one; this list does appear to distinguish them). Finally, I have not included weird shit that probably reflects a regional or dialectal pronunciation; although including these in pronunciation sections without adequate labels would obviously be bad.

The various word lists have notational differences; which is understandable if done by different people at different times, but not ideal if we intend to use it as a reference. More is that in some lists the notational conventions seem to change midway.

Many of the errors are specific to some word lists. This one does not consistently distinguish the phonemes /y/ and /u/. The same list contains dozens of errors with aspirated voiceless plosives; I decided not to include that many.

I think these examples are enough to show that these lists should only be used by people with sufficient knowledge of Dutch phonology. They should not be used in the way a pronunciation dictionary could be used for reference.

The transcriptions from the archive are apparently supposed to be narrow.

Egregious errors[edit]

Innocuous errors or questionable conventions[edit]

  • <fair> given as [fɛr] (doesn't capture the phonemic distinction between /ɛ/ and the more marginal /ɛː/) [25]
  • <katje> given as [kɑtʃjə] [26]
  • <beuk> given as [beɷt] replace ɷ with ʊ, invalid IPA characters (ɷ) (probably a reproduction error) [27]
  • <uw> given as [ɪʉ] (likewise for <duw>) [28]
  • <oude> given as [aʊt] (transcription is for the uninflected form) [29]
  • <boet> given as [buːt] [30]
  • <bedankje> given as [bədɑŋkλə] replace λ with ʎ, invalid IPA characters (λ) [31]
  • <lijfarts> given as [liːfɑʀts] (an affected Low Saxon pronunciation or a transcription error?)[32]
  • <kar> given as [kaːʀ̥] (same for karren, dialectal?) [33]