-1

Consider the Wikipedia article for phoneme, this is in Norwegian but one can easily translate, I will use this example for asking the question.

Fonemer er vanligvis plassert mellom skråstreker i transkripsjon, mens språklyder (foner) er plassert mellom hakeparenteser. Ordet takk vil altså transkriberes fonemisk /tak/ som en sekvens av tre fonemer /t/, /a/, /k/, mens det fonetisk transkriberes [tʰak:]. Forskjellen ligger her i aspirasjonen av "t", [tʰ], som ikke vil utgjøre en meningsforskjell om den ikke aspireres ([tak]).

I am a bit confused. /tak/ is said to be the phonemic transcription of takk, and to be made of a sequence of three phonemes, /t/, /a/, /k/. The phonetic transcription is [tʰak:]. Why isn't the first phoneme /tʰ/ if it's pronounced (if I understand correctly, and from the phonetics) with [tʰ].

More generally, what is transcribed by the phonetic and phonemic transcription respectively?

Nardog
  • 4,931
  • 1
  • 16
  • 36
kiriloff
  • 137
  • 5

1 Answers1

0

The "why" question at least has the unsatisfactory answer "because that's how it is traditionally done". However, writer-inconvenience is a real factor in deciding how to write phonemic transcriptions. Simply put, it is easier to write /tak/ than to write /tʰak:/.

There is a real difference in claims made when you write /tak/ than to write /tʰak:/, and I might add, /tɑk/ or /tʰɑk:/. Since the (presumed – I suppose for the Oslo dialect) phonetic fact is the pronunciation [tʰɑk:], any divergence from the observable should be justified, assuming that you are using /these letters/ and [these letters] to mean the same thing. Then the question is, which analysis gives the best account of the rules of the language? /tʰ/ implies one set of rules, /t/ implies a different set of rules, which is the better set of rules?

Implicit is this kind of analysis is reductionism, that one should discover the analysis that has the fewest phoneme segments, which is why one does not posit /tʰɑk:/ and /stɛ:k/, one posits /tak:/ and /stɛk/ (or, that is one of two options).

In the case of vowel versus consonant length in Norwegian, the details indicate (contrary to the traditional view) that vowel length is "more contrastive". There is a contrast [V:C] vs [VC:] as in [tʰɑ:k] 'roof', [tʰɑk:] 'thank' owing to a general rules. However, the possessive clitic -s does not cause shortening of a vowel when added to a VC-final noun, compare Tors "Tor's" with a long vowel and kurs "course". Vowel length is not actually fully predictable from the phonetic record, it is phonemic, even if it is predictable given non-phonetic factors.

We actually can't say that an author intends by their transcription – you have to ask the author what they were thinking and why. We also don't know what "phonetic" vs. "phonemic" means as used by a particular author, because these concepts don't have uniform definitions and applications in the field. Most often /tɑkk/ would mean "the input to the phonology" and [tʰakk] would mean "the output of the phonology", but sometimes people mean other things, so ask, whenever possible.

user6726
  • 83,066
  • 4
  • 63
  • 181