Recording Practices
There are some difficulties which have needed to be addressed in presenting this language data. These include differences between the written transcriptions and the corresponding audio-tapes, and the difficulty of determining word-boundaries.
Differences between the Transcriptions and the Tapes
Flint’s written transcriptions and the corresponding audio-tapes do not always present identical information. There is sometimes a slight discrepancy between an informant’s or Flint’s pronunciation of a word on the tape, and the corresponding written transcription. For example, Flint transcribes the Yanyula word for "ankle" as nandadangk, but clearly pronounces the word with word-final a on the tape (i.e. nandadangka). In another instance, he transcribes the Yanyula word for "wild fig" as magalyan (transliterated makalyan), but says mangalyan on the recording.
With discrepancies such as these, either the transcription or the tape must be chosen as the basis for our transcription in the \ft field. Whether to use the transcription or the tape as the primary source has been decided on a language by language basis, and is set out under "Method" on each language homepage.
Word-divisions
Determining word-divisions in the Aboriginal words and phrases is sometimes difficult. It is often hard to tell, from the word-spacing, whether Flint has transcribed an utterance as one (polymorphic) word or two. It is also difficult to determine word-boundaries from Flint’s and the informants’ pronunciation on the tape. Referring to the analyses of other linguists is very useful when the transcription and tape are unclear. However, we do not want to lose significant information unique to Flint’s material.
Sometimes constructions which are clearly grammatically parallel appear to be transcribed in different ways - sometimes as a single word and sometimes as two. For example, in Flint’s original transcription the Garawa verbs "to spit" and "to sweat" are ja"bulaba "ngayu and kindibangayu respectively. In the first of these, the first person singular pronominal element ngayu appears to be separate from the verb, whereas in the second, it appears to be connected to the verb.
Thus determining word-divisions for the "flint’s transcription" (\ft) field involves referring to Flint’s transcription, the tape and other sources. Once again, the role played by each of these is decided on a language by language basis and is set out under "Method" on each language homepage.