WIELD’s Americanist transcription system descends from a rich and varied tradition of linguistic work in the Americas going back to the turn of the twentieth century. WIELD recommends the use of a standardized Americanist system for all broad transcriptions, especially in work relating to the languages of the Americas. We recommend that the International Phonetic Alphabet be reserved for narrow transcriptions. The full range of consonants, diacritics, and vowel symbols can be viewed here.
Our recommended Americanist system strikes a balance between idealized symmetry and practical usage with regard to individual symbol choices. Classic Americanist transcription practices such as unitary symbols for certain affricates, a preference for unambiguous diacritics over the creation of new letters, and diacritics which reverse the backness of vowels (e.g. /u/ for IPA [u] but /ü/ for IPA [y]) have been interwoven with common symbols (often shared between the IPA and the Americanist tradition) which do not conform to expected diacritic usage. Thus for a voiceless uvular fricative, we recommend the symbol /x̣/, which is formed by the regular application of the backing underdot diacritic to /x/, but we recommend /q/ be used for IPA [q] rather than /ḳ/; /q/ is the commonest symbol in recent work within both Americanist and IPA traditions.
There is still room for variation in transcription practices, but any differences from the standard proposed here should be due to legitimate language-specific needs. For example, Ventureño Chumash has coronal affricates, which, if represented by the symbols recommended by WIELD, might be written as /c/ and /č/. However, in Ventureño, such affricates regularly alternate with sibilant fricatives. For instance, the third person singular appears as both [ts-] and [s-]; the writing of these as /c-/ and /s-/ obscures the relationship between the two varieties of the prefix. Affricates such as [tʃ] can also sit astride a morpheme boundary, with the [t] being a separate morpheme from the immediately adjacent fricative. Using the single symbol /č/ would obscure those boundaries. Thus it makes more sense to transcribe coronal affricates in Ventureño with two-consonant combinations: /ts/ and /tš/. Similarly, the sequence [ts] in the English word CATS would be transcribed as /kæts/ rather than /kæc/ due to the [s] in CATS being segmentable as a separate morpheme from the root CAT and the lack of a /c/ as an affricate phoneme in English.
Some endangered languages have a strong orthographic tradition. In no cases should such a tradition trump the usage of the transcription system put forth by WIELD. Instead, it is recommended that an additional line be added to examples in order to give the broad transcription in Americanist symbols. (See the sample text in the document linked above for a more detailed discussion.)
WIELD recommends fidelity to the speech sounds of a language over any considerations of parsimonious phonemic representation. Broad transcriptions need not be slavishly bound to phonemic theory. If a phoneme has two or more allophones which are quite different to an English speaker’s ears, researchers are encouraged to consider representing such variation in their broad transcriptions of data. And English bias should, in fact, play a role in the decision to elevate at least some allophonic variation into the broad transcription. English is the most widespread language of scholarship. It is also the first language of virtually all Indigenous peoples north of Mexico. An English-speaking audience will be aided by overtly representing sounds which are not familiar to English readers.
This is a pragmatic policy of scientific description. Researchers should not look for ways to reduce a transcription system as though they were reducing a fraction; rather, they should use the appropriate number of symbols necessary to convey the pronunciation of the language without getting lost in too much detail. This policy is not the giving of carte blanche to avoid detailed phonological analysis, and it should not be used as an excuse to exoticize data.
All work sponsored or published by WIELD is expected to adhere to the recommended transcription system set out here. Any broad transcription practices which deviate from this standard should be justified thoroughly and might require editorial or board approval before being accepted.