The stated goals of the Kirshenbaum transcription are that:

  • It should be usable for both phonemic and narrow phonetic transcription.
  • It should be possible to represent all symbols and diacritics in the IPA.
  • The previous guideline notwithstanding, it is expected that (as in the past) most use will be in transcribing English, so where tradeoffs are necessary, decisions should be made in favor of ease of representation of phonemes which are common in English.
  • The representation should be readable.
  • It should be possible to mechanically translate from the representation to a character set which includes IPA. The reverse would also be nice.

Given the third, fourth, and fifth of the goals, why did Kirshenbaum choose to invert the IPA’s decision to make ‹r› the symbol for the alveolar trill and require ‘fancy characters’ for the other rhotic consonants?

Most IPA phonemic transcription schemes for English sensibly use the more familiar symbol ‹r› for English’s rhotic consonant. If Kirshenbaum transcriptions are to be mechanically transliterated to IPA, then the long-winded ‹r<trl>› form must be used for it to come out correctly.

Furthermore, the IPA’s choice to use ‹r› to represent the alveolar trill is nicely international, since many languages will at least sometimes use it even if it is not the dominant realization of any phoneme (just as in English). Among European languages [ɹ] is unique to English as the ‘normal’ realization. This is a jarring contrast to Kirshenbaum’s general approach, which (despite the third goal) is highly international, arguably more so than IPA, due to its featural nature.

This choice of symbol for the alveolar approximant seems to be excessively focussed on one specific use case: the narrow phonetic transcription of English. For broader purposes it exists in direct conflict with the goals of Kirshenbaum.

Is there any other reason for Kirshenbaum to prefer it this way around?

(Furthermore I noticed while writing this that Appendix D of the specification says that ‹r<trl>› is alveolar while Appendix E incorrectly gives the IPA symbol for the uvular trill as its equivalent.)

  • Same reason broad IPA transcription uses <r> for the English rhotic. It's easier to type and easier to read. – Draconis Sep 14 '17 at 22:47
  • @Draconis Yet as I pointed out, when machine transliterated to IPA it becomes the harder-to-read ‹ɹ› – dpk Sep 15 '17 at 6:52
  • Yes, but that's not Kirshenbaum's fault. If he defined <r> as equivalent to IPA /r/ then that would make things more difficult for English. See goal 3. – Draconis Sep 15 '17 at 16:09

Because of the existence of labial trills, "b-trl-" must also be accounted for. Suppose that "r" is taken to represent a trill, then would a labial trill be "b"? His system somewhat encourages use of phonetic modifiers, akin to the practice of APA, where IPA would use separate letters. If we agree that using "b-reg-" in virtually all instances of a bilabial voiced stop would be absurd, then either the trilling of the labial and lingual trills must be written differently, or preservation of IPA interpretation cannot be dispositive.

The third goal is sufficient to motivate this decision: all things being equal, a system whereby English requires no fancy letters or sequences is preferred. You don't have to represent the flap in English (one-letter * is available if you want to do so), and having to use 5-letter "r-trl-" would probably render the proposal unacceptable for most computational users. I do not see any conflict in his principles, since he doesn't claim that he will maintain IPA values as much as possible. All IPA values can be represented (at least with some augmentation), and although it is not a principle, they are indeed as much as possible the same as IPA. Consistent with his principles, the alveolar trill could be represented as F (which doesn't not appear to be used), which suggests an unspoken principle "pick bas characters that most resemble the IPA character" (though that is violated by the choice "r-lbd-" over "v-lbd-".

IPA "r" theoretically represents a trill, except that most of the time when there is no trill / flap contrast and the rhotic is [ɾ], people simply write "r". Likewise, "a" represents a wide range of phonetic values; it is only when there is a contrast that one must resort to "a ɑ" or "a æ". Typically, if a language has 5 vowels and the mid vowels are actually [ɛ ɔ], people don't write "ɛ ɔ", they use regular letters. The vowel written as "ə", when there is just one non-front unrounded vowel, often is realized as any of [ə ʌ ɨ ɜ ɘ ɤ]. When there is a contrast, or when it is necessary to communicate narrow phonetic value, other letters will be pressed into service.

| improve this answer | |
  • 1
    “a system whereby English requires no fancy letters or sequences is preferred” — But most phonemic IPA schemes use IPA’s ‹r› for English’s [ɹ]. – dpk Sep 15 '17 at 6:54
  • I don’t understand the argument in your first paragraph either, since ‹b› is a stop, not an approximant. You could write [ɹ] as ‹r<apr>› if regular ‹r› were the trilled form. – dpk Sep 15 '17 at 6:56
  • 1
    @DavidP.Kendal While many people use IPA <r> for the English rhotic, this isn't strictly correct (as that letter represents an alveolar trill, which isn't a realization of the phoneme in any dialect of English that I know of). The sound in English is generally an approximant. – Draconis Sep 15 '17 at 16:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.