Author Topic: "Old" (<= 1976) IPA conventions?  (Read 1009 times)

Offline Don

  • New Linguist
  • *
  • Posts: 3
"Old" (<= 1976) IPA conventions?
« on: May 16, 2016, 09:37:16 PM »
Hi,

First post, not a linguist -> excuse any blunders!

I'm evaluating (non-dictionary-based) text-to-phoneme algorithms for USEnglish.  One of the reference documents I'm reviewing makes use of symbology that appears outdated and/or potentially unique (page in question attached).  I'm trying to clarify its intent and ensure its consistency with other related documents (different authors).

I've lots of questions but hope to start out simple :)

The "Rules" column is intended to indicate the ASCII representation used in the document for the sound indicated in the "IPA" column.  The "up arrow" is represented by a circumflex (^) -- I've no idea why that was not used in place of the up-arrow in this table!

Note the IPA symbols associated with the ^L, ^R, YL and YR rule symbols (lower left).  I am guessing that these are intended to represent syllabic versions of those particular sounds (note the schwa definition for 'Y')?  But, two different syllabic versions of each??

I note that a "^M" symbol is also present in the rules -- but omitted in this table (typographical error?)  This might be a syllabic M?

Another document that links with this references an "en" sound -- syllabic N?

On a finer point, are "YL" and "YR" to be treated as distinct from the "Y L" and "Y R" sequences?  I.e., is "schwa-l" subtly different from "syllabic l" (if we decide that the "YL" symbol actually references a syllabic l)?  Ditto for "schwa-r"?

Now, the tougher ones...  :(

Why is "I mapped to two IPA symbols?  Are those intended to be similar sounding? I.e., too close to warrant a different representation in the "rules"?

And, the real puzzler:  what might the intent of the YR,^ rule symbol be?  Note that a "bare" circumflex has no presence in the rules -- unless this is intended to mean "YR followed by any rule symbol that begins with ^".  Note, also, that "YR" already exists in this table!

Thanks for your patience and any clarity you can provide!
--don

Offline Daniel

  • Administrator
  • Experienced Linguist
  • *****
  • Posts: 1491
  • Country: us
    • English
Re: "Old" (<= 1976) IPA conventions?
« Reply #1 on: May 17, 2016, 09:32:45 AM »
I hope one of the sections on this page might answer your questions:
https://en.wikipedia.org/wiki/History_of_the_International_Phonetic_Alphabet#1976_and_1979_revisions

The problem for me is that I'm not sure about the font / handwriting used in that chart.

For "I, it might be the case that it is two different symbols to represent the same sound: an unrounded high mid vowel, modern /ɨ/.

The symbols ž, č, ǰ are not IPA as far as I know. I'm guessing they represent "zh" (voiced postalevolar fricative ʒ), "ch" (voiceless postalveolar affricate tʃ) and "j" (voiced postalveolar affricate dʒ) respectively. That's fairly common usage in a number of traditional (but not IPA) systems. I don't really know if they were mixed into IPA conventions for a while at that time or not. I know they were often used in descriptive grammars of languages (and various other publications). So this means that at least for those and potentially for others the "IPA" symbols are not strictly from the IPA, although overall they are similar to it.
Welcome to Linguist Forum! If you have any questions, please ask.

Offline Don

  • New Linguist
  • *
  • Posts: 3
Re: "Old" (<= 1976) IPA conventions?
« Reply #2 on: May 17, 2016, 01:07:16 PM »
I hope one of the sections on this page might answer your questions:
https://en.wikipedia.org/wiki/History_of_the_International_Phonetic_Alphabet#1976_and_1979_revisions

Yes, thanks -- I'd already been there (and disappointed).  The document is dated 1976 so I am not sure what "reference" was used for the "IPA" column in that table -- hence my continued search!  :(

Quote
The problem for me is that I'm not sure about the font / handwriting used in that chart.

Well, it's (worse than) "Greek" to me!  :(  My best guesstimate is attached -- but I'm not even sure of that as my opinion on these ^L, ^R, ^M symbols throws some of it up in the air...

Quote
For "I, it might be the case that it is two different symbols to represent the same sound: an unrounded high mid vowel, modern /ɨ/.

That's what I was guessing -- the "comma operator" to mean "or".

Unfortunately, the same interpretation for "comma" doesn't seem to apply to the YR,^ symbol!

Quote
The symbols ž, č, ǰ are not IPA as far as I know. I'm guessing they represent "zh" (voiced postalevolar fricative ʒ), "ch" (voiceless postalveolar affricate tʃ) and "j" (voiced postalveolar affricate dʒ) respectively. That's fairly common usage in a number of traditional (but not IPA) systems. I don't really know if they were mixed into IPA conventions for a while at that time or not. I know they were often used in descriptive grammars of languages (and various other publications). So this means that at least for those and potentially for others the "IPA" symbols are not strictly from the IPA, although overall they are similar to it.

Yes, I had reasoned that as well and reinforced that reasoning by examining some of the rules that create those sounds (though I could well be looking for examples that fit my expectations and ignoring those that contradict them!

There are lots of "little things" that aren't quite right in the document. And, hard for me to accept that they wouldn't have been caught prior to publication.

Chasing down the original author 40 years later doesn't seem like a promising avenue, either.

At the end of the day, I may just "take a best guess", implement the rules per that guess and
"score" the algorithm against some lexicon.  Then, try another guess and see if the score improves or worsens.  Thankfully, that sort of thing can be automated.  It's just annoying to have to resort to it at all!

Thanks!
--don

Offline Daniel

  • Administrator
  • Experienced Linguist
  • *****
  • Posts: 1491
  • Country: us
    • English
Re: "Old" (<= 1976) IPA conventions?
« Reply #3 on: May 17, 2016, 09:12:12 PM »
Someone familiar with the general descriptive tradition from that time would be able to help you guess much better than me. An older professor in a relevant subfield would be a good place to start. See where the book was written as a starting point. Beyond that your guesses look educated to me.
Welcome to Linguist Forum! If you have any questions, please ask.

Offline panini

  • Linguist
  • ***
  • Posts: 67
Re: "Old" (<= 1976) IPA conventions?
« Reply #4 on: May 18, 2016, 03:02:21 PM »
If you gave information about the source, that would be helpful. This does not represent any standard transcription system of any era, though it could be an ad hoc transcription system reducing phonetic symbols to sequences of regular letters, along the lines of SAMPA. It isn't one of the major respelling systems used in dictionaries. One is tempted to say that it wasn't devised by a linguist since it is so wrong about what IPA is.

Offline Don

  • New Linguist
  • *
  • Posts: 3
Re: "Old" (<= 1976) IPA conventions?
« Reply #5 on: May 19, 2016, 04:52:50 PM »
Someone familiar with the general descriptive tradition from that time would be able to help you guess much better than me. An older professor in a relevant subfield would be a good place to start. See where the book was written as a starting point. Beyond that your guesses look educated to me.

Thanks, I think I may have found an address for the original author.  I've drafted a note that highlights these issues -- along with other errors in the publication -- accompanied by copies of the pages from the original document (and related documents).  I don't hold out much hope that a 40-year-old memory will yield much authoritative information, though. :(  But, for the price of postage...

I suspect my best approach is to treat it just as a puzzle: assume ANY particular interpretation; run a reasonably large corpus through it and examine how those rules manifest in the resulting pronuncations; consult an authoritative reference for the "correct" pronunciations and correlate these with the empirical results; hope for some consistency!  :-/

At the very least, this will uncover other errors that I can't pick out "from casual inspection"!

Thanks, again!
--don