User talk:Lhynard/Projects/Say What

Phonetic Convention
This is an awesome project, especially for a non-native English speaker like myself. I'd like to contribute to it in a more systematic way, but then I've noticed that there are some entries that don't use the same conventions as others, or the ones established in the Dragon article.

For example, the sound /æ/ is denoted as '@' in the Dragon article, but is instead expressed in the corresponding IPA notation throughout the wiki. That is not the case with the vowel sounds /oʊ/ and /ɔ/, represented in the article as 'o' and '0', respectively, but displayed here with no distinction. The sounds /ə/ and /ʌ/ are indistinctly represented as 'u' both in the article and on the wiki, etc.

Moreover, there is no apparent convention in the representation when the phonetic description comes from different sources. For example, p. 4 of Tomb of Annihilation has a phonetic guide to pronounce the names of many NPCs in the adventure and monster names throughout the book, but uses a different convention when displaying them than that of the Dragon article. And even within the wiki there are examples of different notation, such as stress syllables emphasized with boldface or all-caps (I don't recall seeing the standard IPA symbol /ˈ / anywhere).

I'm not advocating for any particular convention (although my preference does lean towards something closer to IPA), but maybe it would be beneficial if we had something of a policy when adding pronunciations. Any thoughts? &mdash; Sirwhiteout (talk) 18:58, May 1, 2018 (UTC)


 * Thanks!


 * Yes, you raise points that I was already considering. My plan was to just get any pronunciation onto the pages from any source as a first priority and then come up with some sort of wiki-wide convention after that for a round two.


 * The problem, as you noted, is that the sourcebooks do not follow any sort of convention. They are not even consistent. Just look at how many ways we are told how to pronounce Eilistraee. It's ridiculous. :)


 * Like you, I would prefer something closer to IPA. (I was the one who decided to use ash (æ) instead of the at symbol (@), simply because I could not bear using an at symbol like that! :) ) A problem, though, is that most people probably have no idea what the IPA symbols mean. I love linguistics and phonetics, but I don't even remember what ʌ is, though I would much rather use schwa (ə) than u for the so-called "uh" sound so prevalent in English.


 * Anyhow, as I said, in the future, I wanted to come back and attack the problem, perhaps even coming up with some sort of in-line template to use across the wiki for pronunciations that would link to a key. I was quite happy that&mdash;once again&mdash;you seemed interested in working on one of my more pedantic projects. :)


 * ~ Lhynard (talk) 19:20, May 1, 2018 (UTC)


 * For example, here are the two Wikipedia templates that I was considering implementing here in some form:
 * wikipedia:Template:IPAc-en
 * wikipedia:Template:Respell


 * I really like the IPAc-en template, especially if the alt-text explaining each individual phoneme works here too. However, the second (Respell) appears to be more easily (perhaps trivially) converted from the notation most sources use. Can we implement both? My vote would be to use IPAc-en as the primary one. In case an ambiguity appears or it is not exactly clear how to translate, we could use the second to approximate the original as it appears in the source, perhaps even keeping it as a note.


 * (Curiously, in the IPAc-en template documentation, one of the aliases for schwa is @.)


 * I'm looking forward to working on this one. For some reason, particularly pedantic projects appeal to me. :) &mdash; Sirwhiteout (talk) 04:34, May 5, 2018 (UTC)


 * OK, here's what I have so far:


 * gives.


 * gives.


 * gives.


 * (Once these come out of my sandbox, it will shorten to  of course.)


 * I wanted to get the audio and formatting down first. Next, I'm going to see if I can get it to generate the IPA automatically from the "phonetic spellings" given in all the sourcebooks. This will make it much easier on editors. Wikipedia uses LUA, a coding language; I'm going to avoid that, but I'm going to need to make a large look-up table, which I'll generate with a Python script ahead of time off-line. Ultimately, the plan will be for  to give ( pronounced: hlɪˈnɑːrd hli-NARD ), with links over the IPA symbols to what they mean.


 * Thoughts?


 * ~ Lhynard (talk) 03:34, May 11, 2018 (UTC)


 * I love it! I was considering writing something like that myself, but real-world obligations pretty much precluded learning the basics of wiki template development, much less LUA (you youngsters and your modern programming languages… why not C?). In any case, this looks amazing. I particularly like changing the pronunciation keys to in-universe words, like /ɑː/: a as in "bard". That added a pretty cool touch.


 * I'd say that, as far as the respelling component is concerned, this template is good to go.


 * As for future development, I have a few suggestions/questions about the IPA conversion:


 * Enclosing the IPA respelling in slashes, so it looks like ( pronounced: /hlɪˈ nɑːrd/ hli-NARD )
 * Is there a way for the tooltip not to be a link? Or is this on purpose?
 * I was wondering for a moment about how you'd make the automatic conversion to IPA from the respelling key, but I guess it would just be a (simple, but tedious) matter of including all syllables listed in the references. It does risk coming into something new from time to time, though, and I don't know how it would handle primary and secondary stresses. For example, our favorite nightmare word, Eilistraee, has one pronunciation that includes a primary and a secondary stress: /ˌiːlɪsˈ tɹeɪjiː/. How would the automatic conversion handle that?


 * &mdash; Sirwhiteout (talk) 17:27, May 11, 2018 (UTC)


 * Thanks for the feedback.


 * I'm hardly a youngster! :)


 * (I actually know nothing about LUA and have no desire to learn that one. My languages are C++ and Python.)


 * Yes, I can (and should) certainly enclose the IPA in slashes.


 * I'll have to look into the tooltips not being links. My thought was to hack it with links and to make every link a redirect to the correct spot on a single page with a table of all the IPA sounds with examples. This way, one would not know that they were all separate links; it would just seem like a single link to the IPA table. I'll have to play around with that.

Yes, stress is something I need to think about more. My first idea is to solve it with an extra, optional  field, though.
 * Yes, I was going for the "simple, but tedious" method; that's why I plan to have a script generate the switching template for me. The trickier part will be coding up the single-character sounds separately from the two-character sounds.


 * ~ Lhynard (talk) 17:10, May 14, 2018 (UTC)


 * I think the current version of the template is sufficiently complete and elegant to be implemented throughout the wiki. The IPA implementation can wait, in my opinion, since it involves a considerable amount of work (and is non-essential in a first approximation). The template documentation could include a note about using the Respell guidelines to make sure that transcriptions are uniform across the pages.


 * Any more thoughts? Is there anything you feel should be adjusted before deploying the template? &mdash; Sirwhiteout (talk) 22:10, May 15, 2018 (UTC)