Romanizing ɑ

[gestaltist]

I like making phonologies that contrast /a/ and /ɑ/, and I have yet to find a nice way of romanizing that. So I thought I'd ask: how do you guys romanize your languages that have that contrast?

[clawgrip]

The closest I have is /æ/ vs. /ɑ/, and I Romanize them as <ë> and <a>.

[Creyeditor]

I often use <å> for /ɑ/, <a> for /a/ and <ä> for /æ/ following some German dialectologists.

[clawgrip]

I should point out that the language I mentioned has vowel harmony and only uses umlauts, so I have:

/i y e o/ <i ü e o>
and
/ɨ ʊ æ ɑ/ <ï u ë a>

So the use of <ë> helps it match up with <e> in the other vowel set (as with i/ï and u/ü).

[Sumelic]

I think most of the strategies for representing /æ/ also work for /a/. The realization of these phonemes in different languages will probably overlap anyway. However, it depends a lot on the distribution of the phoneme and the rest of the phoneme inventory, specifically the other vowels.

/a ɑ/ <a o> or <a å> works in languages that don't have other distinct /o/-like vowels, or where /a/ phonologically patterns as central and /ɑ/ phonologically patterns like /ᴐ/.

/a ɑ/ <e a> works in languages where the value of /a/ is closer to [æ], and where /ɑ/ phonologically patterns as central and /a/ phonologically patterns like /ɛ/.

/a ɑ/ <æ a> is fairly simple, and approximately what was used in Old English. It treats them fairly evenly, so it can fit well even in an inventory without many other parallel contrasts, such as /a ɑ e i o u aː ɑː eː iː oː uː/ <æ a e i o u ǣ ā ē ī ū>

/a ɑ/ <ä a> is also fairly simple. It goes well with languages that use other umlauted vowels; however, it can also be used as the only umlauted vowel, as in Skolt Sami.

/a ɑ/ <á a> or <aa a> makes sense for languages such as standard Dutch and Hungarian where /a/ is treated as phonologically "longer" than /ɑ/. It's also used in Northern Sami.

/a ɑ/ <a â> is used in French; I think it works well in languages where /ɑ/ is treated as phonologically "longer" than /a/~/æ/, such as French, English, and Persian. It might look a bit unbalanced if you don't use the circumflex on any other vowels, though.

[opipik]

/a ɑ/ <a ạ>

/a ɑ/ <a a̱>

/a ɑ/ <a ao>

/a ɑ/ <a a̧>

/a ɑ/ <a ą>

/a ɑ/ <a ā>

/a ɑ/ <a á>

/a ɑ/ <a à>

/a ɑ/ <a ȧ>

[RobElks]

In Shesti: http://relkton.blogspot.co.uk/2015/11/the-shesti.html, I romanised /a/ as "a", and /ɑ/ as "ó", partly because in some dialects, it is pronounced as /ɒ/.

[Lao Kou]

In the very inelegant system I use for my own notes on Cantonese (in notebooks, and usually written in cursive), I use the cursive form of <ɑ> for [a], and actually write out a physical <a> for what the Wikipedia article on Cantonese describes as [ɐ]. Sidney Lau's system uses <aai> and <aau> vs. <ai> and <au> for the contrast (don't remember what Yale does, but I think it's similar), whereas the mainland dictionary I have uses <ai> and <ao> (Mandarin pinyin values) vs. <ei> and <eo> (which I personally find counter-intuitive). Frankly, I'm personally hard-pressed most of the time to distinguish 街 (gaai/gai) ("street") and 雞 (gai/gei) ("chicken") or 教 (gaau/gao) ("teach") and 够 (gau/geo) ("enough"), and back in the day, nothing would have pleased me more than to just collapse them in my personal pinyin and call it a day. But it remains a sticking point in my attempt at a personal cross-dialect pinyin that works for the dialects I dabble in, so I keep it in, both for in case I can one day consistently hear the diff in Cantonese and because it helps predict cross-dialect readings of characters I know in Mandarin.

So for now, it's 街 (cursive <gɑi>) vs. 雞 (mangled cursive <gai>) and 教 (cursive <gɑo>) vs. 够 (mangled cursive <gao>)

Diacritics are a hassle if you have tone markers (though I use <ö> and <ü>, go figger. Or is it because I love 'em? mwa ha ha ha ha (to wit, a Portuguese/IPA tilde for nasalized vowels of Taiwanese plus tone markers is a Vietnameseque nightmare - but what to do?). I shudder to think what that would entail in actually typing my notes. But it's for personal use, so who cares?

To your point, though, why not just use the <á a> option proposed by sumelic or the <a á> by opipik and use the diacritic on the one (seemingly) less frequent?

[alynnidalar]

Azen uses <a o> for /a ɑ/, but that's mostly because /ɑ/ originated from /o/ anyway.

[Avo]

Persian-style <a ā> or <a â>.

[Squall]

This is the same problem as romanizing the contrast between /e/ and /ɛ/ and between /o/ and /ɔ/.

We can present better ideas if you present your entire vowel inventory.

An example I have:
a â /æ ɒ:/
e ê /ɛ e:/
o ô /ɔ o:/
i î /ɪ i:/
u û /ʊ u:/


Other solutions for /a ɒ/:

Diacritic: <a â>
Silent consonant: <a ah>
Double vowel: <a aa>
Double consonant: <a aC> Ex: ap app; appap apap

[gestaltist]

Thanks for the input, everyone, a lot of good ideas.

To those of you asking for my full phonology: I was asking this as a general question. I was more interested in seeing how you guys solve the problem.

In my current project (Tsoketa - not yet published here), I pointedly got rid of ɑ (it surfaces due to a diachronic process but disappears later) to avoid having to think about that.

One another idea I have which nobody has mentioned: both could be romanized simply as <a>. We conlangers tend to want to have a clean grapheme-to-phoneme correspondence, but natlangs rarely do it. Next time I make a language with a/ɑ, I might try just that - and maybe only have some way of disambiguating for minimal pairs.

[Lao Kou]

Next time I make a language with a/ɑ, I might try just that - and maybe only have some way of disambiguating for minimal pairs.

Silent "e".

Spoiler:

SILENT "E" IS PEOPLE!!

[G64]

romanizing the contrast between /e/ and /ɛ/ and between /o/ and /ɔ/.

As an Italian, I always use è /ɛ/ and é /e/, ò /ɔ/ and ó /o/.
Funny thing is, even though ó is used in Italian when clarifying the pronounciation (for example in dictionaries), it doesn't appear on keyboards and I have to copypaste it

Because even though it has minimal pairs ([bottɛ] barrel vs. [bɔttɛ] hits), they are considered allophones and spelt the same way (<botte>)

[Omzinesý]

In one of my unnamed langs, I used <ia> for the low front vowel and <a> for the back vowel. Similarly <e> was used for schwa and <ie> for /e/. The front vowels cause palatalization of the preceding consonant so the <i> works similarly to Polish. In that language the vowels, however, are phonemic and palatalization secondary. Actually I copied the orthography from somebody's conlang.

[Curlyjimsam]

<æ> for /a/ and <a> for /ɑ/ would work, I think, at the risk of being slightly confusing for people familiar with the IPA. A language with both phonemes would probably have quite a fronted /a/ anyway.

One another idea I have which nobody has mentioned: both could be romanized simply as <a>. We conlangers tend to want to have a clean grapheme-to-phoneme correspondence, but natlangs rarely do it. Next time I make a language with a/ɑ, I might try just that - and maybe only have some way of disambiguating for minimal pairs.

That would be a nice thing to do too - and why bother even distinguishing minimal pairs, necessarily? Presumably you'll still want a way of notating the difference in your notes, however.

[jute]

I use or have used <a> for /a/ + /ɐ/ and <o> for /ɑ/ in Jutean

[xroox]

Deinau uses <ä> for the front one and <a> for the back one.