I agree that it may be a good idea to have a conventional transliteration of Ethiopian languages for non-linguists.
Thus, I think that ch for č, sh for š, j for ǧ, etc. are good choices. On the other hand, I don't find diacritics so disturbing as you do; The rule of thumb should be, in my opinion, to use diacritics as long as ignoring them won't make a far away pronunciation (such as s for š).
So, I think that ä for the modern languages is preferable to e, while sticking to ā for the 4th order vowel. This may seem redundant, but it's better than e which may be misunderstood for the 5th or 6th order vowels. In Ge'ez, Wolf Leslau chose to transcribe the 1st order vowel with a, so as to stay indifferent to the effect of laryngeals; e.g. kahāli vs. hallawa (in fact, ä:a shows the contrast of the vowel's timbre, whereas a:ā shows the contrast of length; usually the two contrasts go together, but not next to a laryngeal). This may have been good also for Amharic, but a seems too far away from the actual pronunciation, whereas Ge'ez is an extinct language whose original pronunciation cannot be known for sure.
Another thing is that I believe ñ is a widely known character, so I don't think that ny is necessary.
Now, transcribing the 6th order vowel with ě and the 5th with ē is more than enough to make the distinction and can be recognized by most people. In fact, the macron on ē is not necessary. Using i for the 6th order vowel seems to me too far away and even ridiculous sometimes (consider: wiha for ውሃ). e is actually commonly used for the 6th order vowel when spelling Ge'ez and Amharic words in Englsih: e.g. negus, selassie, Gebre Krestos (I would suggest: něgus, Sěllasē, Gäbrä Krěstos or: Sěllāsē, Gabra Krěstos à la Ge'ez).
One last thing, why should the gemination be neglected? The only reason I can see is with digraphs such as sh where both shsh and ssh are strange; but otherwise I think it gives important information on the pronunciation and shouldn't disturb the non-linguist. Consider the spellings Muhammad, Allah, hajj, assalamu alaikum, ummah, etc. in Arabic; if it's good for Arabic why isn't it good for the Ethiopian languages? yhever13:04, 30 January 2006 (UTC)[reply]
Thanks for your comments (since I hadn't announced this anywhere, I'm surprised anybody found it). I think I'm being driven (guided?) here by what is easy for non-linguists, especially native speakers of the languages, and I think ignoring diacritics would be the norm for them; they're just harder to type (and may not even be in the character set the person is using). As I'm sure you know (well, I don't know much about you so I'm guessing here), some distinctions that are contrastive in languages are routinely ignored in orthography, in transcription used in non-linguistic contexts, and even in transcriptions used in linguistic articles (examples of the last are tone in Chinese and Bantu and pitch accent in Japanese).
As for "ä", I must confess I never liked this, for linguistic articles or otherwise, because it doesn't suggest anything like the correct pronunciation (in languages that I know of where it's used, German and Finnish, it's something else altogether). Maybe that's a good thing; I don't know. At least people won't have any biases. For a non-linguistic transcription, Leslau's reasons for choosing "a" (which I didn't know, BTW; I know very little about Ge'ez in fact) aren't really relevant, it seems to me. My reason for preferring "e" is just that, I think, without a diacritic, it suggests the correct pronunciation to a completely ignorant reader. That's all. But here's an argument in your favor: in Oromo, the vowel that's closest to the first order Amharic vowel is written "a"; for example, Wallagga ወለጋ, Shaashamannee ሻሸመኔ.
As for "i", I don't quite agree on how inappropriate it is. Yes, if people who know nothing of the language are treating the "i" in, say, wiha as a Spanish or French "i", they will be pretty far off. But lots of English "i"s, especially unstressed ones in American dialects, are close to the 6th order Ethiopian Semitic vowel. But you're right, as I said in the page, this sound is often "conventionally" represented with "e". (In this case, Oromo is on my side though: Xilaahuun ጥላሁን, Dirree Dhaawaa ድሬ ዳዋ.)
I agree with you on ñ; this is familiar from Spanish (and much preferable to the Italian/French "gn", which I've seen in many places misspelled as "ng", leading to the completely wrong pronunciation).
As for gemination, my reason for proposing that we ignore it is simple. Unlike Arabic speakers, who learn to mark gemination when they learn to write Arabic, Ethiopians and Eritreans learning to write Amharic and Tigrinya do not learn this (at least this used to be true), and, in my experience, it's not something people are aware of (though they do learn about the concept of መላላት and መጥበቅ in school). There are also lots of cases that would be hard to agree on (like final geminated consonants; is the final /l/ in ይገባል geminated?). I think it would be hard to maintain the standards; that's all. That said, I admit I really don't like the looks of adis or welega/wäläga. One question we need to think about is whether (as for the librarians) we are transcribing Amharic/Tigrinya orthography, in which case there's no reason to mark gemination, or phonology, in which case there in.
Anyway I don't want to give the impression that I'm strongly committed to the proposals I made. My main interest is in agreeing on one system and then sticking with it. -- MikeGasser(talk)05:27, 1 February 2006 (UTC)[reply]
