Theiling Online    Sitemap    Conlang Mailing List HQ   

Re: Irish Gaelic is evil!

From:Stephen Mulraney <ataltanie@...>
Date:Friday, February 18, 2005, 20:07
Carsten Becker wrote:
> Hey!
First: "Gaelic" or "Scots Gaelic" (pronounced ["galIk]) is the name of the Goidelic language of Scotland. In Ireland we call our Goidelic language "Irish", not "Gaelic" or "Irish Gaelic". The word "Gaelic" is still used (e.g. in the name of the Gaelic Athletic Association), but it's then pronounced ["gajlIk]).
> Since we're going to deal with Ireland the rest of this > semester IIUC, I had a look at Wikipedia for information > about the Irish orthography (or should I rather write > áeghtdhúaekreímfbhiagh?) and phonology. Yikes! They must be > crazy. Now I understand the concept behind Christophe's > Maggel and why it is thought to be a parody of (Old?) > Irish. Wikipedia writes something about velar offglides and > palatalization, but how is that realised? And how can I see > if a consonant is palatalized or velarized? The chart for > the orthography always gives e.g. c = [k], [k_j], [k_G] > without further explanation.
Well, the chart is incomplete. Actually, if you take into account the "context" column of the vowel chart, you can figure it out, but it helps if you already understand the system. There are two series of consonants in Irish, slender and broad. I'm deliberately avoiding phonetic descriptions such as "palatalized" and something else ("plain", "unpalatalised", "velarised") since these phonetic descriptions don't always describe it properly. But basically consonants occur in pairs, for example /t_d/ is paired with /c/ (or /tS/ in Ulster), but only one grapheme is used to represent each pair (in this case <t>). Incidently, /t_d/ is the broad member of the pair, and /c/ (or /tS/) the slender. Vowel *graphemes* are also divided into two classes, sometimes also called "broad" and "slender"; Basically, "broad" graphemes are those for back vowels (<a>, <o>, <u>), and "slender" are those for front vowels (<i>, <e>). When a consonant grapheme is next to a "broad" vowel grapheme, it represents the broad consonant sound of the pair; and when next to a "slender" vowel grapheme, it represents the slender consonant sound. As for what "next to" means, there's an orthographical rule of Irish spelling which says "caol le caol agus leathan le leathan", i.e. "slender with slender and broad with broad". The meaning is that a consonant grapheme (or a cluster of them) must be flanked by *either* "broad" vowel graphemes or "slender" vowel graphemes, so extra vowel graphemes are inserted to make it would (giving rise to the infamous triple & quadruple series of vowel symbols... :)). In practice there's little confusion. For example, a long vowel always has to be a "real" vowel, rather than an extra symbol. Some examples of the whole process: in the phonetic transcription I'll use /t/ for the broad value, and /t'/ for the slender value (and so on for other consonants), rather than getting into the complicated business of trying to explain right now which slender sounds are paired with which broad sounds. The word for "man" is /f'ar/. Obviously our spelling will start with something like <far>; but this doesn't indicate that the <f> represents a slender sound, so we insert an <e>: <fear>. But could be analysed as /f'er/, too? Well, I don't think that's a possible syllable; although /f'e:r/ is. In the case of /f'e:r/, we could begin to spell it with <fér>, and finalise it with the insertion of an <a> to keep the <r> broad: <féar>. The <é> has to be a "real" vowel sound, since you'd never insert a long vowel for purely orthographic reasons. As you can see, it's impossible to spell the impossible syllable /f'er/. This is because the orthography has developed in tandem with the language over a millenium and a half, and by now fits it very closely.
> IIRC, palatalized means that you pronounce a sound with the > tounge aiming for [j], but not touching the palate > actually. But velarized offglide? Is pronouncing a [G] very > shortly after another sound?
No. It's easiest if you forget all about this "velarisation" stuff. If you concentrate on differentiating the slender sounds, mostly palatalised, from the broad sounds, which you can think of as "plain", then you'll do better. Note that the dialects vary widely in their phonology, which brings up another advantage of the orthography: it's dialect neutral. In fact, most of my comments here, if we exclude the ones that deal with actual phonetic values, apply equally to Gaelic in Scotland too. But, the examples: To give you an idea of what the sounds are like, in other words to explain what the wiki pages means by those palatalised apostrophes, here are some phonetic values from one particular dialect (Cois Fharraige, in Connemara). grapheme - broad value - slender value <p> - [p] - [p_j] <b> - [b] - [b_j] <f> - [f] - [f_j] <m> - [m] - [m_j] <t> - [t_d] - [c] <d> - [d_d] - [J\] <c> - [k_-] - [k_j] <g> - [g_-] - [g_j] <s> - [s] - [S] <ph> = <f> <fh> = 0 <bh> - [w] - [v'] <mh> = <bh> <gh> - [G] - [j] <dh> = <gh> <th> between vowels: 0 <th> elsewhere: [h] - [h_j] <ch> between vowels: [x] - 0 <ch> elsewhere: [x] - [C] <h> only occurs at beginning of word: [h] A curiousity of this dialect is that the letters <l>, <r> & <n> each represent three sounds, not two. There's a broad emphatic, a slender emphatic & a slender lenited form (no broad lenited form exists). A single slender <l>, <n> or <r> not at the beginning of a word represents the slender lenited form; everywhere else, it's emphatic (in particular, a doubled slender <l>, <n> or <r> not at the beginning of a word is emphatic slender. Here's what these terms mean, though the emphatic versions feel quite "tense", and I'm not sure how to transcribe that. grapheme - broad emphatic - slender emphatic - slender lenited <n> - [n_e] - [n_e_j] - [n_j] <l> - [l_e] - [l_e_j] - [l_j] <r> - [r_e] - [r_e_j] - [r_j] Phew! I hope I've got everything. Except the vowels, that is. It's more than you asked about, but at least I can use it as a reference the next time the subject comes up... s. -- To be sure Stephen Mulraney to be sure ataltane@ataltane.net