Quote:
Led thoil, a Chaoimhín, cur chughainn anso é, do bheinn-se fíoghar-bhuidheach díot
Here's what I have. Now that I look at it again, I think a big chunk of it actually came from a site I found online which may no longer be available (which is why I always download cool things like this), and what I did was to add into it some explanations from posts on the old forum, which fit in nicely:
Causes of Initial Mutations in Irish
The main cause is the Irish way of speaking: the words of a sentence are pulled together and spoken without a pause. Pauses are almost always avoided, and some sentences seem as if spoken as a single word (there seems to be a quasi “horror vacui” = a fear of nothingness). The individual words affect each other in pronunciation (what is generally known as “Sandhi”), which also occurs in French as “liaison” (e.g. les amis [“lezami”]), in order to enable this unhindered flow of speech. By contrast, in German there is a short pause made after almost every word, so the beginnings and ends of the words remain separate and unchanged. It’s important to note that those words in German beginning in a vowel actually don’t begin with a vowel sound, but with a glottal stop, which is not the case in Irish. As is also the case in French, certain unspoken phonemes “reappear” under certain circumstances (e.g. the normally mute “s” in les). However, as opposed to French, in Irish these letters are for the most par no longer included in the spelling. They make themselves known via a “t-, h-, n-prefix”, via “eclipsis”, or via similar phenomena. Also different from French is the fact that these changes are made to the beginnings of words and are used to express grammatical relationships to other words.
Séimhiú [Lenition]
The h after a consonant (previously, and occasionally also currently, written as the consonant with a dot over it) is to indicate that the consonant is lenited, i.e., pronounced ‘softer’ than it would otherwise be. These mutations eased the flow of speech, because by using them the obstacles presented by the stop sounds disappeared. Stops cause a stop in the flow of air, and in turn hinder the flow of speech. Lenition originally occurred in Early Irish if the consonant ended up between two vowels. If a word ended in a vowel and the next word began in a consonant + vowel (which was mostly the case), that consonant was now between two vowels and was lenited. When that happened, the consonant would become weaker than in other positions. So, a “t” would become the sound represented in English as ‹th› (as in ‘thing’, not as in ‘this’), a “p” would become an “f” (just like ‹ph› in English represents the sound f), and a “b” would become a sound that’s written phonetically as [β], which can be loosely described as being halfway between a “w” and a “v” (as pronounced in English). As in Spanish, this was originally an automatic rule, so there was no reason to represent it in writing. The word beir was pronounced /ber/, since the b was at the beginning of a word, but if you added a prefix to it that ended in a consonant, you’d end up with, for example, do-beir, in which the b was suddenly between two vowels and the word would be pronounced /do-βer/, while still being written the same way.
Today the once final vowels have been mostly discarded, but lenition became in time a tool of grammar, being used where once a vowel ending once stood. For example, the feminine article was originally longer and ended in the vowel “-a” (*sinda). In time, it was shortened to an, but the following lenition was retained. This explains lenition of feminine nouns after the article. The masculine article (*sindos) was *sindi in the genitive singular and the “i” ending in that case thus caused lenition of the initial consonant in the following word. The masculine possessive pronoun “a” is an old genitive singular of the personal pronoun. Just like the genitive singular of the masculine articles, lenition is also required afterwards. Feminine nouns once ended in the nominative preferably in the vowel “-a” (as is the case in Latin), and this is why the following adjectives and nouns in the genitive are still lenited today, although the lenition today is generalized and also occurs following those feminine nouns which didn’t once end in “-a”).
As the language evolved and time passed by, many of the vowels that had otherwise surrounded these consonants were lost, and the reason that the consonant was ‘softened’ (lenited) was no longer apparent. Basically, sounds like /th/, /ph/, and /β/ stopped being just automatic substitutions for other sounds in some contexts and became separate sounds of their own. Since they were now separate sounds, they could also contrast with the ‘original’ sounds (t, p, and b here), which means that a way was needed to distinguish them in writing. Therefore, they came up with the simple idea of adding a little dot over the lenited letters.
Since the advent of typewriters and computers, which tend to be based on the English alphabet and its limitations, this adding of dots over consonants became bothersome. Glyphs like ṗ ṫ ċ ḃ ḋ ġ ḟ ṡ ṁ do exist for computers nowadays, but they’re not available in many computer fonts, and on old-fashioned typewriters they were impossible without special keyboard layouts. Therefore, it was decided that these dotted forms would instead be written as the consonant followed by an “h”.
The only lenited consonant that is always silent is f (a lenited /fh/ is always silent, no matter where in the word it is). It’s still written, because lenition now occurs as a grammatical process, and you’d otherwise end up with pairs such as fear ‘man’ vs. m’ear (“my man”). The latter is written m’fhear so you can tell the root word is fear, not ear.
Most of the other digraphs are not generally silent, though some of them are silent in certain positions in the word, varying by dialect. Most commonly, /dh/ and /gh/ have both melted together to become a sort of guttural soft ‘gurgling’ sound, written in phonetic writing as [ɣ] or [ɰ] (depending on the dialect), but this sound is very often lost in the end of words, sometimes after affecting the previous vowel. Similarly, /sh/ and /th/ have both developed into being pronounced like a regular h, and if that appears at the end of a word or before something like an n, there’s a very natural tendency for this sound to disappear as well – just try saying a word like ‘meh’ and actually pronounce the h at the end there and see how much unnecessary effort it takes.
Both “f” and “ph” were pronounced bilabially in Irish (using only the lips, not the tongue). The “ph” combination arose, as a softening of “p”, but “f” arose as a strengthening of the original [w] (word-initially and in certain other positions). That this is the case, and that it must have gone through an intermediate stage of [ɣw] or [xw] or something like that, can be seen from the fact that the same words in Welsh now start not with an [f], but with a [gw]. Take the word ‘man’ as an example:
-- Latin vir retains the original [w]
-- Old English wer (still found in ‘werewolf’, a ‘man-wolf’) [the “i” became “e” because it’s followed by an “r”, which is regular in Germanic]
-- Common Celtic *wiros (with the typical masculine ending -os)
-- Insular Celtic *ɣwer(os) (“i” is lowered to “e” because of the low vowel “o” in the following syllable -- -os is subsequently lost)
-- Goidelic *xʷer ➔ Old Irish fer ➔ Modern Irish fear (the “a” is inserted to indicate that the “r” is broad, although “e” before a broad consonant is subsequently lowered to be pronounced more like an “a” anyway)
-- Brythonic *ɣwer ➔ Old Welsh gwər (probably) ➔ Modern Welsh gwr -- the “e” is lost subsequently and the [w] glide is then turned into a full vowel [u], as is regular in Welsh.
Ùrú [Eclipsis]
Eclipsis and the n-prefix occur where today or once the preceding word ended in a nasal (m, n, or ng, but mostly n). For example, the endings of the numbers seacht (seven), naoi (nine) and deich (ten) ended earlier in a nasal (compare Latin septem, novem, decem). This “-m” ending has long since disappeared; but it lives on in the modern eclipsis, as in seacht mbád (seven boats). Eclipsis also occurs after ocht (eight), by analogy to the numbers 7, 9, and 10, although the number 8 never ended in a nasal (eclipsis has spread by analogy in other ways as well). Similarly, the preposition “i” was once, as in German and English, written as “in”. Since it ended in a nasal, it caused eclipsis. However, the eclipsis also made the -n ending in in superfluous, so that the preposition is today whittled down to just i (except before a vowel, when it resurfaces, as it does with the “n-prefix” before eclipsed vowels). Another example is the conjunction go (also today a rare preposition in the sense of “with”), which causes eclipsis. As with the cognate related Latin word cum, the conjunction/preposition go (Old Irish co) once ended once in a nasal, and thus caused eclipsis, which go still does today.
Eclipsis was produced by these final nasals as a way of simplifying speech. A voiceless consonant is difficult to speak after “-m” or “-n”, so it was simply re-voiced. The sound “p” became “b”, “t” became “d”, and “c” (=“k”) became “g”, because “-mb”- and “-nd”- are easier to speak than “-mp-” and “-nt-”. In time, many of the end nasals disappeared, but the initial mutation remained and was then generalized as a grammatical tool (e.g. also stretched on other words, e.g. ocht (eight), as in ocht mbád. The sequence of a nasal followed by an unvoiced stop (that is, a d, t, or c) or an f developed into the voiced counterpart of the stop. So, nc became g; mp became b; and nt became d. This took place inside words (as is common in languages), so the word idir (“between”) is cognate to Latin inter-, and cúig(e) (“five”) is cognate to Latin quinque, which is easy to see once you know that d is the Irish outcome of nt, and g is the outcome of nc (or, as here, of nqu[e]).
However, unlike the situation in most languages where such changes happen, in the Celtic languages these changes also happened and stuck across word boundaries. So if you had a word that originally ended in n or m, and then the next word begins in a p, t, or c, then the change occurs across the two words as well. Now, the word ár was at that point something like *áron, and croí begins with a c. So (ignoring the spaces between the words, and the fact that croí is not how that word would have looked two thousand years ago), *áron croí became *áro groí. This concept of the first consonant in a word being voiced like that because of a previous word is what is known as eclipsis. At some point, *áron lost its final syllable and became just ár, but the effect it had on the following word still lingered on. As a result, people continued to say the g instead of a c. They couldn’t quite figure out how to best represent that in writing for many hundred years, but nowadays it’s consistently written so that you first write the consonant which is actually spoken (g), and then the consonant which the word begins with if it’s not next to a word that triggers eclipsis. So, when you write ár gcroí, the c is silent. It’s just there to let you know that the basic word actually begins with a c, which has now been eclipsed by a g.
The definite article an, conversely, never ended in a nasal consonant, so no such change ever took place there: an croí just remains an croí. However, after the combination of a preposition and an article, either eclipsis or lenition can occur, depending on the dialect. Why? The original accusative singular form of the articles were *sindon (masculine) and *sindan (feminine). Both ended with an “-n” nasal, and this is why today eclipsis follows, as in the expression ar an mbord (on the table). Just as in German, there was the possibility of the accusative or the dative after certain prepositions, depending on the desired meaning. In German, for example, they can be used to distinguish motion or position, as in auf dem Tisch (position on the table) and auf den Tisch (motion onto the table). The dative form of the article was then *sindu (masculine) or *sinda (feminine), which of course caused lenition.
This is why in Ulster today lenition (rather than eclipsis) is used following the combination of a preposition and an article. Today, there is no lexical differentiation between dative and accusative following a preposition. Most prepositions are now followed by the dative (or what remains of the dative). If lenition or eclipsis is used, it is now more a question of dialect, which is why there is in Connacht/Munster ar an mbord (on(to) the table), compared with the Ulster ar an bhord.
In the genitive plural, the article once was also *sindan, from which eclipsis resulted, due to the “-n” nasal ending, as discussed above. The plural possessive pronouns “ár”, “bhur”, and “a” are old genitive plural forms of the personal pronoun, and probably ended in a nasal as well, so they still cause eclipsis today.
The rule for eclipsis is that unvoiced stops become voiced and voiced consonants become nasals:
Unvoiced stops
p -> bp
t -> dt
c -> gc
f -> bhf
Voiced stops
b -> mb
d -> nd
g -> ng
Here's how changes developed over time, using two examples (I can't reproduce the chart cleanly here, so I'm using dashes):
Masculine forms:
Indo-European/Old Irish -- Translation -- What’s going on
sind-os fer-os -- the man -- two consonants, so nothing happens, but the initial s is lost at some point in history
ind-os fer-os -- the man -- now the endings are lost
ind fer -- the man -- now the -d is lost
in fer -- the man -- change the spelling and the sounds a bit
an fear -- the man -- Irish
am fear -- the man -- Gaelic assimilates the an to am
Feminine forms
Indo-European/Old Irish -- Translation -- What’s going on
sind-a ben-a -- the woman -- consonant-vowel-consonant, so lenition (and again the initial s is lost)
ind-a bhen-a -- the woman -- now the endings are lost
ind bhen -- the woman -- now the d- is lost
in bhen -- the woman -- change the spelling and the sounds a bit
an bhean -- the woman -- Irish
a’ bhean -- the wife -- Gaelic now loses the –n
Login
Register
FAQ
Search
