Having read in full one of the threads where LPA are thrown about, and IPA too, it strikes me that both are pointless. The following is my logic:

IPA
Most people starting cannot hear separately the sounds. They also don't know why there would be such a thing as separate classes of l, n, r, s, w etc. IPA is complex to start with, with the extra diacritics used. In other words, there is no perceptive ability to hear in ones own speech or others, plus there is no understanding of the rationale of the system (several pieces of information at one.)

LPA
Irish and English are not pronounced the same, so explaining one in terms of the others is hardly credible. Using chains of relative instruction every single time with the result that no symbol has a particular referent, but a dithered set , makes it unusable. In a sense, it is the opposite of IPA. LPA has too much information, much of it low quality, and no superstructure behind to to hang to together. IPA is very succinct, but its usage is contingent on prior understanding.

How to make sounds
This nonsense is going on for a long time. Placed relatively to one another, the languages sounds have 4 possible relations:
1) exactly the same (slender ch and h in 'human')
2) same/almost same tongue position, different articulation (y in yes and strong slender Irish l)
3) similar tongue position (c and ch)
4) different (slender bilabials)

I have mapped out the transformations needed -they are simple and only take practice -less than some of the grammar.

The proposed solution
a) Person learning come to terms with Irish not been English.
b) Understand the basic utility in grammar of having 2 sets of consonants
c) Understand the system of the mechanics of Irish (b -->bh etc)
d) Understand that to communicate this with others, one needs to use some sort of a system. From the point of view of phonology, sounds can be corralled by taxonomy, that is, classes based on degree of difference or similarity, like animals in nature. Broad ba is in opposition to slender be, ar to er and so on
e) Split Irish sounds into 3 main class (normal/soft) (broad/slender) (lips, front, middle, back).
f) Use the modified IPA of Irish dialect studies (broad l is then /L/, trilled r /R/, tapped r /r/.
g) Learn why Irish orthography is the way it is. You do not need to be able to speak, just understand the logic. The reason people are memorizing Irish words like English, instead of making them on the fly is that the rules are misunderstood

Basically, one needs to understand the purpose of the difference prior to using any symbolic system for representing the sounds. Then one needs to learn a system to represent the distinctions and once known allows discourse to begin. Next one needs to learn to make the sounds. Only then, can one realistically write them, LPA, IPA, or modified-IPA

Practical solution for speech
The best thing would be like I did starting -make a plain and slender distinction. Only learn the slender sounds first, as they are more important in Irish than broad. Use the nearest sounds in English, but use the modified-IPA to tell other people about it. The result would be people on different pronunciations still communicating on the same wavelength about important distinctions. In other words, the modified-IPA is a transducer that lets use show the broad/slender, normal (b)/soft (bh), and position (lips, front, mid, back) distinctions.

So you're suggesting that people use this modified-IPA instead of LPA and IPA?

Or do you mean people should use modified-IPA only when starting out and then move on to the more complex pronunciation systems once they have mastered the differences?

Yes, in a way. The truth is the level of detail given by Lughaidh or myself with IPA is not needed as it is false anyway -sounds run into one another. PLus, I am giving a 'perfect north Connacht' rule based one, and Lughaidh is giving a traditional NW Donegal version. This is not to everyones taste (such as Munster speakers). Indeed, all you are getting with the [IPA] is a somewhat detailed transcription of static speech, which changes when you start to speak.

The old /Irish dialect book/ slashes give a less detailed transcription, BUT they show important differences, as in /L/ vs /l'/ (strong and weak slender l).

The utility comes from been able to use it before you can make all the sounds, as you come to find the logic of the sound system of Irish more sensible, and that way people can talk about the sound of a new word, and just get a workable version of it, without either been completely like English, or absolutely native. Of course, you still need to learn the sounds, but not all of them.

I understand the difficulties of learning it -I just happen to live near Ulster and between that and myself, a good chunk of Irish sounds exist in some shape of form. However, I found relative sentences harder, for contingent reasons, so tis swings and roundabouts.

As for learning the sounds, most English speakers have some sound that can be modified to make it Gaelic -the y in yes, for example, with a little modification lets you make a slender l, n, (historical strong) r, z, t, d, g, c, ng, s. That is apart from slender b, f, m, p, v/w almost all of the slender sounds. The broad sounds are harder, but I has simple instructions for them too

quote:

The old /Irish dialect book/ slashes give a less detailed transcription, BUT they show important differences

You are inventing bicycle. The thing you are looking for is called phonemic transcription as opposed to phonetic. Phoneme - is a smallest distinct part of word substitution of which changes the meaning of the word. Each phoneme can have several actual realisations depending on surrounding sounds. In that sense it is futile to give "gamma" index after broad consonants as Lughaidh does. The marked member of broad-slender pair is the slender one, so it is enough to mark the slender consonants, the other consonants are broad by default as there are only 2 different types of consonants.

Other point - if someone pronounces either thrill-r or flap-r - it is semantically insignificant. There are no Irish words that would be distinguished only by flap-r vs. thrill-r contrast . The actual realisation of [r] phoneme just influences person's accent, while not impeding the understanding. Therefore, it is also futile to show the precise sound which is realised as [r] phoneme's allophone.

The last point allows you to get rid of plenty of unfamiliar IPA sign. There is no point to clutter transcription with strange symbols if 95% of reader have not a clue what those signs mean, and even if they knew - no guarantee they could pronounce all those "uvular", "postalveolar fricatives" and other "goodies" correctly.

So phonemic transcription is the way forward. The precise phonetic transcription is too ephemeric, the same speaker can pronounce the same word differently while phonemically it is the same for sure.

(Message edited by Róman on July 02, 2007)

"The thing you are looking for is called phonemic transcription as opposed to phonetic."

I know the distinction, but I am interested in pedagogy here, not linguistics

"The marked member of broad-slender pair is the slender one, so it is enough to mark the slender consonants, the other consonants are broad by default as there are only 2 different types of consonants."

Indeed, thus goes my logic. With such a system, the sound system is clearer for learning

So phonemic transcription is the way forward

Given you explain everytime how every phoneme has to be realized !

For me, phonemic transcription is sort of step coding that one learns in conjunction with pronunciation. The relation between broad and slender is easier to see. It is not an IPA replacer

quote:

Given you explain everytime how every phoneme has to be realized !

Precise pronunciation of every allophone is pointless, as it neither enhances understanding nor is practically attainable. You can use the same allophone for every phoneme and be doing very fine in comparison with anyone saying "jia gwitch" and the likes. And again - the point was that 99% of discussants can't read all IPA signs anyway. So if you are pedantic and want to show all possible intricacies, your audience doesn't follow you. Then what is the sense of doing this?

(Message edited by Róman on July 04, 2007)

Do you really think that a learner knows how /t'/ or /r'/ etc have to be pronounced if you don't explain it anywhere?

Lads, maybe as a psychology graduate here I have a different perspective to linguists, but *my* point was that you TRAIN the sounds and LOCK IT IN with a symbol as referent, so that people associate a symbol with a sound. It helps to have a proprioceptive element too in that the learner can feel and hear themselves making the sounds. To enable this, a map must be made for English speakers to get to Irish sounds. A different map would be need for Francophones, Mongolians etc

With a broad phonemic transcription, you can use 1 symbol =1 sound. With IPA diacritics that modify the basic 'meaning', you mix basic incomprehensibility with a feeling that there is a whole lot of mysterious variability involved. IT turns people off.

Granted, people don't learn the sounds, but that is down to the poor teaching materials. It has not been considered an area of importance, plus a lot of the people involved in Irish teaching materials are not very competent nor practical/research focussed. The result is the blind leading the blind.

I have made over 70 palletographs and mapped out how to go between the two languages, plus how to relate them to phonemic codes and later on to IPA. I'm using tiddly wiki to make a guide. I think it is the most detailed and realistic guide ever done in Irish. For example, I've found a savage way to make the strong broad dental set; slender r has 3 different stratagems. It should hopefully be possible to learn what it is all about then!

So what you're describing is something like the system used in dialect books? You know, where they tell you the vowel sound in "bán" is /a:/ and take it for granted that you know (or will find out some other way) what an /a:/ actually sounds like.

I find that extraordinarily convenient, compared to either of the alternatives. Once you've got a basic handle on the sounds of the language, it's the easiest way to explain a particular word.

But of course, if I were trying to explain how my /t'/ differs from, say, Róman's /t'/, it wouldn't do at all. That's probably where IPA comes into its own.

(Message edited by Abigail on July 04, 2007)

quote:

if I were trying to explain how my /t'/ differs from, say, Róman's /t'/, it wouldn't do at all. That's probably where IPA comes into its own.

Very true. The question is - how often (if at all) learners need such information? /t'/ will cover a range of sounds as produced by different learners and as long their variety of sound is identifiable as /t'/ by any reasonable speaker - does it really matter if it is "uvular"-enough or too "fricative"? We (I hope you are with me at this) strive for speaking intelligible Irish with easy communication to natives and learners alike. So if you pronounce sound X which is understood as /t'/ by 100 people in the row does it really matter if it is not "alveo-palatal" enough or too "bilabial"?


p.s. I know that my frivolous use of phonetic terms turned the sentences ad absurdum, by my point was to convey the futility of striving for some idealized pronunciation if your aim is to speak intelligibly.

Exactly! I wouldn't say that that sort of striving is futile necessarily - I suppose it depends to some extent on your purpose in learning the language - but to me it's a bit of a sideshow.

When I look at my Irish, the biggest problems I see in it (by which I mean the biggest obstacles to my living through Irish 24/7 if I had the chance) are limited vocabulary, faulty grammar and lack of confidence. So I'll spend more time and energy working on those. If that means I wind up speaking Irish with an accent, well, at least I'll be speaking Irish - and in time and with continued exposure, I expect some of the accent will smooth itself out.

Regardless of accent, you need a dual system of normal vs soft (/b/ vs /bh/) and broad vs. slender (/b'/ vs /bh'/). Such fundamental structural differences, once set up should allow a degree of 'graceful degradation' that hopefully will forgive variance within a tolerance

I am a fan of the IPA, and although there is such a thing as a phonemic transcription, the learner needs a phonetic transcription.

If you are told the English word "red" is pronounced /red/ - and that this phonemic transcription can basically cover all dialects, then fine. But to learn English you need to decide which "r" to use: /ɹ/ or /r/ or /ʁ/ or /w/ or /ʋ/. [= Southern English r, Scottish r, Northumbrian burr r, pre-war schoolboy foppish w for r, or my own labiodental r] A phonemic r corresponds to any of these phonetic ones and more.

I agree on that, but at no point are people told how to make the sound so what symbol is used becomes irrelevant

So if you pronounce sound X which is understood as /t'/ by 100 people in the row

How will learners know the difference between /t/ and /t'/ if nobody explains it?

The problem is even worse with broad bilabial consonants. For example, in phonology, faoi would be written /fiː/. If you never explain that /f/ is [ɸʷ], the learners will just pronounce "fee". Do you see what I mean? I'm sure all the people who come here and ask for pronunciations don't all have books where this is explained. So, phonology isn't enough. Some phonemes are obvious, but many aren't. You have to explain how these phonemes are pronounced.

You could try /fw/ and explain that one must pout while saying 'fú'.
For slender f, then 'purse' lips (/f'u:/ or /fju:/) for 'fiú'.

Practically, there must be reflexivity where the person can hear and feel the sound difference between 'acub' and aguib', and a logical dimension where the intellect can see that the difference has a functional role in grammar. Such dimensions must be approached by teaching and training.

In essence, my concern is for a map to get to where IPA is both used and understood, but it cannot be the start

You could try /fw/ and explain that one must pout while saying 'fú'.
For slender f, then 'purse' lips (/f'u:/ or /fju:/) for 'fiú'.

Can't work: it's not phonology. If you write just how it is pronounced, then it is phonetics.

OK. I will go and write up my strategy in the week coming and post it along with pictures. After that I will do mp3s. I think I am confusing you here as to my intent

quote:

the learners will just pronounce "fee".

If I am not mistaken this word is felt like [f`i:] ('fí') by any native. I was never advocating ditching broad vs slender contrast - you are just changing the topic. The question is other way. In proper Irish this sound is bilabial, but I pronounce it labiodental. Does it really matter if it is strongly velarised in my mouth? You will never mistake it for [f'], yet it is not exactly the same phonetic sound, although surely it is the same phonem.

A Pheadair,

I didn't get your point. We used to have PWC consultant from Highlands that had a strong rolling-r (like Spanish rr) - and yet she is a native speaker and I never had any problem understanding what is she saying although her "r" don't match the mould of hot-potato-in-the-mouth way as "r" are pronounced in BBC. So? If a learner says a rolling "r" - does it harm his English? No, it doesn't. Any learner has an accent so one sound more or less - it really doesn't matter as long as it doesn't impede understanding. And rolling r's has never got anyone into trouble.

It's very important to have a system that accurately describes the basic utterance of a sound. For instance if you asked an English speaker how they make things plural, they would surely respond with, "add an s most of the time." They wouldn't say sometimes we add [z], other times [s], sometimes we changed the length of the vowel, sometimes we add an ee sound, and still sometimes we add [en]. This is way too confusing. Also, it's completely pointless. You teach someone to add an 's', and then when they say "oxes", you say, no, actually, that's an exception, it's really, "oxen." Or if they happen to say "zebras" with an [s] at the end, we say, not it's actually [z]. Same with "princes" and a great many other examples. But, it's really not too big a deal. If a guy said, "oxes" or "zebras" with a [s], (which is pretty annoying to do any way) we'd all understand him. There's probably already a big group of English speakers that do say childs and oxes rather than children and oxen.
The solution is to teach people the basics and then loads of vocabulary, so as to get them speaking as quickly as possible. Then, when they get comfortable with this, you start correcting small things here and there. Of course, there will be a student or two who will pick up on these differences right away and may have a bigger interest in the phonology or phonetics of the language, but this is the best way for the vast majority of language students. This is also not very practical for self-learners. Self-learners must learn the IPA, or some shortened form of it, to have any hope in learning the language. Not neccesarily in the beginning. But, as your knowledge and use of the language expand, the self-learner must reassess his pronunciation and check it against the ipa. Alternatively, a self-learner could just find someone else who knows the language and ask them, as this is really the best way, but that would make him into a non-self-learner.

Where I'm from, we tend not to use plural forms for animals, unless they're animals which we're very familiar with, like dogs, cats, horses, cows, donkeys. Elsewhere though, we don't pluralise: The field was full of giraffe, zebra and buffalo. We also say "zebra" with a flat initial vowel.

Dorothy: Do you think there could be wild animals in here?
Tin Woodsman: Perhaps.
Scarecrow: Even ones that, that eat... straw?
Tin Woodsman: Some, but mostly lions and tigers and bears.
Dorothy: Lions?
Scarecrow: And tigers?
Tin Woodsman: And bears.

Lions and tigers and bears oh my!

I think the usage of a "weak plural" in English for animals gives a stronger sense of infamiliarity and distance. Consider the difference between: That field is full of zebra and That field is full of zebras. The former is more clinical, I think.

Does it really matter if it is strongly velarised in my mouth?

In Irish, bilabial consonants are never velarised when they are broad, they are bilabialised, ie. followed by [ʷ].

Cibé ar bith, i ndiaidh ar mhínigh mé duit i gcuid mhaith teachtaireacht, leig tú ort nár thuig tú mo chaint. Tá mé ’gabháil a stad dó sin, níl ann ach am amú: is maith ar thuig tú a rabh mé ’mhíniú, gan amhras, ach is cosúil liom go leigeann tú ort nach dtuigeann tú mé siocair nach mian leat aidmheáil gurb agamsa ’tá an ceart. Ar aš turiu rašyti lietuviškai, kad tu mane supranti?

quote:

bilabial consonants are never velarised when they are broad, they are bilabialised

What a rubbish! Bilabial consonants are bilabial by their nature, so you can't "bilabialize" them even more for broadness. Broadness is achieved by straining back of your tongue towards velum and this is called velarization.

I the point is that, the important difference is that Irish broad bi-labials are very strongly pouted -and that is what matters. Given the broad and slender context, the tongue will often be raised at the back, but this is not important. See English b in 'but' and Irish 'buí', the latter is like 'bwee'.

The slender variety before front vowels have no palitisation , they are just pursed. It is the lip shape that matters, more so than any secondary -ization

IN the Old Irish book I think it points out that labialisation was late in the day, and originally separate from broad and slender, but got pulled into the system.

What a rubbish! Bilabial consonants are bilabial by their nature, so you can't "bilabialize" them even more for broadness.

Of course you can. Just have a look at any IPA chart and you'll see the symbol [ʷ], for example here: http://www.paulmeier.com/ipa/charts.html . They say "labialized". In Irish you never have [bˠ], [ɸˠ], [mˠ] etc, but [bʷ], [ɸʷ], [mʷ] etc. Can't you hear that?


Broadness is achieved by straining back of your tongue towards velum and this is called velarization.

Not for bilabial consonants, as I said. Or you don't hear properly.