[Spellyans] th/dh

Michael Everson everson at evertype.com
Sun Nov 21 16:18:27 GMT 2010

On 21 Nov 2010, at 15:13, Daniel Prohaska wrote:

> Yes, but the mistake you made has been in interpreting what Lhuyd's transcriptions reveal. As Nicholas showed:”
> And you are not interpreting Lhuyd’s writings?

Syntax issue. It is not the fact that one reads and interprets Lhuyd's transcriptions that is mistaken, but your interpretation of them is mistaken, in my view. Nicholas' arguments are simply better than yours, and he supplies lists of examples. You have not. 

> “pazuârdhak ‘fourteen’ AB: 134c
> pemdhak ‘fifteen’ AB: 135a
> huettag ‘sixteen’ AB: 147b
> seitag ‘seventeen’ AB: 148c
> eitag ‘eighteen’ AB: 105b”
> You say Lhuyd is influenced by his native Welsh. Why then does he write pemdhak ‘fifteen’ (cf. W pymtheg), what is it Welsh influence or not? Just an example.

Are you then arguing that we should trust Lhyud here and write peswardheg and pemdheg?

I have already said: when Lhuyd shows devoicing in final unstressed syllables, he is describing what is going on in Cornish. When he is not, he is being influenced by his Welsh. 

If, on the other hand, Cornish and Welsh are the same, and the Cornish are saying -dh and -g and the Welsh are saying -dd and -g, then you have to explain why Lhuyd writes -th and -k.

And I don't think you can. 

> “Lhuyd is not consistent with final consonants in unstressed position. This isn't surprising: he wasn't attempting to do what we attempt to do, and the linguistic environment he did his remarkable work in was one in which people were still comparing Irish and Phoenician.
> You've over-interpreted Lhuyd,”
> This is an empty claim. I could as easily say, you are over-interpreting the MSS texts, as they don’t distinguish <dh> and <th>. How do you arrive at a phonemic /ð/ : /θ/ contrast at all? Argumentation has to go both ways.

Arguing about whether there are phonemes at all is a bit of a red herring. It is easy to trust BM and the Charter Fragment and Lhuyd in terms of initial and medial th- and dh- for instance. Arludhy (BM arlyȝy, Lhuyd Arlẏdhi) well attested, as is dhodho (BM ȝoȝo, dhodho). We have consensus about this. Everybody has consensus about this. 

The difficulty is in final unstressed syllables, where Lhuyd is inconsistent. An explanation must be found for the inconsistency. 

We hold that devoicing is systematic in Cornish as it is in German and Breton and Russian and other languages. We hold that Lhuyd's -th shows that he heard the devoicing, and that -dh (and -g in eitag) shows influence from his Welsh. 

If devoicing is not systematic in Cornish, then what is your explanation for Cornish -k and -dh in Lhuyd? And if you want to write the latter <dh>, why do you not want to do as Bailey does and write Kernoweg with the same rule to de-voice that?

> “and you haven't offered a counter explanation for why -th turns up in some words and -dh in the same words.”
> Yes, I have. You archive everything, I’m sure you’ll find it.

That's the dodge Bailey tries. Give your explanation or admit that you haven't got one. If devoicing is not systematic in Cornish, then what is your explanation for Cornish -k and -dh in Lhuyd?

> “We have: Welsh influence.”
> Hence <pemdhak> and <noweτ>… unconvincing.

Those are examples of Lhuyd getting it right. 

>> “If you take that into consideration, you're back at a logical description with a voiced/stressed v. unvoiced/unstressed opposition. If you don't, then you still have to explain the voiceless transcriptions.”
> And guess where <noweτ> occurs, at the end of the sentence, before the full stop.

Then why is this alternation not found for all the rest of the consonants? The model you propose is extremely unlikely. Other languages which devoice regularly do so... regularly. Your analysis of Lhuyd makes Cornish into a chimera. 

> “If Lhuyd writes -th where Welsh has -dh, you can be sure that's what he heard, because he'd never make that mistake otherwise.”
> This also works the other way round, so where he writes <dh> where Welsh has <th> he’s also writing what he hears (if this isn’t over-interpretation).

Then you can't explain why he writes -th.

> Why don’t you believe him, when he writes <diuadh> in a colloquial phrase (<pa thera diuadh an vledhan>) that he surely heard? Is <vledhan> also owing to Welsh influence?

You forget that he is not transcribing from tape. He has been working with Cornish for some time, and has been writing it and may have been following orthographic choices he made, consciously or unconsciously. 

> “And given the context of the voiced/stressed v. unvoiced/unstressed opposition, if there are other examples of -dh where we wouldn't expect it, it's not unreasonable again to attribute this to interference from Welsh.”
> I find this “influence of Welsh” argument to be special pleading. Wherever it suits your theory

Our theory holds for a wide variety of languages which devoice consonants in final position regularly.  

> the examples are accepted, but where they don’t they are attributed to Welsh influence, except he doesn’t always write <δ> and <τ> where Welsh has <dd> and <th>, not in stressed nor in unstressed syllables.

No, he sometimes writes -dh and he sometimes writes -th even for the same words. Throughout his book, not just in one rather short transcription. And many times the words are written in pausa (as in the vocabulary lists) where your theory ought to be borne out -- but it isn't. 

> Maybe the development and situation in LC was a little more complex that your explanation. Anyway, there are a lot of maybe, too many for me, to accept the implementation of one man’s theory.

That's a pity, but I am sure that eventually the penny will drop. 

> “And the k/g distinction above bears exactly the same thing out.”
> The k/g distribution is not in dispute; it already looks less sure as far as the p/b distribution is concerned, and is even more difficult to ascertain with the conflicting evidence, or rather the evidence conflicting with lack of evidence in the MC MSS, what the situation really was. I remain unconvinced and highly sceptical.

So instead of a regular linguistic pattern you have a chimera, a language which doesn't devoice regularly, but does so by treating one consonant differently from the others. And the phonological motivation for that is...?

> “Essentially Lhuyd started from Welsh, and saw Cornish as a set of derogations from that.”
> I do not believe that this statement is true. He went to Cornwall specifically to record Cornish, not Welsh.

That doesn't mean he didn't treat Cornish as a variety of Welsh. I speak Spanish and German, and when I approach Portuguese and Dutch, I see them through the lenses of my Spanish and German. One resists interference, but it is inevitable until one attains fluency. 

> I believe you are over-interpreting his work. Nowhere did he state that he took Welsh and derogated when weir Cornish forms cropped up. Unconvincing, Michael.

Lordy, Dan. He hasn't got to state this for observers after the fact to *see* Welsh influence in his phonology, grammar, and syntax. You make it sound as though we want to rubbish Lhuyd as a whole. No. We don't. But there is evidence for interference from Welsh. Indeed there would *have* to be interference from Welsh. It could hardly be otherwise. As good investigators we have to take this into account in order to do a better job rather than a worse one at interpreting Lhuyd. 

In the matter of final unstressed -dh/-th he is inconsistent, and the simplest explanation is not some complex array of sandhi environments operative only on /ð/ but simply that he is being influenced by Welsh. 

You already agreed that he was influenced by Welsh in writing -ev since the segment had become -e by his time. 

> “This explains many inconsistencies which otherwise have pretty much no explanation at all.”
> Have you considered the influence the MSS spellings may have had on his spellings.


> After all he was confronted with MSS that didn’t distinguish /ð/ from /θ/ at all and wrote <th>. Only PA has <ȝ> but this is also inconsistently used.

This is incorrect. Yogh is used in the Charter Fragment. in PA, and in BM. There are 284 different words using yogh in PA and 60 different words in BM. 

Michael Everson * http://www.evertype.com/

More information about the Spellyans mailing list