[Spellyans] <dh> ~ <th>

Michael Everson everson at evertype.com
Mon Nov 29 14:17:35 GMT 2010

On 29 Nov 2010, at 13:22, A. J. Trim wrote:

> Could there be two forms, pempthek and pem'dhek?

More likely pempthek and pemthek.

> Perhaps some people pronounced the p and others didn't.

The p there is a part of a bridging consonant cluster, homorganic with the preceding nasal and voiceless like the following . In English you see similar things in Thomson and Thompson, Samson and Sampson, [sæmsən] and [sæmpsən]: the p may be inserted, but the following consonant is not voiced. We don't get *Samzon or *Sambzon. Now, if *pembdhek were found that would be another thing. 'Tisn't though. 

> I would expect the p to affect the voicing of the th/dh. Perhaps the full version was used in counting and the other used in connected speech.

A homorganic consonant inserted in such a position doesn't cause the devoicing of the following consonant.. it reflects the nature of that consonant. So we have Samson and Sampson, but chimney and chimbley. Now in that example, the n is converted to an l, but the bridging consonant is voiced like the following consonant, not voiceless.

> The evidence just reported is good, and it tells us how these numbers were pronounced when counting (pilchards?) in one location in the 1870s. It may be more reliable than Lhuyd's evidence.

AB doesn't tell us the context of many of its words. Some were collected in the field by him. Some by others. Some were rewritten by him in his phonetic script on the basis of manuscript readings. It's a mistake to try to treat AB in the same way we might treat one of Leonard Bloomfield's collected transcriptions of Algonquin languages. And it's a mistake to underestimate L1 interference from Welsh, as Nicholas has discussed. 

And he was not a trained phonetician such as we have today (or as we had even in Jenner's time). Modern phonetics didn't exist then. Indeed he was a pioneer. But so many examples of e.g. meneth vs menedh in stand-alone citation position has to be explained -- and can't be explained by external sandhi as Dan claims to find in JCH, because there is no external sandhi in citation forms. 

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

More information about the Spellyans mailing list