About the Yew1

[Douglas G Kilday]

Patrick C. Ryan (20 Jul 2001) wrote:

>> [DGK]

>> Regarding PCR's objection that the long vowel in Lat. <pe:s> 'foot' arose by
>> compensative lengthening from *peds instead of gradation: Palmer gives this
>> as an example of gradation. Apart from this crude appeal to authority, we
>> may note <capis> 'one-handled sacrificial vessel', gen. <capidis>, which
>> shows that Latin d-stems did _not_ regularly undergo compensation of *-ds in
>> the nom. sg.

>[PCR]

>The example you give, capis, seems to me to be different. if R = root vowel,
>it is CRCid-.

>pe:s, of course, would be CRd-.

>This could, of course, be a Dehnstufe but does not the absence of /d/ make one
>question that a bit. Furthermore, gradation is normally associated with
>syntactic or grammatical usages. What function would the Dehnstufe serve here?

[DGK]

Given the structural difference between <capis> and <pe:s>, my argument was
rather weak. A better one occurs to me. Latin -ds- is normally assimilated
to -ss-, and this has occurred with <ped-> in <pessimus> 'worst' <
*ped-simos lit. 'lowest, humblest'. Therefore, we should expect *peds >
*pess. But the nouns <as> gen. <assis> 'unit' and <os> gen. <ossis> 'bone'
show that the nom. sg. degeminates the final in Latin ss-stems. Hence, if
all cases of Lat. 'foot' had normal grade, we should expect nom. sg. *peds >
*pess > *pes, not <pe:s>. Gradation explains the observed form better than
compensation. Our <pe:s> presumably represents *pe:ds.

Aren't cases associated with syntax and grammar? Don't we have the same
phenomenon of Dehnstufe restricted to the nom. sg. in other IE words? Palmer
gives <Cere:s>, <pu:be:s>, <arbo:s>, and Gk. <phre:n> as examples. The
original "function" must have been to accommodate changes in the stress and
timing of syllables, for IE Ablaut is broadly parallel to the
vowel-alternations of modern Spanish, e.g. <duermo> 'I sleep', <dormi/> 'I
slept', <durmio/> 'he/she slept'. Short /o/ and /e/, stable in Classical
Latin, underwent severe changes in Vulgar Latin and Continental West Romance
depending on their position with respect to the word-accent, but original
/u/ and /i/ remained relatively stable. A similar episode must have taken
place in late pre-PIE (or whatever you choose to call it).