Final Theta in Spanish

[Eduard Selleslagh]

----- Original Message ----- 
From: "Larry Trask" <larryt at cogs.susx.ac.uk>
Sent: Thursday, November 23, 2000 12:33 PM

> David White writes:

>> Should not all Latin 3rd declension nouns
>> in /k/ show up with final theta in Spanish, if "lux" and "pax" do?  I cannot
>> at the moment think of any more, but there must have been more than 2 ...

[Ed Selleslagh]

e.g. falx > hoz ( -al- > o, cf. French Gaule < Gallia), radix > raíz, rapax
> rapaz (and analogous formations: loquax > locuaz). Actually one should
consider the accusatives:falcem, radicem, rapacem, loquacem... On the other
hand: sex > seis, not *seiz, but here the (non-extant) accusative ending
doesn't come into play, so it would be simply the loss of -k-, a regular
phenomenon.

Other words in -z have different origins: haz < fascem, faz < faciem (this one
does have a -k root)

> There were, but I'm not sure they were all that numerous, and anyway
> not all of them survived into modern Castilian.  For example, Latin
> <ducem> 'leader' should have yielded Castilian *<duz>, but no such
> form is recorded, and modern <duque> 'duke' is clearly borrowed
> from Old French <duc>, which itself appears to be an analogical form.

> Anyway, the history of final /e/-loss in Castilian is a trifle
> complicated.  For example, Latin <dulcem> yielded the expected <duz>,
> well attested in medieval Castilian, but the variant <duce> seemingly
> never disappeared, and modern <dulce> exhibits not only the final /e/
> but also an extra /l/ which it picked up somehow.  As for <doce> '12',
> from <duodecim>, I confess I have little idea why that /e/ is still
> there.

> Larry Trask

[Ed]

'dulce' has all the characteristics of a 'learned' later formation, i.e. after
the Renaissance (like próximo, fructuoso....). In other words: the -l-wasn't
'picked up', it was never lost as it should have, because it didn't undergo
normal evolution: it skipped it. All Latin languages are full of that, e.g.
French: Traditional: moustier/moûtier <> modern: monastère.

It seems there is a 'need' for two syllables, maybe a lingering memory of the
multisyllabic Latin origin, or just false analogy? (In S. American popular
Spanish some say even 'seise' for 'seis'(cf. siete < septem)). There is a
strong incentive for -e in 'doce/trece': avoid confusion with 'dos/tres',
notwithstanding the s-z/c opposition, but that is absent in 'catorce,
quince'... So it seems the -ce is rather the remnant of '-decim', the -d-
having been lost like in 'raíz' and many other words (and in popular speech:
-ado > ao, -ada > aa > á). No wonder, especially if z/c was originally
affricated as is often thought.

Ed.