LINGUIST List 15.696

Wed Feb 25 2004

Sum: Thematic Vowels in Latin Verbs

Editor for this issue: Naomi Fox <>


  1. Bruno Maroneze, Sum: thematic vowels in Latin verbs

Message 1: Sum: thematic vowels in Latin verbs

Date: Tue, 24 Feb 2004 17:58:49 -0500 (EST)
From: Bruno Maroneze <>
Subject: Sum: thematic vowels in Latin verbs

Dear linguists,
On January 31th I posted the following query (Linguist 15.391):

''I have a question on the meaning of the thematic vowels in Latin
verbs. I was always thaught that the only function of thematic vowels
was to indicate the inflection class of the verb (first, second, third
of fourth conjugation). But I think it is very strange that there
exists a morpheme which has only a grammatical function and doesn't
have meaning. My hypothesis is: in earlier stages, the thematic vowel
was a ''full morpheme'', (maybe even unbound), which had a meaning
possibly related to the Aktionsart or the valency of the verb. Later,
this morpheme suffered a grammaticalization process and lost its
meaning. I wish to know if this problem was already studied; could
someone point me some bibliographical references on this matter?''

I received many interesting responses. First of all, here are some
suggested books on this matter: Angelo Mercado and Dag Haug indicated
me the following book: Meiser, Gerhard. _Historische Laut- und
Formenlehre der lateinischen Sprache_. Darmstadt: Wissenschaftliche
Buchgesellschaft, 1998.

Remy Viredaz indicated me these two titles: Antoine MEILLET,
Introduction � l'�tude comparative des langues indo-europ�ennes, 8th
edition, Paris 1938, reprinted by the University of Alabama Press,
pages 197-223 Helmut RIX (ed.), Lexikon der indogermanischen Verba,
2nd edition, Ludwig Reichert publishers, Wiesbaden 2001 (in German),
especially p. 14-25

Dag Haug and Nick Pharris suggested Andrew Sihler, New Comparative
Grammar of Greek and Latin, Oxford 1995, pp. 528-546.

John Koontz also suggested Meillet's book, and L.R. Palmer. Reprint
1988. The Latin Language. U of Oklahoma Press, Norman, Oklahoma.

Pius ten Hacken indicated me Aronoff's analysis: Aronoff, Mark
H. (1994), Morphology by Itself: Stems and Inflectional Classes,
Cambridge (Mass.): MIT Press.

The responses I received may be divided in two ''groups''. First, some
of them were about the diachrony of thematic vowels:

Remy Viredaz wrote a very interesting explanation: ''As for its
origin, a given Latin inflectional class or ''conjugation'' may have
various origins. For instance, -e:- (long e) of the second
conjugation can go back either to Late Proto-Indo-European *-e:- , a
fientive suffix (meaning the subject enters in a new state), or to
Late PIE *-eye- (alternating with *-eyo- depending on person and
number), a suffix that had itself two main functions: either an
intensive or iterative function (often no longer perceptible), or a
causative function (often no longer recognizable, because of semantic
change or/and because the basic verb may have disappeared). Latin
verbs in -a:- can go back to verbs with an athematic suffix Late PIE
*-a:- (factitive, e.g. novare 'to renew', from novus 'new'), or with a
double suffix *-a:-ye- / *-a:-yo- (denominatives from nouns in *-a:-),
or of still different types (e. g. ara:re 'to plow' from the root
Late-PIE *ar- ( y"'schwa'') + suffix *-ye-/*-yo- with no specific
value). (I write ''Late PIE'' meaning the stage after the
disappearance of the so-called laryngeals. At that stage the various
branches of the Indo-European family were already separate dialects.)
The function of Indo-European verbal suffixes cannot always be
recovered. For instance, the suffixes *-sk'e-/*-sk'o- or *-neu-/*-nu-
have largely different values in the different languages that still
have them.''

Dag Haug also explained me these etymologies:

''the Latin conjugations are ''historical accidents'' which reunite
historically different morphemes. Here is a short overview of the main
sources: The third conjugation represents the most productive way of
conjugating a verb in late Proto-Indo-European, with a short vowel
between the stem and the ending. The capio-type is a variant where the
alternating vowel e/o is preceded by a y - if the stem syllable is
long, the verb ends up in the fourth conjugation, like sentire. The
fourth conjugation also contain denominative verbs from nominal
i-stems, like finire to finis. The first conjugation contains
denominatives from nominal a-stems and later also o-stems. It also
contains factitives to adjectives, like novare to novus. In the second
conjugation, you find causatives and statives - these have fallen
together, since -eye- of the causatives gave a long e, which is the
inherited stative suffix.''

Finally, Hayim Sheynin sent me a message in Latin: ''Nescio
publicationes de vocalis thematicis in lingua Latina, sed censeo
vocales thematicae coniugationis primae et secundae (a et e) ante
earum grammaticalisationem indicatae substantiam transitivam fuerunt,
quondam vocales thematici coniugationis tertiae et quartae (brevis e
and i) substantiam intransitivam (sive transitivam non activam). Hic
est prima idea solo.''

With that, it is clear that Latin thematic vowels were full morphemes
(suffixes) in earlier periods, which indicated aspect, transitivity

The second group of responses was about the morpheme status of
thematic vowels. Frank Y. Gladney suggested that they are simply
buffers between morphemes: ''I think thematic vowels (and consonants)
are simply buffers between morphemes that are introduced by
readjustment rules. For example, when Old Russian _^este_ /e:d-te/
'you (pl.) eat' was replaced by _^edite_, all that happened is that
the form was made morphologically more transparent; its morphological
structure was not changed. Another example of the workings of
readjustment rules would be the _-b-_ in the noun _freebie_ 'something
given or gotten free of charge', which has the morphological structure
/fre:/ + /e:/.''

Lisa Bennett told me about morphemes that have no meaning: ''There are
other instances of morphemes that have no meaning, especially in
languages that have a requirement on the minimal size of a word. Some
African languages add yi- to monosyllabic roots when they would
surface as such, that is, in the imperative for instance.''

I am still reluctant in accepting ''morphemes without meaning''.

John Koontz and I discussed a bit about that, and he suggested me an
interesting solution: ''That's an interesting issue. I don't know the
status of thematic vowels in theories of Romance linguistics (as
opposed to PIE linguistics). By analogy with the more elaborate PIE
analysis, the theme vowels might be either whole morphemes or parts of
them, depending upon the stem in question. That is, perhaps in some
cases - here my knowledge of Romance morphology is inadequate - it can
be shown that the theme vowel is part of a larger entity. (Maybe -iza-
or -isc- in Spanish?) Or, perhaps Romance linguistics sees theme
vowels always as morphemes. In the former case one might argue that
they were always a property of the stem, even if they were not always
morphemes in themselves. That is, they would be a property, on the
order of the number of syllables, or the accentual pattern, that had
morphological significance without morpheme status.''

This last sentence is interesting: the thematic vowels would be a
property of the stem, with morphological significance, but without
morpheme status. Thank you very much to all those who answered my

Best regards,
Bruno O. Maroneze
Post-graduate student
University of S�o Paulo - Brazil

Subject-Language: Latin; Code: LTN 
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue