Personal Indices and Personal Pronouns

[Gideon Goldenberg]

Dear typologists,

The discussion about the personal morphemes is important and
welcome, but many of the ideas brought forth have been recycled
throughout the ages too many times. In explaining the constitution
of an inflected verb-form, the basic analysis, at least so far as
recognizing the inflexional marker as subject-pronoun, seems
generally clear, and in no few languages even the etymological
connexion between all personal allomorphs can hardly be ignored.
Synchronically, a recognized common origin of independent personal
pronouns, proclitic or enclitic personal markers, inflexional affixes,
and agglutinated or otherwise incorporated personal morphemes is not
a necessary condition for recognizing them as allomorphs, but where
etymological connexion is obvious, it is easier to be convinced
that personal pronouns, affixes, indices etc. belong in the deep level
to the same morphemes.
The simple fact that an inflected verb-form where both the actor (etc.)
and the verbal idea are represented (nexally related) consists of two
morphemes has been repeated many times in the history of linguistics.
From the 10th cent. Arab grammarian Al-Zajjaji we may quote the statement
that "since the verb would not be devoid of an actor and would in no way
dispense with it, a pronoun was attached to it which became like one of
its letters, and the sentence became one word". No few writers thought
"that the personal pronouns are contained in the Greek and Latin
terminations of the three persons of their verbs", that "in the Latin
Vol-o, Vol is the verb; and o a common removable suffix, with a
separate meaning [viz] Ego". In discussing the origins of inflexion,
linguists who did not accept the agglutinative hypothesis because it
could not account for the fact that the verbal person-endings in Indo-
European were "not self-explanatory, that is to say, did not contain the
singular, dual, and plural forms of the personal pronouns", actually
admitted that if they did so, there would be as little difficulty in
proving the pronominal origin of personal indices as there is in the
case of the first and second persons of the Semitic verb, since "the
amalgamation of a pronoun with a participle to express a verbal idea
is not foreign to Semitic grammar". In the modern linguistic study
of the Semitic languages at least, the very nature of the inflected
verb form was beautifully formulated by David Cohen, stating that
"[le verbe] peut constituer par lui-même un énoncé assertif fini du
fait de sa nature dimorphématique: base verbale + marque personnelle
représentant le sujet". As put by Emile Benveniste, "'Volat avis' ne
signifie pas « l'oiseau vole », mais « il vole, (scil.) l'oiseau ».
La forme volat ... inclut la notion grammaticale de sujet".
A great drawback of twentieth-century morphological analysis was the
conception of personal index in a verb-form as just an agreement-sign
to a preceding independent pronoun, which pron. may then be deleted in
so-called "pro-drop" languages. This conception was motivated by the
assumption, following the tradition of elementary school grammar, that
the basic sentence would universally consist of a noun phrase and a
verb phrase. In a letter to this forum on 2 April 2001, Claude Hagège
has judiciously marked that "the 'pro-drop' issue in formal
linguistics has  brought about a certain amount of confusion: the
so-called 'pro-drop' languages may drop the pronoun, but not always
the personal index. [So] they are not so pro-drop as they appear".
I would not limit the term pronoun, as Hagège does, to the stressed
independent forms only, but it was refreshing to find this comment,
though hyper-delicate, in a general-linguistic discussion. Whatever
term we use for the personal affix, the important point, I should say,
is that it brings the personal actant to which it refers to be bodily
present within the inflected verb-form.
If we accept the inflected finite verb as a predicational complex with
a pronominal (or "indexed") subject, then the agreement of the nominal
subject with the verb is in fact marked by apposition of the nominal (or
independent-pronominal) subject and the personal pronoun incorporated
in the verb-form. We may find that all grammatical agreements involve
in fact apposition and repetition.
The process of weakening inflexions, adding independent personal pronouns,
which again are becoming like inflexional affixes is common enough, and
"English 'I' [which] seems to be on its way to become an index" as
mentioned by Elisa Roma in her letter to this forum last week is no
great surprise; French 'je' has already gone farther.
                                                    Yours,    Gideon.