Personal Pronouns / Ergativity

[Jens Elmegaard Rasmussen]

Dear Wolfgang,

you have expressed your surprise that nobody from the IE quarters replied
to your posting of June 4. I do have a number fo questions and comments.
Some may appear silly, given my poor acquaintance with much of typological
discussion. However, to make a collateral field useful for IE studies, its
practitioners must argue their points in such a fashion that we can
understand them.  Therefore I take the liberty to ask for clarification
where I am doubt.

On Fri, 4 Jun 1999, Wolfgang Schulze wrote:

[...] Diathesis as a referential strategy (phrase internal as well as
> discourse dependent) presupposes a reference-dominated type. Many Modern
> IE languages are of this type [...], but PIE itself
> obviously lacked this strategy which comes clear from the fact that we
> cannot reconstruct a common "passive" for PIE. Diathesis is a very
> active feature in language change. It can come and go, and nothing
> allows us to propose a Passive for PIE just because a number of (modern)
> IE languages share this feature.

I strongly disagree witht he statement that PIE had no passive. In fact it
had several. One of the basic functions of the middle voice was patently
to express the passive use of transitive verbs. It goes so far that
causatives in the middle voice lose their causative meaning: Skt.
pa:ta'yate 'is made fly, flies'; also, while nasal presents from
adjectives are factitive in Hittite (tepu- -> tepnuzzi 'makes small'),
they are ingressive in Balto-Slavic and Germanic (ON rodhna 'turn red')
which must reflect the function of the corresponding middle-voice
inflection. Alongside this, there was the "stative" morpheme //-eH1-//;
this was stative with intransitive verbs, as aorist *sed-eH1- 'enter a
sitting position', prs. *sed-H1-ye'- 'be in a sitting position, sit', but
passive when added to transitive verbal roots, as *k^lu-eH1- 'be heard',
prs. *k^lu-H1-ye'- (Skr. s'ru:ya'te). This inflection combines the
Sanskrit passive and the Greek "e:n-aorist" into a PIE paradigm. The high
age of the passive meaning is proved by the equation of Skt. ja:'yate 'is
created' and Old Irish do-gainethar 'is born' from *g^nH1-H1-ye'-tor,
passive of *g^enH1-'create'. And of course the PPP of transitive verbs was
passive, as *g^nH1-to'-s 'created, born' *{kw} orr-to'-s 'made', and could
be used in combination with the genitive marking the agent. On the basis
of idioms from different parts of IE one can safely posit a structure like
*me'dhu H2nro's {kw}rto'm (H1esti) "the mead has been made by the man"
as belonging to the protolanguage.

> If we look at the question of ERG and passive, we can sometimes observe
> a tendency to establish a pseudo-ergative strategy based on the passive
> (some modern Indo-Iranian languages, partly Cl. Armenian etc.). But this
> is a secondary process often bound to specific (perfective) TAM forms.

In what way is this "pseudo-ergative" different from a "real ergative" -
other than by the fact - not too frequent in typology - that its
prehistory is known?
-  Is TAM "tense-aspect-mood"?

[...]

> Now, IF (I say IF) you want to ascribe some ERG features to the PIE
> system it would make much more sense to declare ALL transitives as old
> (and generalized) antipassives [...]

Would that be possible? Some IE verbs are underived, can there be such a
thing as a morphologically unmarked antipassive verb in a language? But
the idea of antipassive looks good for the marking of the object: could
*-m be an old adverbial ("goal") case originally used with an antipassive
and generalized from there? You give up the idea for other reasons:

> But this does not make sense as long as we don't have substantial
> evidence for ERG strategies elsewhere in the paradigm [...]

I believe we do, but is that really relevant? Could not the _forms_ of the
verb be the sole survivors from an old ergative system?

In an aside, you then speak of:

> Georgian
> which has an ACC (< AP) strategy in the present/future tenses/modes, but
> an ERG strategy still present in the aorist, hence Georgian mirrors say
> some Iranian languages on an ergative basis[].

A burning question to the IE-ist: Why is the ergative part of the Georgian
verb held to be older than the non-ergative part? And why is its
accusative construction taken to be an old antipassive?

You suggest a set of structural combinations for basic sentence types,
allowing for variants, but insist:

> [...] Still, the overall picture remains the same: The operating system
> of PIE clearly showed an ACC strategy in its protypical kernel,
> semantically split according to [±animate] or so. This ACC strategy
> seemed to be dominated by topicalization routines with animates, a clear
> indices for the semantic basis of PIE "case" marking. Finally, AGR does
> not change this picture, even if we assign the *-H2e etc. series to
> statives/inactives, and the *-m etc. series to dynamics/actives: In this
> case, even the dichotomy [±anim] becomes irrelevant, because it does not
> show up in a specific set of clitics. ALL these clitics have an ACC AGR
> scheme...

If I understand this correctly, you are addressing several layers of the
language in one mouthful. I agree that the PIE we reconstruct was not
ergative, but had the same basic syntax as Sanskrit, Greek, and Latin. But
does that exclude the existence of an older structure? Or are older
structures disqualified as ergative for some other reason? What is the
"topicalization" business based upon? If animates as subjects are marked
with an *-s, it would mean that this role was not self-evident for them.
Could that not indicate that the whole statement was primarily about the
object, which was then only minimally marked, viz. in the case of animates
for which this role is not as self-evident as it is with inanimates? A
patient-centered statement structure is exactly what the classical
ergative is. However, I am in great doubt, as I have grave misgivings
about some of the alleged evidence for an IE (better, pre-IE)  ergative;
thus, the two endings containing /s/, the nom.sg. (*-s) and the gen.sg.
(*-os), are not at all identical if one cares to look closer. Still, since
we are groping in the dark anyway, what differences we have found between
them may turn out to be secondary and ultimately not relevant.

In there has been a pre-IE ergative, where does that leave the reflexive?
The reflexive pronoun replaces the anaphoric when referring to the
_subject_, in which case the transitivity or intransitivity of the verb
does not matter. Shockingly, perhaps, this is the same in Eskimo: the
reflexive ("fourth person") is used whenever 3rd person reference hits the
subject - be it transitive or intransitive. In this selection, there is no
companionship between the subject of an intransitive verb and the object
of a transitive one. That would rather indicate that the ergative
constructions are later creations, just as they are in Modern Indic.

Could you elaborate on the clitics you have in mind? What are you talking
about here?

Jens