Re: VSO language evolving from Nom/Acc to Erg/Abs

Peter Bleackley wrote:

>
>
> Now suppose that the language evolves towards an Ergative/Absolute
> argument structure. What changes would you expect to take place? The
> first idea that comes to mind is that the person and number marking on
> transitive verbs is reinterpreted as referring to the object. What
> else might happen?
>
>
Well, in many Indic languages, the past tense is ergative/absolutive and
the present is nominative/accusative. This occured, I believe, because
the past tense is derived in such languages from a past participle with
a passive (or perhaps adjectival) meaning of sorts. This participle is
often used in nominal sentences (that is, sentences without a verb), so,
for example

जितो राक्षसः
jito raakSasaH
ji-taH raakSas-aH
beat-PART.NOM.SG.MASC raakSasa-NOM.SG.MASC
'Raakshasa is beaten'

This can also be interpreted, as 'Raakshasa has been beaten'. Now, when
you want to know who beat Raakshasa, you need to use an instrumental. Hence:

जितो राक्षसश्चाणक्वेन
jito raakSasashcaaNkyena
ji-taH raakSas-aH caaNky-ena
beat-PART.NOM.SG.MASC raakSasa-NOM.SG.MASC caaNky-INST.SG.MASC
'Raakshasa is beaten by Caankya', 'Raakshasa has been beaten by Caankya'

This eventually became more common than the finite past tense, which has
since (I believe) been lost in languages such as Hindi etc. The
participial construction, meanwhile, has been reanalysed as a finite
verb of sorts, which agrees with the object in transitive sentences and
the subject in intransitive sentences. Also resulting from this
construction is an ergative case of sorts, descended from the original
instrumental, while the absolutive is descended from the nominative. In
some Indic languages (such as Nepali), this Ergative/Absolutive
distinction has been leaking into the present tense. Perhaps something
similar could happen in yours?