Ergativity and verb forms

On 18 Jun 2004 Doug Dee <AmateurLinguist@AO...> wrote:

> I've read that some of the Cetic languages have sentences roughly like these:
> I am at eating the pizza (=I'm eating the pizza)
> [...]
> the verb "am" is intransitive, & it's followed by a preposition and
> then a verbal noun/gerund/whatever you call it with its object.
  This construction is typical in Uralic languages and the
"whatever" is classified usually as infinitive. Uralic infinitives
are not premodified by prepositions rather they are suffixed by
case markers, but grammatically they are equal constructions. This
type exists in English too: "I am _about to eat_ some pizza".

  These constructions can be used to express temporal and aspectual
features but they are used much frequently (in Uralic laguages) to
represent subordinative clause, cf. Hungarian "Jo:tto:mben
la'ttalak" 'I saw you when I was coming', lit. 'In-my-coming I-saw-
you'.

  I have no ergative data of the temporal/aspectual usage but I
have found a Basque compound sentence:

[emakumeak gizonari liburua ematea] nahi du umeak
woman-the-ERG man-the-DAT book-the give-INF want has child-the-ERG
'The child wants [the woman to give a book to the man]'

  Thus there are two independent grammatical relations. The focus
of the first one is the main finite verb: Basque "nahi" 'want' is
transitive therefore subject "ume" 'child' will be in ergative and
the direct object, that is the whole subordinate clause, in
absolutive. In order to be able to put a whole clause into
absolutive, its verb "ema" is transformed into a finite form
"emate" 'giving, to give'.

  The second relation is between the verb of the subordinating
clause and its subject and complements. In this case verb "ema"
'give' is transitive, therefore the subject "emakume" 'woman' will
be in ergative, the direct object "liburu" 'book' will be in
absolutive and the indirect object "gizon" 'man' will be in dative.
When transforming finite verbs into non-finite forms, the
grammatical markers of their complements may or may not change
(e.g. accusative into genitive): the above example shows that they
are not changed in Basque.

  To apply the above to your problem, we have also two relations:
(1) "I am at eating pizza" := (2) "I am at X" + (3) "I eat pizza".
These sentences can be transformed into ergative type as: (2a) "I-
NOM am X-LOC" (3a) "I-ERG eat pizza-ABS". Thus if we put together
(2a) and (3a), the result can be (1a) "I-NOM am eat-INF-LOC [I-ERG]
pizza-ABS" (supposing that non-finite vebal forms require the same
grammatical markers as finite ones).

  Comments:
- I distinguished the grammatical subject of the intransitive verbs
(in NOM-nominative) and the on eof the transitive verbs (in ABS-
absolutive): it is possible but not necessary.
- IMHO a typical ergative language is agglutinative, therefore I
supposed a locative case (LOC) marker instead of the preposition
"at".
- I used the term INF-infinitive for the non-finite verbal form but
it may be replaced by supine, gerund, participle, etc. according to
the tools of the actual language (however, I do not propose verbal
nouns: they usually require the remapping of grammatical markers,
e.g. turning accusative into genitive).
- The above word order is just one of the possible solutions.
- You may or may not omit "I-ERG": it depends on the actual
language defaults. This unfilled slot may be filled in ny default
both with the subject of the main clause and with "someone" or
"previously mentioned one".

Reply