Re: Old Albic update: Syntax of the verbal noun

Hallo!

On Tue, 16 Oct 2007 11:54:17 -0700, Elliott Lash wrote:

> Hello!
>
> --- Jörg Rhiemeier <joerg_rhiemeier@...> wrote:
>
> > Hallo!
> >
> > This time I wish to present you some brief notes on
> > the syntax of the
> > Old Albic verbal noun.
>
> Excellent!
Thanks!

> > The Old Albic verb has only one infinite form, the
> > verbal noun (VN),
> > which is of great importance in the syntax of the
> > language.  The VN
> > is formed with the suffix -°nth, e.g. _hatanth_
> > 'biting'.  It is
> > an inanimate noun which has no plural, but is
> > inflected normally
> > for case.
>
> Is it at all related to, or inspired by, Indo-European
> type participles or gerunds such as Latin (-nd- or
> -nt-), Germanic (-nd), etc?  That is, do you think
> that such a morpheme is not only Indo-European, but
> also Proto-Europic in general?
It is indeed related to the IE active participle in *-nt-.
The Proto-Europic form perhaps was a converb (like the
-te form in Japanese).

>
> > The VN is used whenever a clause functions as an
> > object (or subject)
> > of another verb.  In such an infinite clause, the
> > arguments of the
> > VN are coded as possessors, with the agent appearing
> > in the genitive
> > case and the patient in the partitive case if
> > animate and in the
> > locative case if inanimate:
> >
> > (1)  Asalama hatanth as chvanas ol ndarol.
> >      AOR-see-3SG:P-1SG:A bite-VN the:C-GEN dog-GEN
> > the:M-PRT man-PRT
> >      'I saw the dog bite the man.'
> >
> > Such sentences can also be constructed differently,
> > namely by
> > coding the subject of the embedded clause as the
> > object of the
> > matrix clause and linking the embedded clause with
> > the conjunction
> > _a_ 'and' to it:
> >
> > (2)  Asalama om ndarom a hatanth sol as chvanas.
> >      AOR-see-3SG:P-1SG:A the:M-OBJ man-OBJ and
> > bite-VN he-PRT
> >      the:C-GEN dog-GEN
> >      'I saw the man and his biting by the dog.'
>
> Is there any difference in the encoding of the
> argument of intransitive verbs?  For example, in Old
> Irish, intransitive verbs could have a genitive
> subject, whereas transitive verbs could only have an
> agentive oblique subject with the preposition 'do':
>
>   -  is maith lim        a   thuidecht    "I prefer
> that he come"
>       is good with.me his coming
>
>   -  ni       a/il lim           mo marbad duit   "I
> don't want you to kill me"
>       not.is good with.me my killing   to.you
The intransitive subject is in the genitive if it is
an agent, and in the partitive/locative if not:

(15) nacvamanth bratos malos
     arrive-VN brother-GEN 1SG-PRT-M-GEN
     'the arrival of my brother'

(16) nacvamanth gratathal thalamal
     arrive-VN letter-LOC 2SG-PRT-LOC
     'the arrival of your letter'

The brother is an agent (he has been actively travelling
all the way, more or less), the letter of course not
(an inanimate object that has been carried about).
These examples show that the semantic role of the noun
is important; even the same verb may govern both cases.

(The partitive case of the pronoun indicates inalienable
possession - kinship in case of the brother, origin in
case of the letter.)

> Later on, when the accusative verbal noun construction
> was replaced (in some instances) by the dative verbal
> noun with 'do', the situation changed. There were two
> options: a) the object was expressed in the accusative
> before the verbal noun and the subject in an agentive
> oblique phrase after. Alternatively, the subject could
> precede, in the accusative and the object could follow
> in the genitive. With intransitive verbs, option a)
> was taken (although obviously, the argument in the
> front of the verbal noun was the subject not the
> object).
>
>
>    - is maith lim Craiptine do dul
>       is good with.me Craiptine to go
>   "I prefer that Craiptine goes"
>
>    - roscnai                   a n-echu           do
> marbad don rig
>     prf.3pl.bothered.3s their horses.acc to kill
> to-the king
>   "It bothered them that the king killed their horses"
>
>   - roscnai                   Craiptine        do
> seinm suantraige
>      prf.3pl.bothered.3s Craiptine.acc to sing
> sleepmusic.gen
>   "It bothered them that Craiptine sang sleepmusic"
>
> (note, these are all made up examples, but created by
> analogy with the patterns in a corpus I've built)
Interesting.  I am not very familiar with Irish, I must admit.

>  Anyways, I'm interested if you've thought about how
> the various case markings developed and if they'll be
> differentiated according to verb type, basically?
The active-stative morphosyntactic alignment of Old Albic
is inherited from Proto-Europic (there are traces of it
in PIE).  The usage of the cases with the verbal noun is
one of those design issues where a solution suggested
itself.  I wanted the same alignment as with finite
clauses with the verbal nouns.  Then I realized that I
had two kinds of possession - alienable and inalienable -
in the language and could use these to encode the semantic
roles of the arguments.

>
> > The verbal noun is also used in three periphrastic
> > stative verb
> > aspects: the progressive, the perfect and the
> > prospective.
> >
> > The progressive aspect expresses that the subject is
> > in a state
> > of doing (or undergoing) something.  It is formed
> > with the verb
> > _b-_ 'to be (temporary, cf. Spanish _estar_)' and
> > the locative
> > of the verbal noun:
>
> > (3)  Baha matanthal bradal.
> >      be-PRS-1SG:P eat-VN-LOC bread-LOC
> >      'I am eating bread.'
> >      (lit. 'I am in the eating of bread.')
> >
> > The perfect aspect expresses that the subject is in
> > a state
> > resulting from a past action or event.  It is formed
> > similarly,
> > but with the ablative of the VN:
> >
> > (4)  Baha matanthad bradal.
> >      be-PRS-1SG:P eat-VN-ABL bread-LOC
> >      'I have eaten bread.'
> >      (lit. 'I am from the eating of bread.';
> >       cf. Irish English 'I am after eating bread.')
> >
> > The prospective aspect expresses a state of being
> > about to do
> > something; it is formed with the allative case of
> > the VN:
> >
> > (5)  Baha matanthan bradal.
> >      be-PRS-1S:P eat-VN-ALL bread-LOC
> >      'I am about to eat bread.'
> >      (lit. 'I am to the eating of bread.')
>
>
> These strike me as being vaguelly Finnishlike - that
> is, similar to the usage of the various cases of the
> infinitive in Finnish.
Finnish is another language I don't know the details of;
if it does the same, the better.  Shows that my solution
is naturalistic.  Actually, I did not think of Finnish;
I simply thought about how constructions like "I am
after eating" work in a language with verbal nouns and
a robust case system.

>       What about a gerundive "X to be
> done"? In Silindion, this is formed by declining the
> gerund in the dative case:
>
>    miossi       i     siri              fantanu
>    this/these the things.nom  do.gerund.dative
>    "these are the things to be done..."
I haven't thought about a gerundive yet.

> > Case forms of the verbal noun also take the role of
> > participles:
> >
> > (6)  a chvana hatanthala (sas)
> >      the:C-AGT dog-AGT bite-VN-LOC-AGT (it-GEN)
> >      'the biting dog'
> >
> > (7)  a chvana hatanthada (sas)
> >      the:C-AGT dog-AGT bite-VN-ABL-AGT (it-GEN)
> >      'the dog that has bitten'
> >
> > (8)  a chvana hatanthana (sas)
> >      the:C-AGT dog-AGT bite-VN-ALL-AGT (it-GEN)
> >      'the dog that is about to bite'
> >
> > (9)  o ndaro hatanthala sol
> >      the:M-AGT man-AGT bite-VN-LOC-AGT he-PRT
> >      'the man that is bitten'
> >
> > (10) o ndaro hatanthada sol
> >      the:M-AGT man-AGT bite-VN-ABL-AGT he-PRT
> >      'the man that has been betten'
> >
> > (11) o ndaro hatanthana sol
> >      the:M-AGT man-AGT bite-VN-ALL-AGT he-PRT
> >      'the man about to be bitten'
>
> If I'm reading these right, the passive is expressed
> by declining the agreeing resumptive pronoun as if it
> were an object?
Yes.  In the active examples (6)-(8), the resumptive
pronoun is in the genitive because the head noun is the
agent of the VN; in the passive examples (9)-(11), it
is in the partitive bexause it is the patient.  The
agentive case markings in the examples have nothing
to do with the semantic role relative to the VN; it
is simply the citation form of animate nouns.  The
whole shebang can be in another case:

(17) Adanama am blaum on ndaron matanthon bradal
     AOR-give-3SG:P-1SG:A the:I flower the:M-DAT man-DAT
     eat-VN-M-DAT bread-LOC
     'I gave the flower to the man eating bread.'

> > The locative case of the verbal noun is also used in
> > a
> > converbial construction, expressing that something
> > happened
> > while something else was happening.
> >
> > (12) Matanthal mas bradal, lingena am dvarling.
> >      eat-VN-LOC 1SG-GEN bread-LOC ring-IPF-3SG:P
> > the:I-OBJ doorbell
> >      'As I ate bread, the doorbell rang.'
> >      (lit. 'In my eating of bread, the dorbell was
> > rung.')
> >
> > Similarly:
> >
> > (13) Matanthad mas bradal, lingena am dvarling.
> >      eat-VN-ABL 1SG-GEN bread-LOC ring-IPF-3SG:P
> > the:I-OBJ doorbell
> >      'After I ate bread, the doorbell rang.'
> >
> > And also:
> >
> > (14) Matanthan mas bradal, lingena am dvarling.
> >      eat-VN-ALL 1SG-GEN bread-LOC ring-IPF-3SG:P
> > the:I-OBJ doorbell
> >      'As I was about to eat bread, the doorbell
> > rang.'
> >
>
> How about the conjunctive passive?    "As he was about
> to be beaten...", etc.
(18) Hatanthan sol as chvanas, lingena am dvarling.
     bite-VN-ALL he-PRT the:C-GEN dog-GEN ring-IPF-3SG:P
     the:I-OBJ doorbell
     'As he was about to be bitten by the dog, the doorbell rang.'

Works of course also without the agent:

(19) Hatanthan sol, lingena am dvarling.
     bite-VN-ALL he-PRT ring-IPF-3SG:P the:I-OBJ doorbell
     'As he was about to be bitten, the doorbell rang.'

Accordingly with the other participial constructions.

... brought to you by the Weeping Elf