Re: Infinitives & gerunds

On Thu, Oct 05, 2006 at 10:39:20AM +0100, R A Brown wrote:
> H. S. Teoh wrote:
> >On Wed, Oct 04, 2006 at 07:46:52PM +0100, R A Brown wrote:
[...]
> >>(Thinks: I suppose the subject line should be changed again)
> > So the original subject line was appropriate after all. :-)
>
> Possibly - if TF's gerund(ive)s aren't really infinitives in disguise   :)
That's true...


[...]
> >>Now does TF really have gerunds (which are also verbal nouns)? Or is
> >>'gerundive' really a better term after all? Or are they something else
> >>:)
> >Well, I'll let you be the judge. What I currently call gerunds (or
> >"gerundives", but I think you're right that it's not a good term for
> >them) work as follows:
> >1) As a nominal verb (like the English gerund):
> >	hapasi    nijibin   so   sianas   baibai.
> >	(harm-GND RCP:child CVY) shameful COMPL
> >	Harming a child (or, the harming of a child) is shameful.
> Latin:  puerum laedere est foedum
>
> >2) As reported action:
> >	tara' kei tsana huu na  ka'ami  nitiki     ipuresi    so  aniin.
> >	3sp   ORG speak 1sp RCP eat:GND RCP:rabbit CVY:carrot CVY COMPL
> >	She tells me that the rabbit eats a carrot.
> >	(Lit., she tells me of the eating of a carrot by the rabbit.)
> Latin:  illa mihi dicit cuniculum carotam esse
>         /e:sse/ "to eat" - not /esse/ "to be"!!!
That's interesting... is it a common IE phenomenon that the root for 'to
eat' and 'to be' have many close, or even lookalike forms? For example,
in Russian есть is both the 3ps form of "to be", and the infinitive form
of "to eat". Context is required to disambiguate between them.


> >3) As a nominalized clause (syntactically identical to (2)):
> >	huu na  hamra hapasi    nijibin   ko   aram.
> >	1sp RCP see   (harm-GND RCP:child ORG) COMPL
> >	I see a child being harmed (or, I see the harming of a child).
> Latin:  uideo puerum laedi [passive infinitive]
>
> Note: 'puer' in the earliest Latin was an epicene noun.
>
> There are all nominalized clauses, aren't they? Certainly they must all
> be expressed in Latin using an infinitive; a gerund is not possible for
> any of those examples.
Hmm. Now I'm confused about the difference between a gerund and an
infinitive... Surely (1) at least would be a gerund?


> >4) As a purpose clause (similar to using a subordinative verb):
> >	diru kei anai    na'iat sei hamra'i be  asamat  nijibin   utu eimei.
> >	girl ORG hold_up cloth  CVY see:GND NEG ORG:man RCP:child for COMPL
> >	The girl holds up the cloth so that the man would not see the child.
> >	(Lit., "... for the purpose of the not seeing of the child by
> >	the man.")
>
> Here Classical Latin does not use an infinitive (tho Medieval Latin
> does so quite happily). But the problem as far as the classical
> language was concerned is that (a) we will need some preposition
> before the verbal noun to indicate purpose, and (b) Latin does not
> allow prepositions to govern infinitives - it has to use gerund or
> gerundive construction instead. *But that is a peculiarity of Latin*.
> The Romance language happily put prepositions before infinitives. In
> French 'pour' (for)+ infinitive is a common way of expressing purpose.
>
> Yes, I have said in other mails in the original thread that where
> there is only one (type of) verbal noun it is preferable IMO to call
> it an infinitive. I would reserve 'gerund' to describe a second type
> of verbal noun if, like English & Latin, it has a different formation
> quite distinct from the infinitives.
Ah, OK. So maybe the TF "gerunds" should really be called infinitives?


> Yes, I think TF's gerund(ive)s are really infinitives   :)
Hmm. OK. I'm still stuck on the point that infinitives have no time
element associated with them, whereas in TF, esp. in reported speech, a
time element could certainly be present.


[...]
> >>But I must say I am enjoying this thread. TF is such an unusual
> >>language in many respects. Yet it does seem to work. I would dearly
> >>love to have it spoken by a community of people to see how it worked
> >>out in practice - but I guess that's not likely to happen {sigh}
> >[...]
> >Well, if we could convince enough people to learn and master TF ...
> >:-) (One can dream, right?)
>
> I think to test fully one needs a L1 TF population - not likely, but
> indeed one can dream   ;)
Ah, so if I speak TF to my children, and they speak TF to *their*
children, and ... Yeah, not likely.


[...]
> >Seriously, though, although there are many things I like about TF,
> >there are also places where I feel like the seams are stretching
> >uncomfortably.
>
> I think this is almost bound to happen if you depart from the familiar
> grammatical structures of natlangs. But this is IMO one of the
> exciting things about conlanging, one can try out original structures
> - even if they often turn out to be ANDEWisms.
>
> >The prospect of an overly long relative clause sitting between a noun
> >and its case particle seems to be pushing the limit of how much one
> >needs to keep in mind at any one time --- especially when so much
> >semantic load is dependent on that last little word.
>
> One could move the subordinative verb to the front of its clause?
Actually, I was referring to the distance between the head noun and its
associated case particle. I think the overt use of secondary case
inflections for the subordinative verb's arguments already adequately
prepares the listener to expect a subordinative verb coming at the end.
For the head noun, though, one has to keep it "on hold", not knowing its
case function, until we reach the end of the phrase, even though in the
middle of it the case functions of the subordinate clause's arguments
are already made clear.


> >It could very well turn out to be a postposition instead, which would
> >significantly change the interpretation of the whole phrase. Too much
> >of this may lead to repeated cognitive clash at the end of each NP. I
> >think Tagalog had it right with putting case markers in front of the
> >noun. :-)
>
> Now that's another interesting language! Does it actually have
> (finite) verbs? But we have, I think, discussed that before on
> Conlang.
[...]

I don't remember what the consensus was... although I do remember
reading an interesting analysis on some website somewhere that proposes
that Tagalog really does not distinguish between noun and verb at the
root level, and therefore has to use derivational morphology to turn a
root into a noun or a verb.


T

--
Don't drink and derive. Alcohol and algebra don't mix.

Replies