Lojban (was: Verbs derived from noun cases)

Henrik Theiling scripsit:

> Speaking of Lojban, I have a question: how did people assign the
> argument structure to verbs, i.e, why does 'go/come' have these and
> not more/less?  E.g. more: 'reason', less: 'means of going'?  And a
> curious sub-question: What's the maximal valence and why?
The maximal valence for morphologically primitive verbs is five, but
morphologically derived verbs can have any number.  There is no
particular reason why five should be maximal: it just works out that
way.  There are only 18 primitive verbs that actually have five places:

	transfer(agent, object, receiver, transmitter, means)
	carry(agent, cargo, destination, origin, path)
	be_wave(wave, medium, waveform, wavelength, frequency)
	learn(learner, what_is_learned, subject, source, method)
	be_school(school, location, subject, student_body, operator)
	teach(teacher, audience, content, subject, method)
	be_book(physical_object, content, author, audience, medium)
	be_drama(drama, theme, dramatist, audience, actors)
	translate(translator, source, target_language, source_language, result)
	be_a_law(law, prescription, community, conditions, lawgiver)
	compare(observer, comparand_1, comparand_2, property, result)
	come_go(agent, destination, origin, route, means)
	measure(agent, quantity, measurement, scale, accuracy)
	send_mail(agent, object, recipient, sender, network)
	pump(pump, fluid, destination, origin, means)
	leap(agent, destination, origin, max_height, propellant)
	be_umbrella(umbrella, protectee, noxious_influence, material, support)
	be_phrasal_compound(compound, modifier, modificand, meaning, instance)

The assignment of arguments to each verb is a compromise between
four pressures: brevity, convenience, metaphysical necessity, and
regularity.  See http://www.lojban.org/files/reference-grammar/chap12.html#s16

> I take it that additional adjuncts are possible (like 'because of').
Yes.  Indeed, Lojban dissects cause into four separate relationships:
"physically caused by", "logically entailed by", "motivated by", and
"justified by".

> But why is 'means of going' a core argument to 'go/come' and not
> realised as an adjunct?  What's the criterion for being a core
> argument?
See above.  Means is included as a result of the pressure toward
metaphysical necessity:  a going without a means is not really a
going at all.  (For specialized purposes, there is a defined
mechanism for creating verbs of lower valency, but the more usual
tactic is simply to allow the argument to default.)

> And apart from that, in what way are the arguments ordered?  If I
> understand it correctly, the cases are distinguished by position and
> when skipping a position for a case, special word has to be used (and
> a tail of these words at the end of phrases may be left out).  E.g.
>
>   go/come I Paris SKIP Metz.
>   'I went to Paris via Metz.'
>
> Is that correct?  (I took the order from the list on your posting).
Positional order is the dominant mechanism, backed up by particles
that give explicit case information, which can be used to alter the
mechanism.  The default is x1 verb x2 x3 ..., a variant of SVO, so

	mi klama la paris. zo'e la mets.
	I go/come the Paris (skip) the Metz

is the unmarked way of saying the above.  To move the goer after the
verb, the particle "fa" meaning "first argument" must be prefixed:

	klama fa mi la paris. zo'e la mets.

> How's that order defined?  Why is 'destination' after 'origin'?  Or is
> it chaotically lexicalised (unlikely for Lojban, I think).
Some regularities are visible (actors tend to come first, destinations
are always before (not after) origins, and so on), but argument orders,
like argument meanings, must ultimately be either memorized or intuited.
Fortunately, the morphologically underived verbs are limited to about 1500,
many of which (the more noun-y ones like "be a bear", e.g.) fall into
fairly regular categories.  Derived verbs have argument structures and
orderings that are fairly systematically derivable from the underived
verbs, although not rigidly so.

--
John Cowan                              jcowan@reutershealth.com
http://www.ccil.org/~cowan              http://www.reutershealth.com
Thor Heyerdahl recounts his attempt to prove Rudyard Kipling's theory
that the mongoose first came to India on a raft from Polynesia.
        --blurb for Rikki-Kon-Tiki-Tavi