Re: MNCL5 really long

* Jeffrey Jones said on 2007-12-25 09:54:33 +0100
> What the heck, maybe somebody will have a comment on one or
> two of these.
Optimist :) Most are busy with christmas I'd wager :)

Wow, long indeed, but it's *morphosyntax*! *swoon* Thx for the
christmas-gift :)

> MNCL5 has accumulated a large number of unsolved problems. I'm
> going to try to list all the current ones here, so they'll all
> be in one place.
>
> Note: "medial" is my term for "non-final suffix".
I'll be calling your lang "5" so as to avoid typos.

> 1. Negative Requests
> Currently, requests are made by combining the question medial
> -uk- with the imperative final -u. However, there's a rule
> prohibiting the use of the negative medial -un- in the same
> word as -uk-, requiring an awkward auxiliary + complement
> clause construction. Possible changes are (a) changing the
> rule, (b) using something other than -uk-, or (c) a distinct
> negative request medial.
Is there an underlying reason for preventing -un- and -uk- in
the same word:
- because they both contain {u}?
- because they'd have to occupy the same slot? (does 5 use slots?)
- semantic/pragmatic reasons?
- authorial fiat?

Finding the answer to that might lead to the answer you seek.

> 2. Argument Structure Conflicts
> In some situations, certain verbs need a different argument
> structure from what they currently have. For example, lez-
> "read" and rait- "write" could use a core argument for the
> person read to or written to,
a target/goal/benefactive?

I'd need to dig in a lot more to help with this 'un I'm
afraid...

> which would require that the existing patientive argument
> become thematic. Sail- "sing" has so far been treated as
> monovalent, but could use both/either a person sung to and
> something sung. This change would require an obligatory object
> for non-verb forms.
Dos 5 allow implicit default (fallback) objects, like: (3rd
person) animate for this set of verbs, (3rd person) inanimate for
this other set?

> Under the same heading, the morphemes for beneficial (ufr-)
> and harmful (ivl-) require that what/whom they're beneficial
> or harmful to be thematic and what's beneficial or harmful be
> patientive/agentive in order to be used as the benefactive and
> the malefactive. This conflicts with their potential use as
> judgemental adjectives, where what's beneficial or harmful
> must be thematic and the person making the judgement must be
> patientive (with no provision for what/whom they're beneficial
> or harmful to). Probably, I'll have to resort to (a)
> suppletion (if there aren't too many like these) or (b) an
> argument structure changing medial.
Have you tested that this is an unwanted ambiguity? With several
sentences, to see how much of a drawback changing nothing would
be?

> 3. Trivalent Imperatives
> Currently, a number of trivalent verbs have two translations,
> depending on either grammatical voice or phrase order. These
> include geb- "give/receive", vend- "sell/buy", and lern-
> "teach/learn". There's a problem, however, which of the two
> meanings is to be used with -u forms (these are primarily
> imperatives and requests, with implied 2nd person arguments,
> but may be used with 3rd person or 1st person inclusive).
> Possible solutions include (a) a new medial (like the
> grammatical voice medials, but used on -u forms instead of
> non-verb forms) such as -s- for the 2nd translation of each
> pair, with the unmarked form used for the 1st translation, and
> (b) allow the existing grammatical voice medial -m- to be used
> on -u forms to indicate the 1st translation, with the unmarked
> form used for the 2nd translation.
Or c), inject something from natural languages (trivalent verbs
are rare) and have each verb has a preferred imperative that
fits with that verb's meaning, and an inverse marker if you want
to preserve the oppsite possibility.

> 4. Habitual and Frequentive
> The question here is how to determine whether adverbial
> expressions such as "often" or those specifying how many times
> refer to a single occasion (frequentive) or multiple occasions
> (habitual).
Why adverbs instead of TAM?

> Related to this is under which circumstances can the habitual
> medial (-oft-) or the frequentive medial (-ebd-) be omitted?
Adverbs means non-affix in my world...

This also depends on the prototypical meaning of the verb: to
chop firewood, you don't chop just once, and it's commonly done
habitually... Can you have both freq. and hab. marker on the
same verb?

> 5. Conatives
> This deals with the morphosyntax used for expressing something
> like "try to", where either success hasn't been determined or
> the attempt has failed.i
Wouldn't "try but definitely fail" be something else?

> Since the person trying is already referred to by one of the
> existing arguments, a medial, rather than an auxiliary, might
> be used.
>
> Another question is for which TAM combinations has the attempt
> failed and for which has success not yet been determined?
> Complicating this is the fact that tense marking is relative
> rather than absolute.
You really need to Venn-diagram your verbs to get an overview of
the types. Try drawing the semantic map for English TAM to see
exactly where 5 differs.

Reference: http://www.ling.canterbury.ac.nz/documents/Adams.pdf

Where'd I put the copy of the adverb/adjective/complement
semantic map prototype...

> 6. Spatial Relations
> I'm trying to figure what spatial relations I need distinct
> words for; I have three kinds so far: (a) directions such as
> north, south, etc., (b) those relating to an object's
> structure, such as front and back, and (c) those relative to
> someone's viewpoint. An additional problem with the last kind
> is that there is a 2nd object specifying whose viewpoint is
> used.
What meaning of object are you using here? Unclear. Do you
operate with default viewpoints? Do all things have the same
dimension or do some (like a road) just have a subset? Said in a
different manner: Do you map from the viewpoint to the thing
observed (possibly with a transform) or does the thing observed
have inherent left, right, up, length, number,
togetherness/scatteredness, etc...

> 7. Tetravalent Verbs
> Obviously, I'll have to use an adverbial or secondary
> predicate final (-i or -in) for the 4th argument. The main
> question is which?
"bet" is 5-valent, which 4-valent verbs are there in 5?

> 8. Adverbs
> I need to figure out the different kinds of adverbs that must
> be distinguished (such as manner, instrument, time when, how
> long, etc.) so that I can assign medials to them, and decide
> which can be used without any such medial.
You are aware of a difference between langs as close as English
and French: whether manner/path is adverb(ial) or verb?

French/romance style: "The branch seawardsmoved floatingly"
English style: "The branch floated towards the sea"

> 9. Partitive and Superlative Constructions
> I came up with some morphosyntactical possibilities for the
> partitive construction, but I don't like them.
What's the semantics of the partitive? It's too vague a word to
just bandy about. Maybe there are several partitives, each with
its own best solution?

> In some of my other languages, the superlative construction is
> formed by adding a simple (lexical) adjective to the partitive
> construction. I'm not sure that will work here.
Maybe superlative is not right for this language. Not all langs
have it, after all. How would you do without?

> 10. Compound Phrases
> What I mean is expressions such as "John and Tom". I'm
> thinking of creating a new final (maybe -al) to be used for
> "and" on each subphrase but the last. The same final might
> also be used in compound numbers.
Compound number: higher than the base? number+fraction?

> That's all I've found in my notes, except for a neverending
> list of lexical decisions.
Those, like the poor, you'll always have with you...


t.