Re: Ambiguities (was Re: Vocab 2.4)
From: | H. S. Teoh <hsteoh@...> |
Date: | Tuesday, November 12, 2002, 2:14 |
On Mon, Nov 11, 2002 at 07:54:06PM -0500, Roger Mills wrote:
> All due respect to Aidan-- here are some sentences that are considerably
> more ambiguous; all, of course, depend on context--
In Ebisedian, there is considerably less room for ambiguities.
> 1. She gave him her picture =
> a. Mary gave John a picture of herself.
In Ebisedian, "noun association tags" would be prefixed to each pronoun.
Also, reflexive references are explicitly marked. In (a), the Ebisedian
would render it as:
jhit0' byy'jh chi'du n3 s0lu d3 <picture><cvy>
she(org) give him(rcp) - herself(rcp) -
(No word for "picture" yet, so I'm just transliterating it.) In this
particular example, although the reflexive particle _s0lu_ does not
inflect for gender, it can only refer to "she" because of case agreement
(originative-originative).
> b. Mary gave John a picture of Sonia
kijhit0' byy'jh chi'du n3 cujitu d3 <picture><cvy>
<ki>she(org) give him(rcp) - <cu>her(rcp) -
Here, the prefixes ki- and cu- are the associative prefixes. Associative
prefixes are attached to pronouns to disambiguate them. They are
"assigned" by previous mention; e.g.,
`ykimari' kijhiti'. kijhit0' tww'ma cujhitu'
<ki>Mary [is] she. <ki>She speak <cu>her
"<ki>she is <ki>Mary. <ki>she(Mary) says to <cu>her(someone else) ..."
Here, the associative tag <ki> appears as an infix. The Ebisedian
equivalent of "Mary" is _`ymari'_ [Hyma"r`i]. In the second sentence, the
second pronoun is tagged differently; and hence it refers to someone else.
> c. Mary gave John a picture that Mary owned. (implies she only owns one!)
In this case, there is ambiguity, as this would be identical to (a).
> d, Mary gave John a picture that Sonia owned (ditto)
This would be identical to (b).
> e.. Mary gave John a picture that Mary had drawn
> f.. Mary gave John a picture that Sonia had drawn.
> (Also depends on the polysemy of "picture")
Ebisedian probably would not exhibit polysemy for "picture", so this
doesn't really apply here.
> And I think the all-time winner, so n-ambiguous that I lose count:
> "The police were ordered to stop drinking on campus after midnight."
In Ebisedian, this would be quite unambigous:
1. Police(rcp) ordered <t3> myso' drink(v) campus(loc)
after-midnight(loc) <t3m3>
The <t3> and <t3m3> are subordinating particles that delimit exact
boundaries for the subordinate clause. So (1) implies that the order was
for them (the police) to stop drinking on campus after midnight. _myso'_
is a negative optative particle meaning "let it not be so". It implies an
order directed at the listeners (the police themselves).
2. Police(rcp) ordered <t3> stop(v) drinking(cvy) campus(loc)
after-midnight(loc) <t3m3>
Here, the verb stop (_p0'ju_) is an imperative, and "drinking" is a
gerund. Hence, it means for them to interrupt the act of drinking (by
others, although it does not exclude the police themselves).
3. Police(rcp) ordered <t3> stop(v) drink(cvy) campus(loc) <t3m3>
after-midnight(loc)
The order is that for the police to interrupt the act of drinking on
campus; and the order was given after midnight.
4. Police(rcp) ordered <t3> myso' drinking <t3m3> campus(loc)
after-midnight(loc)
The order is that the police stop drinking, and the order was given on
campus after midnight.
Note, of course, that this is stretching Ebisedian grammar a bit.
Generally, Ebisedian dislikes long, complex sentences like this. In this
particular case, it would like to break up the sentence depending on the
intended semantics:
1) The police were ordered to stop (their own) drinking. In this case, the
sentence as it stands is perfectly acceptable, with _myso'_, the
prohibitive, replacing the English verb "stop". Literally, it would
read something like "let it not be so, that [you] drink".
2) The police were ordered to stop others' drinking. Although the versions
I give above are plausible, a native speaker would paraphrase the
sentence as:
If you(pl,police) see(v) [ drinking(instr) who ] person,
then stop the person(cvy).
In orthography,
ana 3co'miu fww't3 n0 <drinking,instr> du bis33'di.
keve p0'ju b3s33'd3.
Ebisedian does not distinguish between direct and indirect discourse; so
this can be transplanted into the main sentence:
Tww'ma 3co'miu t3
ana 3co'miu fww't3 n0 <drinking,instr> du bis33'di.
keve p0'ju b3s33'd3.
t3m3.
"It was asserted to the-close-ones[*] [ If the-close-ones see
the-drinking-person. Then stop person. ]"
I formatted this a bit creatively so that you can see the nesting enclosed
by t3...t3m3.
[*] Raw transliteration of the Ebisedian pronoun. Here, I'm substituting a
pronoun for "police", since I don't have a word for "police" yet. :-)
> Kash at least can disambiguate this a little (litterally, without bothering
> to translate):
[snip]
> The disadvantages of lacking a passive, and the two adverbial clauses create
> a raft of other problems.........
*Tries to resist urge to boast about Ebisedian's active/passive-less
grammar* ;-)
T
--
People say I'm indecisive, but I'm not sure about that. -- YHL, CONLANG