Re: Mapwords

En réponse à Peter Bleackley :

>The recent discussion of parts of speech has inspired this idea for a very
>weird one.
>
>A mapword is a word whose entire purpose is to define the grammatical
>structure of a sentence. It is a polysynthetic compound of particles, each
>morpheme corresponding to the function, role and gramatical relations of
>the words following it. Each sentence begins with such a monstrosity, the
>rest of the sentence consisting of isolating semantic words which are its
>arguments. Here's an example (in English gloss).
>
>n-pat.adj-attrib-pat-sup.vb-pt.adj-attrib-agt-comp.n-agt dog big buy small boy
>
>The smaller boy bought the biggest dog.
>
>Word order is simply mapword : everything else.
>
>Of course, when you start using subclauses things can get seriously
>complicated.
>
>Any thoughts?
This is an interesting idea, which is not completely unknown (I think), but
I've never seen it used up to this extent! I think it's something which is
at the frontier of being humanly possible (it stretches at the limits of
what the human memory can do. Also, it obliges you to define the whole
sentence before you pronounce it, since you have to give the grammatical
relations first - of course, if you choose everything else : mapword word
order instead, you would get this freedom again, but memory would be
rapidly a problem, especially if subclauses are involved -). But then, I've
seen a natlang marking the function of nouns in a noun phrase at the
opposite end of that phrase! (meaning that in such a language, a noun
phrase like "the woman's husband's house" is rendered "woman2 husband1
house of1 of2" where figures indicate what function word is connected to
what noun!)

As for subclauses, I don't think it would be a problem. Define each clause
to have a mapword, and the function of each subclause is defined in the
mapword of the main clause. If you have mapword : everything else word
order, you will begin the sentence with the main clause mapword, which will
indicate everything, including the function and position of each subclause.
At the position of the beginning of a subclause, you just have to insert
the mapword of that subclause, which acts also as a marker for the
beginning of that subclause, and by giving the number of arguments in the
subclause, indicates also unambiguously where it ends. This use is the best
I can think of. It's best to make one mapword per clause. Making a single
mapword for a full sentence would be just humanly impossible.

So subclauses are not the main problem. Human memory may be one :))
(sentences work mainly with your immediate memory, which can handle around
7 items at a time at most. That's why in speech, long subclauses are never
embedded inside the main clause. By the end of the subclause, you have
forgotten what the main clause was about :)) ). I guess this pragmatic
consideration would limit a little the use of such a construction (giving
pragmatic limits to an otherwise completely free word order). However,
poetry and literature (which are not dependent on this pragmatic
consideration) could use such a structure for very nice effects (with my
proposal with inverse word order from yours, you could get very nice
effects by introducing words in some order that people would tend to
translate in advance into a meaningful sentence according to pragmatic
considerations, and then create surprise by giving a mapword showing that
this interpretation is completely wrong! :)) . It would be the effect of
saying "dog bite man", leading people to think - naturally, since it's what
usually happens - that a dog bit a man, and then surprise them with the
mapword "n-pat.v-past.n-agt"!!!! Could be very useful for jokes, poetry,
and political speeches :))))) ).

Christophe Grandsire.

http://rainbow.conlang.free.fr

You need a straight mind to invent a twisted conlang.