Re: Linguistic knowledge and conlanging (was Explaining linguistic...)
From: | Gary Shannon <fiziwig@...> |
Date: | Sunday, July 25, 2004, 5:52 |
--- "Mark P. Line" <mark@...> wrote:
> Gary Shannon said:
> >
> > Here's the process I'm using for my present
> > experiment. I choose a piece of text and just
> write
> > the translation. Since this is the first piece of
> > text nothing is known about the language yet, so
> the
> > translation can look like anything; just a string
> of
> > made-up words.
> > [snip]
> > I am also free to refer to the existing corpus
> > for examples of sentence structure,
>
>
> How do you know what the sentence structures are?
>
>
> -- Mark
>
The same way a four-year-old knows the sentence
structure of his native language; by unconsciously
generalizing from previous examples, or inventing a
structure when no applicable previous examples can be
recalled. It's a matter of "training the ear" on the
corpus examples and then making utterences that "sound
right" based on that ear training.
Of course this can result in sentences like "Him gived
it to me the book." But that's fine. Such sentences
would be perfectly acceptable in the formative stages
of the language simply because there are, as yet, no
alternate forms in the corpus to choose from.
Very chaotic. Very natural.
--gary
Reply