# Bootstrapping a cooperative conlang - theory vs practice

From: | Gary Shannon <fiziwig@...> |

Date: | Monday, November 19, 2007, 20:47 |

--- Mr Veoler <veoler@...> wrote:
> MorphemeAddict wrote:
> > AW gives the following (partial) explication for red:
> >
> > X is red. =
> > when one sees things like X one can think of fire
> > when one sees things like X one can think of blood
>
> [...]
>
> > Notice that there is no mention of the word "color". Naturally, the words
> > used above that are not primes have to be explicated too.
>
> This is one part where I expand my core vocabulary: adding various hypernyms,
> such as a word for "color". That makes it easier to define words as "red",
> "blue", "green" etc, when there are a word for "color" to use.
>
> -- Veoler
>

I am reminded of an article I read many years ago about the theoretical minimum
number of different kinds of machine language computer instructions necessary
to carry out any operation. As I recall, every possible computer instruction
can be built up from just the "subtract" instruction. You add 3 + 4 by starting
with 3, then subtracting 4 from 0 to get -4 and then subtracting -4 from 3 to
get 7. You jump to a new program instruction location (branch) by subtracting
some constant from the internal program counter register, etc.
Theoretically interesting, but myself, I'd rather be able to say "A =
sqrt(3*B+C)/9;" than to have to write out the 1852 subtract instructions it
would take to get the same result!
I'm beginning to think that NSM is the "Turing machine" of linguistics.
Interesting things can be proven with it, but it is of little or no practical
value in everyday use.
--gary