On 2/28/07, Wayne S. Rossi <[log in to unmask]> wrote:

Hello, Wayne!

I agree with the "Uncanny Valley hipothesis", but with some
reservations. Firstly, sometimes a word is off-putting to a speaker
because he doesn't know other languages in which the word behaves that
way. I'll use some examples you've mentioned in order to illustrate

> When reading Occidental, for instance, words like
> "naturalmen" and "presc" and "ultim" look and feel
> inappropriate, because in the context they look like
> they should be "naturalmente" and "presque" and
> "ultima".

In the case of "naturalmen", it wouldn't be off-putting for someone
that knows a bit of French. IIRC, in French, the "t" of "naturalement"
is not pronounced ("-ment" is pronounced somewhat like "*mun" in
English). I believe that the "-men" suffix was based on a behavior
that de Wahl observed in spoken French. I am pretty sure that, if my
explanation was correct, your reservation against "naturalmen" in
Occidental will disappear because from now on you will find it more
"natural". This is the expected psychological effect of the Principle
of Precedent in Natural Languages. :)

> the plural
> present indicative of the Interlingua copula ("illes
> esse," often switched to "illes son");

If you were Brazilian, you wouldn't find "illes es" strange. In
Standard Portuguese, the "official" way is "eles são", but in the
Portuguese variant spoken in my region, we often say "eles é". Since
you probably does not speak Brazilian Portuguese natively, I can
completely understand your reservation against "illes es".

The lesson is that the psychological effect of "unnaturality" occurs
when the person is in a context that does not have contact with a
language in which that phenomenon is natural. As language engineers
are humans, and we know humans often make mistakes, we tend to find
that the language engineer inserted by "shortsightedness" an
"unnatural" element in the language which would be better replaced by
a more "natural" one. But often the fact is that the language engineer
indeed adopted a feature which has a precedent in a natlang, it is
just that we didn't know it before.

BTW, it reminds me of "k" in Esperanto, which has a precedent in
Polish, language which certainly was spoken fluently by Z. :) Ask some
Brazilians how natural they find a language that "replaces 'k' for
'c'". Most of them have never seen a word written in Polish, so most
of them will probably find the Esperanto solution of 'k' "unnatural".
I found it before I knew about Polish. ;)

But all this comment does not invalidate your point of "Uncanny
Valey"; it just complements it with the explanation of a behavior we,
language engineers, should be aware of.

Antonielly Garcia Rodrigues