To go back to the typecasting metaphor, the idea behind typecasting is
that it represents the identity operation.  The copula, as it were.
Sure, there are infinitely many functions that receive a character
string as input and return a floating point number as output, but the
"obvious" one that needs no qualification is the one that returns the
"value" of the string - presumably the one it  would represent if
entered as a numeric literal in the source code.

This is not always clear-cut, and is therefore a good metaphor for the
sort of munging under discussion.  The regular part-of-speech endings
in E-o inspired me to do likewise in my early conlangs...  e.g. all
nouns in Shalakar end in "-wa"... and I very quickly ran up against
the problem of defining a general rule for deriving the meaning of a
trans-part-of-speeched root.  There is no such rule that is wholly
satisfactory, IMO.

On 8/10/08, Dana Nutter <[log in to unmask]> wrote:
>> [mailto:[log in to unmask]] On Behalf Of
> caeruleancentaur
>> > <Liase> is quite interesting, since <-on> isn't (AFAIK) a
>> > derivational affix in English, such that it could be removed
> from
>> > the word, the way that <-ance> is. I guess that's similar to
>> > <-aholic> from <alcoholic>. I am trying to think of other
> examples
>> > of that same thing happening, but they aren't coming to me.
>> These come to my mind:
>> hamburger > cheeseburger (named after the German city
> Cheeseburg)
>> execute > electrocute
> Or "-gate" (< Watergate) for scandals like "Billygate" or
> "Contragate".

Sent from Gmail for mobile |

Mark J. Reed <[log in to unmask]>