Re: Different Words with Large Common Substrings
From: | Alex Fink <000024@...> |
Date: | Monday, October 13, 2008, 20:51 |
On Mon, 13 Oct 2008 15:40:49 -0400, Eldin Raigmore
<eldin_raigmore@...> wrote:
>...>Inspired by this:
>
><
http://conlang.wikia.com/wiki/List_of_self-segregating_morphology_methods
>>
>
>|3 Alternatives
>|And Rosta's Livagian uses another method which, though not a self-
>|segregating morphology in the strict sense, partly serves the same
>|purpose with less restriction in the phonological shape of words. It
>|requires a full knowledge of the lexicon to parse unambiguously, however.
>|The key is that no actual morpheme must look like a prefix or suffix
>|substring of another actual morpheme.
Note that the "or" here has high scope. To obtain unambiguously parsable
utterances all you need is either that no word is a prefix or another, or
that no word is a suffix of another. In data compression and similar
contexts they call the first a "prefix code" among other names; the second I
suppose would be called a "suffix code" though I can't attest having heard
that.
E.g. the language whose three morphemes are "n" "\n" "\\" is a prefix code
and therefore is uniquely decodable = segregable into morphemes with
complete knowledge of the lexicon, despite that "n" is a suffix of "\n".
Prefix codes are, loosely, more important than suffix codes in practice (and
so a more sensible choice for an engelang) because an incomplete message is
a prefix of a complete one: that is, with a prefix code you can start
breaking a message into morphemes as it comes in, but not so with a suffix code.
>...>I have been considering rules roughly similar to:
>"No two distinct words can have a common initial substring which is longer
>than half the length of one of them and longer than one-third the lengths of
>the other; nor can any two distinct words have a common final substring
>which is longer than half as long as one and longer than one-third as long as
>the other."
Anyway, the real point of this message is to express my confusion. I don't
understand why you have been considering these rules, and thus your musings
don't make a lot of sense to me. What do you want to achieve -- some weaker
or stronger variation on self-segregating morphology?
Alex