On 4/18/06, Dana Nutter <[log in to unmask]> wrote:
> > [mailto:[log in to unmask]] On Behalf Of Jim Henry

> > On 4/17/06, Dana Nutter <[log in to unmask]> wrote:

> > > Roots are not created that may conflict with existing root+suffix
> > > patterns to help avoid misinterpretation and minimize the chances of
> > > having a "sukero" problem.

> > Does this mean that no morpheme in Sasxsek contains a
> > prefix or suffix substring that looks like another real morpheme?
> > Or does it just mean that such substring matches are few?

> Substring matches (ex: "saf" and "safat") are still possible.  Suffix
> conflicts are specifically avoided.  Prefixes don't exist.
> For example, I can have a word like "finin" ("endless, infinite" <-
> fin=end, -in=no, non, un-, -less, etc.) which means I will not make any
> more roots which begin with "fin+{any suffix}" making it impossible to
> have another root containing "finin".

So you have "saf" and "safat" -- but can I infer that there is
no "-at" suffix that could cause real ambiguity there?
And do you avoid ending any roots with "-in" and all
other actually used suffixes?  Have you ever had conflicts
with old roots containing a substring match for a newly
introduced suffix?  If so did you remake the old roots
to eliminate such conflicts?

Jim Henry