I’ve been interested for a while in vague terms that don’t generate Sorites arguments. I hadn’t had much success in coming up with precise terms that do generate Sorites arguments. I was convinced by discussions with Josh Parsons that such terms must exist, but I could never come up with one on my own. The next best thing to discovering one is finding a paper in which one appears. The following is from a paper by “Gerald Hull”:http://home.stny.rr.com/ghull/ on “Vagueness and ‘Vague’: A Reply to Varzi”:http://home.stny.rr.com/ghull/site/mind2005.pdf, soon to appear in _Mind_.

bq.. Let ‘approximately n’ be defined as ‘n plus or minus a tiny amount’. We can now construct the following sorites paradox:

Approximately 0 is less than 1000.
If approximately n is less than 1000, then approximately n+1 is less than 1000.
Therefore, approximately 10000 is less than 1000.

Clearly the first premise is true and the conclusion false. In between there are values, e.g. approximately 1000, that result in borderline cases.

p. You’ll have to read the whole thing to see the point Hull wants to make with this, but I think it’s a perfectly successful example of a Sorites argument where the major predicate is _not_ vague.