Along the lines of the apparent legitimacy of words like "preventative"..... I find myself running across more and more bits of language usage that seem strange to me -- it seems the language is morphing exponentially (along with everything else) as web-use becomes primary in so many areas and ways (people advertising items on-line don't even make a cursory attempt at giving general locations -- they merely post the physical address and know that "map this" will do the rest). Examples -- "disambiguation" -- <<Disambiguation in Wikipedia is the process of resolving conflicts in article titles that occur when a single term can be associated with more than one topic, making that term likely to be the natural title for more than one article. In other words, disambiguations are paths leading to different article pages which could, in principle, have the same title.>>. Is this Wikipedia coining a term, which will become part of general usage? Or sloppy useage which becomes legitimate over time? (I'm thinking the variations of "prevent", etc. fall into this category.) "unrational" -- the most recent example I've come across in what seems to be a general move to remove (how d'ya like that one?) the prefix "a" -- as in "arational", "amoral", etc. (My auto-spell-check led me to this on "arational" -- <<In Lee Mosher's talk at U. of Chicago on 2/27/03, he gave an example demonstrating how to tell if a lamination is arational. One way to define arational is that the lamination meets every simple closed curve essentially. On the torus, these correspond to measured laminations of irrational slope, so are a generalization of irrational numbers. A lamination may be specified by a sequence of splittings of train tracks, where each train track is some coarse view of the lamination, where one can resolve distinct leaves only up to a certain scale, after which leaves merge together at branches, called cusps. Mosher discussed a particular kind of sequence of splittings, where one has a distinguished cusp, marked by ∗. One splits the train track sequentially at the cusp ∗, and each splitting is specified by an L or R, depending on whether the branch splits to the left or right (see figure 1). When the two cusps agree, there is only one way to split, so we do not need to record it. If the train track fully carries a lamination (meaning that there is a smooth homotopy of the lamination into the train track)*, then this sequence of splittings is uniquely determined by the lamination. Thus, a sequence LRLLRRRL... defines a unique sequence of splittings of the train track along the distinguished cusp.>> (* That certainly cleared THAT up! I wouldn't have necessarily thought immediately of a smooth homotopy of lamination...perhaps lamentation, but not lamination...I wonder what the implications would be if one were to laminatate the train track? Someone please implicatate that into an implicatature.) Who knew? That, then led me to the most wonderful phrase "-i is also -1. I think you may begin to see why my dishes aren't done...). Julie Krueger Flinging syllables to the wind, to see who catches them.