Date: Thu, 24 Feb 1994 14:09:31 EST From: Larry Horn Subject: croissants In response to Dennis's post, I'd figure that the expectation that 'croissant'-->[kroysaent] would be based on either spelling or (a reversal of) history, in the sense that the [oy] diphthong represents the earlier stage of what developed into MFr. [wa] (around the Revolution?). But in fact this particular change doesn't strike me as being plausibly reversible, and in any case from a phonological (rather than orthographic) point of view, his attested [krusaent] is a better representation of the source. Or the other half of the diphthong could be retained, and indeed I've heard something like [kr[AT SYMBOL GOES HERE]SANT] even more frequently than the previously noted simplification. \ schwa Larry