> you were shewing the effects of the patterns /in combination > with values of \LHM and \RHM that were different for the two > dialects/ on those different non-words. Yes, that's exactly what I wrote earlier. Arthur