I've become accustomed to referring to - - coherent logic, with connectives true, /\, false, \/, = and (exists); and - geometric logic that also admits infinitary \/. I think I got this sense of "geometric" from Mike Fourman, and then "coherent" is natural because coherent theories are classified by coherent toposes. But the published works vary considerably. In particular, where I have coherent geometric, Makkai and Reyes have finitary coherent coherent, Johnstone has finitary geometric generalized geometric MacLane and Moerdijk have geometric (not referred to). Is usage as chaotic as it appears? Steve Vickers. ==============================================================================