| xorxes’ usage is just that; it is not official Lojbangrammar nor semantics. In the latter,as worked out over the years on this andpredecessor lists, {lo’e} is indeedfor the average whatsit, and thus not what is typically being hunted – anywhatsit will do,even a very untypical one, and {lo’e broda} is by rule not tobe treated as an individual but as a fictive stand in for a more complex(hopelessly so, so far as expanding it is concerned) locution. By the samehistory, {tu’a x} is short for {le nu x co’e} or something like it, showingthat the x is not on the surface of the sentence butembedded a layer down, notin the world of direct reference but an indirect one – which may, depending onthe brivla to which it is subordinated, not have any veridical connection tothe world of direct reference. We can oftenignore this fact, resulting only in anomalies of sumti restrictions (or takingthem as implicit) but whenever the possibility of logical error arises, we haveto fall back (in a logical language) on explicitness. Note, by the way, that we do not have a special grammarcategory for the concept of average, but rather only a special member of thecategory gadri for a particular way oftalking of certain statistical information. That way is (like the x in {tu’a x} ) not open to quantification nor toclaims that it exists as an individual, so xorxes’ usage is at leastanalogical, though inaccurate. Ofcourse, none of this applies to {sisku}, which got defined in this messy way inan earlier attempt to avoid the same problem that {tu’a} finally solve moregenerally, and probably should be moved back to something more natural, sinceit is hard to say what one wants now. The temptationis always to move to some other brivla in these cases. Hope we are not going to get bogged down in this one again. |