| Since I feel somewhat responsisble for this mess, introducing "superfective"
from a draft of my dissertation, I think I should shoulder up the old job of summarizing where thing got to last time. First a bit of background: the idea of a superfective is that an event continues after a "natural end," where that is contextually defined. For precesses, where there is a natural end, completion, what continues is not (usually, at least) the process (which can't continue) but some related essential activity: "He has run the mile but he keeps on running." For other kinds of events, end is even more arbitrary and may need explicit stating somewhere: "He's eighty-five, but he keeps on running" "I've put on three coats of white paint, but it's still blue." Or it may be implicit but clear "He just keeps on talking" (long after I have stopped being able to listen). As these examples suggest, one of the expressions that goes with the superfective is "still", as well as "overlong," "excessive." and the like ("too much" as opposed to the "just right" of {mo'u}, and like those factors depending upon some unstated standard or purpose). This suggests a "too little" ~ "not enough," stopping before the {mo'u} is reached. Although strictly {co'u} marks any end and {mo'u} only makes separate sense for processes, Gricean forces have taken {co'u} to mean "before {mo'u}," since saying {mo'u} if it applied would be fuller information (but this does ignore the possibility that one does not know what the status was at the stop). So, {co'u} took on the "too little" sense, "quit too soon," and {mo'u} expanded to cover non-processes when there was a possibility of the notion of "enough" applying. This led to the question about the other end of the event, the beginning, where there were not two notions, corresponding to {mo'u} and {co'u} but there was definitely the possibility of starting too soon by some external standard or too late ("already" and "not yet"). Under one possible way of understanding what the Book says, these could be understood as aspects of negative events: starting p late is keeping on not-p and starting p early is cutting not-p short, {za'o na} and {co'u na}. This compounds the problem with whether this is really what {co'u} means with the problem about negation, but both of these issues are at least largely open to usage decision. All of this was cross-fertilized by a set of terms already mentioned here: "still, " "not yet," "already," and "no longer." Which seemed to fit in with these in various ways. "Still" seemed to have ovious correlation with {za'o} , "not yet" was. of course, related to {pu'o} but also, as noted, to {za'o na}. "Already" had a use related to {co'u na} (premature) and so maybe "no longer" was related to {co'u} , but certainly to {ba'o}. But then, it was noted that in these notions (and, indeed, the earlier more purely aspectual ones), the connection was not quite right: {co'u} marked the transition, the stopping, not the range after the stopping, which is what "already" and "no longer" aimed at (on one reading, at least). Whence came the proposed experimental {xa'o} which was to hold during the time between when something actually starts and the (later) time it should start, a "mirror" of {za'o}. "Already" applied in that time, at least. There have been a variety of fiddling with this to get just what is wanted in each case, but no definitive answers to how to do all the things that some people have tried to do with these devices. On the other hand, some poeople objected to these devices to start with. One view was that at least the "still" etc. set is not aspect at all, but largely things from attitudinals: something you expect doesn't happen and you are impatient so you say it hasn't happened yet, a combination of {pu'o} and maybe a weakened form of {o'onai}. And so on. This involves no change in grammar or vocabulary. Another line was that these things should not be handled in the aspect system but in the bridi (with a slight whiff of the idea that ALL aspects really belong in bridi), explicitly spelling out how the events are related and to what: continues after expected end, and the like. This involves no changes in grammar or vocabulary, but does some damage to several linguistic theories we used to play with. The new possibility, sketched by xod today, is that the tense/aspect system should have a (maybe more?) mark for the normal time of an event and then make a compound tense-aspect to cover starting before that beginning, running on after after than end and so on {co'a pu x}, {ca'o pu co' x}, {ca'o ba mu'o x} and so on, eliminating the need for {za'o} in the process. This involves a new vocabulary item (from the xes I suppose) and possibly some complication of the grammar, but does not seem to involve major problems, only additions. |