Wednesday, August 20, 2014

Note on temporal relations

Part of working on prepositions and on 'la'

Known:
Vector tense/implicit particular quantifiers over occasions.
X la = lon X
tenpo pini (P)/ tenpo ni ()/tenpo kama (F)

(fussinesses: 1) the scope of CP (condition phrases, 'la' phrases) are not defined, so subsequent unmarked sentences might be narratively continuing these times or revert to the moment of speech or some other contextually defined time. Similarly, subordinate sentences with these marks might be secondary tenses off the implicit axis or redefined from the absolute base: 'tenpo pini la jan li toki e ni: tenpo pini la meli li kama', her coming before person speech or just before utterance of this complex? I am inclined to think that a temporal in a CP would restrict one in a PP in the sentnece to which the CP was attrached or in a CP that followed it attached to the same sentence, but I can imagine a case being made for the opposite positions.
2) Strictly, 'tenpo ni' can mean “then, at the time of the just mentioned event” as well as “now” and so be a temporal connective. To avoid confusion – and because having a way to get back to the now is always handy – I recommend using 'tenpo sama' for the relative meaning.)


Extended vector tense (relative tensors)
Same pattern (?)
tenpo pini/kama lili/suli  "a little/long time ago, in a little while, a long time hence"
     
      (fussiness:  not sure how apt 'lon' is here. since it is not in the stretch of time put rather at the end of it --
      which is, admittedly probably 'lon' as well. The pattern 'tenpo lili pini' makes as much sense, apparently,
      but this one is established.)

Other quantifier cases
(universal)
presumably also X la = lon X
tenpo ala/ ali “never/always”

(fussiness: I suppose this are usually in fact only about past occasions (and the present?) but that is only implied. The explicit cases do not seem to have occurred (and are rare in English, too) but would presumably be formed just by adding 'pini' or 'kama' to the end of the forms just given to make G and H:”has never/always” and “will never/always”. It is arguable that the forms should be with the opposite arrangement of quantifier and direction, e.g. 'tenpo pini ali' for G. Since there are no occasions, it is hard to figure out which is right or even what principle to use to decide. But see later. At least it is clear that dividing the quantifier and the vector into two phrases does not help: 'tenpo pini la tenpo ali la' and 'tenpo ali la tenpo pini la' bear undetermined relations to one another, some bound up in the metaphysics of time but also in the issue of the scope of 'la' phrases, but none of them giving the equivalent of what is intended here. And they raise issues about the scope of PPs which we really don't want to mess with now.))

(numerical) iterations
Still the same pattern, I suppose
tenpo N/mute/lili “N/many/(a) few times” (maybe also 'mute lili' and 'lili mute' ""several times”/”occasionally” and “rarely” - not, apparently set off with 'pi').

(fussiness: All the problems about past and future cases apply here, though the implication that this is about the past is even sharper. But the two prenex solution, assuming 'tenpo pini/kama la' subordinate 'tenpo N la', works better here. 'tenpo lili' can also mean “a short time” and so belongs in the duration section as well as here. Perhaps the PP form is distinctive.
Iteration needs to be distinguished from repetition, the same action done repeatedly on a single occasion. Contrast: “The officer has shot suspects on three occasions” with “The officer shot the suspect three times” It is not clear just how to do repetitions but apparently either an adverbial phrase 'pana pi tenpo tu wan' or even just 'pana tu wan' or some modification in the various nominal complements, 'pana e sike kiwen tu wan' or 'utala e ona kepeken sike pana tu wan' are adequate.)

Unitized occasions
presumably same pattern
tenpo suno/pimeja/esun/mun pini/ni/kama “yesterday, today, tomorrow, last night, tonight, tomorrow night, last week, this week, next week, last month, this month, next month”
sike suno pini/ni/kama “last year, this year, next year”

(fussiness: All non-ni strictly mean “some past/future day/week/month/year” but are used for the
proximal one by convention/implication. The more general understanding adds to the simple 'tenpo' only a sense of the length of the displacement from the present, which is not strictly relevant here. 'ni' again has the possible sense of “the same as that containing the latest mentioned event” and the 'sama' solution is recommended again.


Known phrase in 'la', preposition unknown

Duration
tenpo suli/lili “for a long/short time”
tenpo suno/esun/mun N (including 'mute', lili' and compounds) “for n days/weeks/months”
sike suno N “for n years”

The issue of the correct PP gets into the whole matter of prepositions in tp, which is the subject of another study under way. The “for” in the English for this notion means that the event referred to occupies the whole of that time and no tp preposition is obviously telic in this way ('tawa', for example, does not assert that goal is actually reached, so is “toward” as much as “to”). Parallels in other languages (“por” or “para” in Spanish, in particular, but also similar patterns in German and French and Latin) give varying suggestions for use in tp, based on equally dubious associations. Personally, I feel like promoting 'awen' to prepositional status for the occasion, but that is a bit radical until other options have played out. Even the use of 'lon' is not automatically excluded, since it is areal as well as punctal (“in” as well as “at”) and so is at least a partial fit (and as good as other obvious choices).

My attention has been called to the use of 'kepeken tenpo mute' as part of the expression for "slow" 'tawa ma lili kepeken tenpo mute'  Whether this can be generalized (maybe with 'mute' changed to suli') offers a way out of the usual preposition debate (and a nice counter to the arguments about how to indicate the language used).


Tp doesn't have words for hours, minutes and second, but, if it did, presumably this pattern would apply with them as well.

(fussiness: the numerical cases could also be read as iterations of specified periods. The difference seems to be mainly that duration has the event taking place throughout consecutive stretches of the sort mentioned, N weeks but attached end to end, where as iteration assumes some gaps between the weeks involved (and that the whole event is repeated, not part in one week and a further part in the next, although all of this is open to context. It might take me three months to walk the Appalachian Trail, even though the months were in fact divided into several disjoint stretches of a week or so each, so 'tenpo esun mute la mi tawa noka lon nasin Apalasi' is true in both senses.) Again, the PP might be different, but I am not sure.)

Temporal overlap
S1 la S2 “When/while/if S1, then S2”
Presumably the PP format is 'S2 Prep ni: S1'

This relation is so weak that it is enough for it to be true that there is one instant during S1 holding that S2 hold as well (and, of course, there can be true cases where even this does not apply). We usually (I think) have something more explicit in mind. And it may be that sorting these out is a necessary step before doing much more about the PP. We can clearly separate out the general claims as being 'tempo ali/mute' (and maybe a few non-temporal factors) with scope over the whole complex (which means we have some clues about scope!). What we seem to be left with are that either S2 holds at some point during S1 or that it follows “immediately” on S1. The first gets subdivided among cases where S2 and S1 coincide exactly or one falls wholly within the other or they overlap with a bit of one preceding the other and a bit of the other following the one. But this expressions does not indicate any of these differences, so we need just the minimal claim of a relationship and so are back to just what the prepositions say. It may be, of course, that the prepositions are all more explicit than 'la' and that this is a 'la' expression which does not have a prepositional analog.

Unknown
Displacement (tensored vectors)

So, an event is in the past/future, but how far specifically? The answer is, presumably, in terms of some unit and some number of that unit, combined with the direction in time. We have the pattern already of unit + direction = displacement by one unit in that direction: 'tenpo suno pini', say. The only apparent question is where does the N, in this case the implicit one, go? That is, is “three days ago” 'tenpo suno tu wan pini' or 'tenpo suno pini tu wan'? And, of course, does it make a difference? It would, certainly, if there were another notion around that would take up the other form. In this case, the possible other notions seem to be directional duration, which seems highly unlikely (“for three past days”) or iteration with specified periods, which seems somewhat more likely (“in three separate past days”) but allows of a two-prenex solution. S, it seems just a matter of making a choice (and maybe coming up with a principle afterwords to justify it – no prior principle is obvious). The earlier case of relative displacement suggests puttting the number last, but it is itself open to question. Once the phrase is set, the X la = lon X patterns seems appropriate.

Relative position.
How are two event related in time, more precisely than the S la S pattern and, particular, relative to one another. The English patterns to be dealt with are
After S1, S2
Before S2, S1
While S1, S2 (with some of the subdivision mentioned under 'la' spelled out here)

Put this way, this appears to be about conjunctions, but, given the short supply of those in tp (and the unclarity about how to extend the list without creating more problems than it is worth), it is easier to think of the problem as being about PP, on the model of “because”, 'tan ni', or as introducing a new category of CP.
On the preposition side, for the first two cases, a single pattern presents itself, since 'tan' is pretty clearly retrospective, looking backward toward the beginning, That means that “after that” could be just 'tan ni', but, to keep our “post hoc” separated from our”propter hoc”, we would go for 'tan tempo ni'. By parity of reasoning (or parody), “before this” would be 'tawa tenpo ni' Thus we get the four patterns for the order S1 S2
S1. tan tenpo ni la S2
S 1 tawa tenpo ni: S2
S2. tawa tenpo ni la S1
S2 tan tenpo ni: S1
The other possible pattern takes up the general notion of a condition. To say that S2 is later than S1 is to say that S2 occurs in a condition in which S1 is done or that S1 occurs in a condition in which S2 is yet to come. Alternatively, we can say that S1 is in S2's past and S2 is in S1's future. This gives the following patterns
S1. ni pini la S2
S2. pini ni la S1
S1. kama ni la S2
S2. ni kama la S1
It must be said that the reading of 'ni pini' as “this being over” and “pini ni' as “in the past of this” is open to challenge with exactly the opposite readings. This is a reason to prefer the prepositional solution, although the fact that the PP with opposite results occur at virtually the same place (the difference between ':' and '.' being not very great and the 'la' coming late) makes it tricky as well (but the same situation has not caused obvious problems with 'tan ni').
The simple 'ni' and 'kama/pini' phrases could obviously be moved to PP, presumably with 'lon', but then the temptation might be to make that, like 'tan ni', into version which picks up the next sentence, reversing the order meaning yet again. That is, S1 lon pini ni: S2 = S1. kama ni S2 and so on.

The final, “while” case raises a number of problems, some already mentioned, others involving things like the aspect modals ('open, awen, pini, kama'), that lie outside this paper. The solutions are more or less the same as the other two. In the second version, for example, 'awen' can fill in for 'pini/kama'. The prepositional form is less obvious, but 'lon', being areal and also the only spatial preposition left, might work. But many questions seem to remain (mostly probably pure fussiness, the Lojban syndrome).

It would be nice if we had a firm grasp on sentential nominalization so that we could extend this discussion from the case of two sentences to to a sentence and a noun "after his coming" "before the battle" (don't even mention the "just" -- 'taso' just doesn't work).  I suppose the second could be 'utala kama' or 'pini utala' and the former similar things with 'kama ona', but problems seem sure to arise (as here already) and we don't know how to handle most sentences, "John hunted deer on Thursday," for example.


Displacement again
So, one event comes after the other; how soon after? Both parts of this issue have been dealt with (more or less), so the question here is how to combine them. They do appear to be independent in the sense that displacement does not depend upon whether we are talking about S1 before S2 or S2 after S1. The only significant fact is that the earlier discussion of displacement was in terms of past and future relative to the present (or whatever the axis is) and that factor drops out. But, on the other hand, we cannot, apparently use just the numbered units (or 'tenpo suli/lili'), since that is duration of the event, not of the gap between events. Using some preposition raises the question of which one again. Using the desperation 'pi' raises the question of what to attach it to. This seems the most open question.


No comments:

Post a Comment