From @YaleVM.YCC.YALE.EDU:LOJBAN@CUVMB.BITNET Tue Sep 7 06:01:03 1993 Received: from ELI.CS.YALE.EDU by NEBULA.SYSTEMSZ.CS.YALE.EDU via SMTP; Tue, 7 Sep 1993 10:00:22 -0400 Received: from YALEVM.YCC.YALE.EDU by eli.CS.YALE.EDU via SMTP; Tue, 7 Sep 1993 10:00:14 -0400 Message-Id: <199309071400.AA12429@eli.CS.YALE.EDU> Received: from CUVMB.CC.COLUMBIA.EDU by YaleVM.YCC.Yale.Edu (IBM VM SMTP V2R2) with BSMTP id 0699; Tue, 07 Sep 93 09:58:37 EDT Received: from CUVMB.COLUMBIA.EDU by CUVMB.CC.COLUMBIA.EDU (Mailer R2.07) with BSMTP id 2059; Tue, 07 Sep 93 10:01:09 EDT Date: Tue, 7 Sep 1993 10:01:03 -0400 Reply-To: vilva@viikki21.helsinki.fi Sender: Lojban list From: Veijo Vilva Subject: My last of ZAhO , sumti places To: Erik Rauch Status: RO X-Status: I hope to be forgiven still one posting on this theme. I sincerely hope that my ponderings about the sumti places do not start a YAET (yet another endless thread). > Date: Mon, 6 Sep 1993 18:58:28 EDT > From: Jorge LLambias > Subject: Re: Event contours and ZAhO tcita > I don't agree with this. I think that a sumti place left out of the > definition changes the meaning of the relationship. {klama} and {cliva} > are not just two ways of writing the same thing. They are not the same. {cliva} emphasizes the fact that k2 is not relevant. However, the relationship between {k1,k3,k4 and k5} is more or less the same as that between {c1,c2,c3 and c4}. [] > Of course, but you were using the rule as a justification. If it can > only possibly apply meaningfully to ZAhOs then the fact that it applies > in general does not explain why it applies to ZAhOs. I think I wasn't trying to use it as a justification. I tried to state MY present conclusions concerning the consequencies. I ought to be more careful with my formulations :-) > > BTW. I do agree with John Cowan in that I see a deeper connection > > between the outer point event and the phase of the sumti event > > in addition to the temporal sync. The formulation in the last > > revision of the 'Imaginary Journeys' (Rev. 3.17, available from > > the FTP server) seems to me quite adequate. > > > I also agree about the deeper connection, I just would like it in the > reverse way: between the sumti and the phase of the outer event. OK. And I prefer the present way :-) > > (3) __mi dunda le cakla do [bai] zo'e__ > > > > This to me means something different. The idea of coercion is totally > absent in the first sentence, present in the second, and hinted at in > the third. [] > > because all the sumti places are there even if they are not > > expressed in the definition of "dunda". The question of coersion > > is left open in the definition but the possibility is not excluded > > -- a "bai noda" is required if we are to state explicitly that > > there is no coersion involved. > > I agree that it is not excluded, but it is not left open in the same > sense as when a sumti place is left vacant. I think the question is rather vexed. We have to intertwining levels which affect the choice of the sumti places to keep. On the one hand we have the underlying semantics, on the other the economics of pragmatics. With the limited number of gismu, the requirement of unique meanings and the quest for an optimal set of gismu places from the pragmatic viewpoint (this is a minimax problem) we end up with rather 'loose' definitions. This results in gismu which give a rather Platonian view of the world. The gismu represent very idealized objects and ideas, they are a compromise between conflicting goals. The problem is further aggravated by the fact that the definitions are available only in English -- and in a rather terse English at that. Every time you and me interpret the gismu list we are bound to make different judgments -- even John and Colin, e.g., do have differences of opinion every now and then. This means that the gismu are in a way lacking in nuance, they must be -- at least for the time being and probably for ever, given the limited numbers. Most of the beauty of Lojban, for me, lies just in the fact that it forces me to see the general in the particular. The particular is expressed to the necessary degree of accuracy by elaborating on the general. I didn't take part in the conversation about a cmavo to delete an existing sumti place. I see separate sets of sumti places as more of a convenience feature than as actually defining a different relation. The selection of the sumti places to retain is rather arbitrary and sometimes also dictated by the need to achieve a certain degree of conformity. There are also situations where certain subsets of the sumti places of a basic relation occure frequently enough to warrant appointing a new alias just to achieve a better economy of speech. Sometimes the apparent need to delete a sumti place arises from the use of an unsuitable metaphora, e.g. all the talk about deleting the maker from {zbasu} when the required relation is more like 'to consist of', {ciste} or {stura}, i.e. we have all the idiosyncracies of natlangs to cater for. > > c1 cliva c2 c3 c4 seka'a k2 = c1 klama k2 c2 c3 c4 > > > > or to put it in reverse > > > > k1 klama zo'e k3 k4 k5 = k1 cliva k3 k4 k5 [seka'a zo'e] > > > > Basically both klama and cliva are incarnations of an underlying > > super-klama relation which describes movement and includes all > > imaginable sumti places expressible with BAI, ZAhO tags etc. ad nauseam. > > If this is true, then thanks to fi'o, all selbri are just incarnations > of one single mega-super-broda. I.e. they all mean the same thing, and > observatives where no sumti is specified are all synonymous. I guess the main difference between mi klama ti and ka'a mi co'e seka'a ti is that with klama you know what has been left out. Basically a gismu definition does 3 things (1) it gets rid of a bunch of tags (2) it specifies which tags have been made redundant by defining fixed sumti positions for the corresponding sumti (3) as a consequence it tells the listener goi ko'a what sumti ko'a can expect to hear so as to be able to tell what has been left out The gismu mechanism is actually quite clever. It achieves two mutually contradictory goals at a same time -- brevity and a kind of redundancy through elision. > The definitions of klama, cliva and other movement gismu just pick a > suitable subset of these myriad places. The important thing to note is > that inserting a sumti into anyone of the underlying invisible sumti > places will have no effect either on the relationship between the already > expressed sumti or the contour of the event. > {mi klama} and {mi cliva} say exactly the same thing? I think the only difference between cliva and klama is that the use of cliva tells ko'a that k2 is deemed entirely irrelevant in the context -- otherwise both the bridi can tell exactly the same thing. Now, my thoughts keep evolving. We have actually two choices (1) mi cliva = mi klama ne'e (2) mi cliva = mi klama zo'e (1) corresponds to theory/your view and (2) to practice. To be exact we ought to use {cliva} only when there really is no destination. Due to natlang influences we are, however, quite often using it in in sense (2) -- which is nearer to the view I have represented. {litru} belongs to the same category. In ko'a litru le dargu le karce we do not usually suppose that there is no origin or destination in the absolute, we just either do not know them (?the original intention) or are using {litru} as an alias for {klama zo'e zo'e}. Actually a {litru} without x3 would be quite useful -- it would describe movement in a medium with no implication of a means. I think pragmatic reasons will quite often dictate the choice of gismu and the use of the existing ones. Also, I think that the later sumti places will quite often be just chopped off instead of filling them with {ne'e}, e.g. litru le ricfoi It will be much easier to concile theory and practice if we adopt a certain flexibility regarding the relationship between meaning and the specific set of sumti places. I think it will take at least a second or third generation native speaker of Lojban to keep {klama}, {cliva} and {litru} strictly apart and even then at the expense of undue verbosity. > > (2) it can - and almost invariably does - specify a 'phase' > > for the sumti itself. Each time a sumti is inserted into > > a sumti place in a bridi a specific aspect/phase/contour > > of the sumti itself is raised. It doesn't matter whether > > the sumti place in question is a standard place or a tagged > > place - there is no principal difference, standard places > > just omit the tags. > > I accept your explanation of this one. However, the property of the > sumti that is raised (I prefer "property" rather than "phase", and this > is what you use in your explanation) need not be the phase that the ZAhO > points to, but could be the marking of that phase. I agree if it is the phase of the sumti event -- and you want it to be the phase of the selbri :-) > > mi zutse le srasu za'o[ku] ?!?! le nu co'a carvi > > ti'u !!!! > > This would work for all time tenses, just as tu'i would for space tenses. Make it {seti'u}, then it will work. If you insert {seti'u}, the {za'o} will be parsed as a tense, not as a sumti tcita. The same happens if you insert about anything else but another ZAhO. With {seti'u} in there we agree, I think :-) But this still leaves the question of ZAhO tcita :-( > > I would express your example in the following way > > > > mi za'o zutse le srasu ba'o le nu co'a carvi > Yes, this is good with the current interpretation. As I would like it, > it is also meaningful. It would say "My sitting on the grass for too > long is over, by the time it begins to rain". insert the {ti'u} and you get what you are after, leave it out and I get what I'm after :-) [If we suppose that double ZAhOs work this way] > > I go to the store via the tiny space of your dancing to the > > sound of ice. > > You want to join the {ve'i} with the x4? Nope. Just couldn't pick up a better expression in a hurry. My English has it's limits :-) > I still don't see how dancing to the sound of ice can determine > a tiny space for the {klama le zarci} to take place, unless much > more context is provided. Well, let's say it's rather ethereal :-). Could be it creates a locality with a certain atmosphere and I cross this locality on my way. There are writers who tend to use expressions like this. > > I think it might actually be quite useful to limit the dancing to > > the tiny space -- the spatial extent of the dancing would then define > > the size of the otherwise rather vague interval. > You want both events in the same space? I'd accept that this may be the > case for some sumti, but I'd prefer the sumti to simply indicate the > space, not necessarily to take place in it. I wasn't too happy with the word 'limit'. The latter half of my expression is better. If we change it to 'The spatial extent of the dancing does vaguely define the size of the otherwise even more vague interval.' you might agree. I'd hesitate to imply, however, that the interval is necessarily smaller than the extent. > > mi zbasu le karce ve'i le kumfa > > I assemble the car within the small space of the room > > Right, but the room need not be small, it is small as far as assembling > cars goes, or it may only have little free space, but the smallness is > relevant only to the assembling, not to the room. I agree :-) with the slight reservation made above :-| > > as opposed to the following three variations on the same theme > > > > mi zbasu le karce ne'i le cmalu kumfa > > I assemble the car in the small room > > > > mi ve'i zbasu le karce ne'i le kumfa > > I assemble the car in a small space within the room > > > > mi zbasu le karce ne'i le ve'i kumfa > > I assemble the car within the room occupying a small space > > > > How about it? There are, of course, other possibilities, but I > > think this interpretation wouldn't contradict anything and > > would have at least some expressional merit. > Absolutely. And it agrees well with what I'd like for the ZAhOs. mi zbasu le karce ne'i le kumfa I assemble the car in the inside of the room inside the room mi zbasu le karce ba'o le nu sipna I assemble the car in the aftermath of sleeping after sleeping I get a match, too :-) The following gets what you'd like to have mi zbasu le karce ba'o seti'u le nu sipna I guess this is as far as we can get. It's been nice. :-) I hope the others aren't too fed up with us :-) I'll fall in line in the question about the {klama}/{cliva}/ {litru} type gismu families and the meaning of tagged sumti places. I did perhaps get out of my depth :-) co'o mi'e veion ------------------------------------------------------------------ Veijo Vilva vilva@viikki21.helsinki.fi