Afraid of Horses, was PEG Morphology Algorithm Posted by pycyn on Tue 22 of Feb, 2005 22:25 GMT posts: 2388 Use this thread to discuss the Afraid of Horses, was PEG Morphology Algorithm page.
Posted by pycyn on Tue 22 of Feb, 2005 22:25 GMT posts: 2388 wrote: xorxes: > > > > > > {la prenu poi lo ke'a xirma ji'a sai > cu > > se > > > > > > terpa}, > > > > > > > > > > Not > > > > > very snappy though. > > > > > > > > Just as snappy as the corresponding > > > > description. > > > > > > > In English, I suppose you meant. > > > > No, I mean in Lojban: > > > > {la prenu poi lo ke'a xirma ji'a sai cu se > > terpa} > > vs. {le prenu poi lo ke'a xirma ji'a sai cu > se > > terpa}. > > > > If you can think of a snappier description in > > Lojban, > > you automatically get a snappier name. > > Thanks; this clarifies the problem considerably. The fact that a "meaningful name" is automatically taken to be a mere variant of a description (as it really does seem to be in Lojban) is the ultimate source of the problem, since it makes all names endocentric, the "description" applies to the referent, whereas, in natural languages with functional descriptive names they may be exocentric to varying degrees. The most common endocentric forms name thhings for what they are associated with (usually, for humans, posessions), the two historic classics being "Much Rice" ("bahuvrihi", also the Sanskrit name for these compounds) and "Red Cap," the English example. Both of these have been at least Loglanized, the one as "possessor of much rice" and "redly behatted" (back when clothing had a place for the wearer). Cleaning this up happily takes very little syntactic change, merely a reinterpretation (or rather a ceasing of an enforced interpretation) of {la redro mapku} and {la piso'i rismi}. More remote cases (which come closer to the Afraid-of-horses problem) are more difficult, since the relevant expressions do not compute in the site: "Raven-steals-the sun" is basically a sentence, which cannot come after {la} without converting at least partially to description form: {la la raven zerle'a be la sol cu klama} parses about right, which may explain why book title quotes disappeared at some point. Our NA name then might be something like {ze'a terpa be le ke'a xirma}. Unless there is some trap hidden in the grammar that does not show up here. But all that have turned up so far is that first arguments have also to be nominalized — and shifted back — if they would give rise to the need for a {cu} other wise: {la se terpa be fa le ke'a xirma be'o ze'a cu klama} works but {la le ke'a xirma cu be ze'a cu klama} obviously would not (though it works with {ku} in place of {cu}). So, it seems that — given a few minor changes which are, after all to be expected in going from sentence to name — the problem I saw in Lojban does not exist and only the habit of taking all meaningful names as endocentric gave rise to the appearance of a problem.
Posted by JohnCowan on Tue 22 of Feb, 2005 22:25 GMT posts: 149 John E Clifford scripsit: > Cleaning this up > happily takes very little syntactic change, > merely a reinterpretation (or rather a ceasing of > an enforced interpretation) of {la redro mapku} > and {la piso'i rismi}. No such enforced interpretation of tanru has ever existed. Attempts have been made with fair success to regularize the interpretation of lujvo, a very different thing. > "Raven-steals-the sun" > is basically a sentence, which cannot come after > {la} without converting at least partially to > description form: English is far more loose-jointed than other languages, which are typically far less able to treat whole sentences as NPs, even name NPs. I suspect the true translation is Raven-who-steals-the-sun (i.e. Total-eclipse), which is a proper NP and eminently Lojbanizable. -- John Cowan cowan@ccil.org http://www.ccil.org/~cowan Is it not written, "That which is written, is written"?
Posted by pycyn on Tue 22 of Feb, 2005 22:25 GMT posts: 2388 The real weakness is that Lojban probably needs more UI3 discursives (or previously semantically-analyzed brivla) for things like "even", "mere", "just", along with the po'o "only". There are those who want these logically analytical, and those who want a short form that saves the analysis. Lacking the analysis in a form short enough to say without thinking, people simply skip trying to express it and make errors. I think this is one of the real shortcomings of the language at present; words that are common in other language have no easy expression in Lojban, and few are competent to analyze the logical implications of the words they are trying to translate. Having failed to get cmavo for them, at this point I would try for brivla and not try to make them dikyjvo, but coming up with a list, and a good pattern for making them as lujvo is something I don't feel linguistically or logically competent to do. Tthere are a lot of gaps in Lojban and it would be useful to get a list going of of ones like these that we need. But saying "ones like these" is a bit misleading, since these don't seem to fit into any pattern and so it seems unlikely that any one pattern of solutions will work for all. "only" has, of course, a completely general and adequate expression (for its main uses — there may be some obscure cases) in terms of quantifiers and connectives, but no one uses it (and half the people who look at it deny that it works). It also a cmavo to help with at least one awkward case ({ro klama bi la djan} for "Only John came" becomes {la djan po'o klama}) which has then been extended in questionable ways -- to cover predicates as well sumti and plurals as well as singulars. Aside from legitimating those extended uses (and CLL already goes a long way in that direction), we surely don't need even more for this concept. "even" is trickier. "Even John came" used to be analyzed as "Some other thant John came. John came. It is surprising that John came." More detailed accounts get to "It was prima facie unlikely that John would come, but he did. Most (all, many, enough) of the people that were more likely than John to come came. ??No one less likely to come than John came." So the crucial factors are spread over several areas. 1. John came (bare sentential claim) 2. The speaker's surprise that John came (an attitudinal) based on (it seems) 3. an initial unlikelihood that John would come — surely a different kind of attitudinal, even an evidential or a report of (perceived) probabilities. 4. Other people came and 5. they were a significant portion of those expected (report and probability again). Now, this combination occurs often enough to deserve some collapsed form — and (as the sentence above shows) one that can be used within modal contexts as well as assertoric ones. It would appear to have to be a free modifier, since it can apply to (focus on) any point in the sentence. "mere" (and one sense of "just" — another is a version of "only") seems to add to a straightforward claim an injunction (a negative imperative)to not make a federal case of the relevant event. A mere scratch is one that does not justify whatever behavior is going on: screaming, calling 911, or giving the kid ice cream. A mere graduate student is one who is not to be trusted on the topic, has no authority, and so on. "Mere" is, in other words, an other- directed directive: don't you do anything (that I eould think extreme) about the situation.
Posted by pycyn on Tue 22 of Feb, 2005 22:25 GMT posts: 2388 wrote: Our NA name > then might be something like {ze'a terpa be le > ke'a xirma}. Unless there is some trap hidden > in > the grammar that does not show up here. I do not know how {ze'a} got into my head, but what was intended was one or the other of {zo'e} or {zu'i} and similarly throughout. > But > all > that have turned up so far is that first > arguments have also to be nominalized — and > shifted back — if they would give rise to the > need for a {cu} other wise: {la se terpa be fa > le > ke'a xirma be'o ze'a cu klama} works but {la le > ke'a xirma cu be ze'a cu klama} obviously would > not (though it works with {ku} in place of > {cu}). > So, it seems that — given a few minor changes > which are, after all to be expected in going > from > sentence to name — the problem I saw in > Lojban > does not exist and only the habit of taking all > meaningful names as endocentric gave rise to > the > appearance of a problem. > > > >
Posted by pycyn on Tue 22 of Feb, 2005 22:26 GMT posts: 2388 > John E Clifford scripsit: > > > Cleaning this up > > happily takes very little syntactic change, > > merely a reinterpretation (or rather a > ceasing of > > an enforced interpretation) of {la redro > mapku} > > and {la piso'i rismi}. > > No such enforced interpretation of tanru has > ever > existed. Attempts have been made with fair > success > to regularize the interpretation of lujvo, a > very > different thing. Well, let us just say then that there is a marked tendency for people to read them in that way, as xorxes comments exemplify (and as at least two discussions over the years — about Red Cap and Much Rice, in fact — have also shown). So, the point here is just to remind folks that they need not be bound by this habit. > > > "Raven-steals-the sun" > > is basically a sentence, which cannot come > after > > {la} without converting at least partially to > > description form: > > English is far more loose-jointed than other > languages, > which are typically far less able to treat > whole > sentences as NPs, even name NPs. I suspect the > true translation is Raven-who-steals-the-sun > (i.e. > Total-eclipse), which is a proper NP and > eminently > Lojbanizable. Well, I think that, as a name, "Raven-steals-the-sun" has next to nothing to do with eclipses and quite a bit to do with the range of cultural values which center on Raven as the source of life and light, etc. In any case, as noted, the changes that need to be made are minimal — nominalization — and the result is still an nice obviously exocentric name. Unless I have missed something crucial. > — > John Cowan cowan@ccil.org > http://www.ccil.org/~cowan > Is it not written, "That which is > written, is written"? > > >
Posted by xorxes on Tue 22 of Feb, 2005 22:26 GMT posts: 1912 > --- John Cowan <cowan@ccil.org> wrote: > > John E Clifford scripsit: > > > > > Cleaning this up > > > happily takes very little syntactic change, > > > merely a reinterpretation (or rather a > > ceasing of > > > an enforced interpretation) of {la redro > > mapku} > > > and {la piso'i rismi}. > > > > No such enforced interpretation of tanru has > > ever > > existed. Attempts have been made with fair > > success > > to regularize the interpretation of lujvo, a > > very > > different thing. > > Well, let us just say then that there is a marked > tendency for people to read them in that way, as > xorxes comments exemplify (and as at least two > discussions over the years — about Red Cap and > Much Rice, in fact — have also shown). So, the > point here is just to remind folks that they need > not be bound by this habit. You and John seem to be talking at cross purposes here. John said that there is no fixed rule to interpret how a tanru component modifies another, but your point has nothing to do with tanru, it would apply to {la makpu} just as well. > > > "Raven-steals-the sun" > > > is basically a sentence, which cannot come > > after > > > {la} without converting at least partially to > > > description form: You could use {la du'u ...} mu'o mi'e xorxes __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
Posted by pycyn on Wed 23 of Feb, 2005 06:30 GMT posts: 2388 wrote: > > --- John E Clifford wrote: > > --- John Cowan <cowan@ccil.org> wrote: > > > John E Clifford scripsit: > > > > > > > Cleaning this up > > > > happily takes very little syntactic > change, > > > > merely a reinterpretation (or rather a > > > ceasing of > > > > an enforced interpretation) of {la redro > > > mapku} > > > > and {la piso'i rismi}. > > > > > > No such enforced interpretation of tanru > has > > > ever > > > existed. Attempts have been made with fair > > > success > > > to regularize the interpretation of lujvo, > a > > > very > > > different thing. > > > > Well, let us just say then that there is a > marked > > tendency for people to read them in that way, > as > > xorxes comments exemplify (and as at least > two > > discussions over the years — about Red Cap > and > > Much Rice, in fact — have also shown). So, > the > > point here is just to remind folks that they > need > > not be bound by this habit. > > You and John seem to be talking at cross > purposes here. > John said that there is no fixed rule to > interpret > how a tanru component modifies another, but > your point > has nothing to do with tanru, it would apply to Ooops! I missed that though I was puzzled by his remark about lujvo. > {la makpu} just as well. > > > > > > "Raven-steals-the sun" > > > > is basically a sentence, which cannot > come > > > after > > > > {la} without converting at least > partially to > > > > description form: > > You could use {la du'u ...} > I expect I would prefer {nu} but I don't find {la la raven zerle'a be la sol} bad at all and even {la cadzu be fa le nanmu bei lo djacu} is acceptable, with {la le nanmu ku cadzu be lo djacu} slightly better.