Return-Path: Message-Id: <9208141600.AA15817@relay1.UU.NET> Date: Fri Aug 14 14:12:19 1992 Reply-To: CJ FINE Sender: Lojban list From: CJ FINE Subject: Re: la bradfrd jbogirz X-To: amanda@WAM.UMD.EDU X-Cc: Lojban list To: John Cowan In-Reply-To: ; from "Dryad" at Aug 13, 92 7:44 pm Status: RO X-From-Space-Date: Fri Aug 14 14:12:19 1992 X-From-Space-Address: cbmvax!uunet!cuvmb.bitnet!LOJBAN Amanda asks: > > I, a very beginning lojbanist, read and attempted to translate A.D.M.Smith's > letter, and i have a few questions.... > > Am i right in reading "le mi selci'a" as "my text"? "mi" here serves as a > possessive? Yes. This is an anomalous construction that has been there since the early days. It looks like le/lo eg le ko'a broda which is by definition equivalent to le broda pe ko'a "Its broda" > Why does it start with "coi la tcidu"... tcidu is a gismu, right? Does "la" > put tcidu into the "vocative case"? Yes and no. "coi" can be followed by a number of things, most commonly cmene (names) and sumti (arguments), but also selbri ("tcidu" is a gismu, but that is its word-class. Its grammatical function here is as a selbri). It is the "coi" that puts what follows in the 'vocative case' as you put it (more precisely, specifies that its role in the sentence is to identify the se tavla (the spoken to) rather than any function in the main bridi.) All the following are therefore grammatical: 1) coi .tcid. "Hello, Tcid" - Tcid is a name, and would be interpreted by most lojbanists as derived from "tcidu", hence "Reader"; but it could be an entirely unrelated name. (The pause is required before the name.) 2) coi doi tcid. "Hello O Tcid" - Exactly the same, except that a pause is not required after "doi". It is not clear whether there is a pragmatic difference in the presence or absence of "doi". (If there were no COI present, it would be required, otherwise there would be nothing to indicate the cmene was vocative.) 3) coi la tcid "Hello, one named Tcid" - grammatically, this is different, as the thing following the "coi" is a sumti, but it is not clear to me what semantic difference there is, if any. I regard this form as stylistically dispreferred, because it contains a more complicated structure (LA cmene) than 1) and 2); but it may be there are cases when it makes a useful distinction. And of course others may disagree with my judgement. It does not require a pause after "la". 4) coi le/lo tcidu "Hello reader(s)" - this addresses those designated by the sumti "le/lo tcidu", which is a description: "the (particular) reader(s)" or "some reader(s)" respectively. It is not in any sense naming them. 5) coi la tcidu "Hello, one(s) named 'reader'" - this is similar, but it is making a name out the description. It would most likely apply when addressing a group who are conventionally named "Reader". Semantically this is similar to "coi .tcid", but it is more determinate, in that it is precise that the addressees are named "Reader" rather than, for example, some foreign name that happens to be rendered "tcid" in Lojban. 6) coi tcidu "Hello reader(s)" - I had forgotten this form was available until I started writing this answer. I presume it is identical to "coi lo tcidu". Andrew originally wrote 3). I recommended 1), or, better, 4); he posted 5) in his corrected version; I now favor 6). > What exactly does "bo" do? Extract the x1 of the previous sentence? "bo" occurs in various places in the grammar, and seems to me to have two rather different functions. The less common, in tanru and mex, is a tight-binding grouping operator, so melbi cmalu bo nixli ckule = melbi ke cmalu nixli ke'e ckule = (beautiful (small girls)) school = 'school for small girls who are beautiful' What turns out to be a much more common use, exemplified in Andrew's text, is to make some cmavo apply to a whole sentence, or at any rate to a larger scope than would otherwise be the case. So .ije semu'ibo lemi jibri se nelci mi means "And therefore my job is liked by me". If the "bo" were not there, the "semu'i" would swallow the following sumti, giving: .ije semu'i lemi jibri se nelci mi "And with-motivated-result: my job, (something) is liked by me" > What does he mean by "skami nelci"? ("mi skami nelci") "skami nelci" means "is a (computer for purpose ...) liker of ...", for which a reasonable interpretation is "computer liker". (It could also mean he is a computer that likes something.) > Doesn't "tarmi naldikni" mean something like "shaped disorder"? Or does it > mean "disordered shape" (seemingly a more appropriate description of a fan)? Not quite. "naldikni" means "is other-than regular", so "tarmi naldikni" is "is a shape kind-of irregular-thing". "Disorder" would be an abstraction ("ka naldikni" or "nu naldikni"). Notice Andrew originally had "naldikni tarmi", which in English looks like "irregular shape". However, my suggestion was that there were two problems with that. The more serious is that "tarmi" means "is a/the form of" - a shape, not an object. In English, we can refer to (our perception of) an object as "a shape", but in Lojban this is not appropriate - in fact, better would be to use the x2: "le se tarmi" = "the shaped thing", and "le naldikni se tarmi" would be OK for "the irregular shaped thing". I felt there was a further problem, in that "dikni" is not specific as to the kind of regularity - time, form, sequence, whatever. Probably "le naldikni se tarmi" would do, as the presence of the brivla "tarmi" would bring us into the realm of shapes (it *could* be 'the irregularly-occurring shaped-thing', but that seems less likely), but neater, in my view, is to turn it round. This is partly because of my propensity for using Lojban in distinctly non-English ways: I like "vorme blanu" for "blue door" (which would not be appropriate if I were going on to say what it was a door to, since using "vorme blanu" makes its blueness, and the place-structure of "blanu", salient). > > Finally... how does the last sentence ("ri jo'u gi snuji lei sovda joi tamca > le burna'a gi le ckafi") work? Is "snuji" the selbri here? Why is "gi" > in front of the selbri and in front of a sumti? How does this work? You're right, it's ungrammatical. Andrew's original had ri du snuji ... gi'e le ckafi which was ungrammatical, because "gi'e" must be followed by a selbri (+ following sumti). He meant either ri du le snuji ... .e le ckafi or ri snuji ... gi'e ckafi The latter is very much preferable, because "du" has the function of identifying, not predicating. However, I observed that this form expands to ri snuji ... .ije ri ckafi (ignoring the possibility that "ri" would pick up something else in the second sentence). I suggested that he did not mean to say that his lunch was a sandwich and it was also coffee, but that it was jointly both: "jo'u". Probably the best way to say this would be to get these in sumti form, repeating a selbri: .i mi citka le snuji ... ku jo'u le ckafi or better .i sanmi le snuji ... ku jo'u le ckafi "(Something) is a meal of a sandwich ... and coffee" but we went for a connected selbri. Using a selbri is a little problematical for two reasons. First, the semantics of non-logically connected selbri are a bit odd: "is a-sandwich-and-jointly-coffee" (?). Secondly, there does not appear to be a way to do it in afterthought. (I think this is a hole in the grammar) We found that we could do it by a special form of the forethought connective ge ... gi ... the special form being needed for a non-logical connective: jo'u gi ... gi ... almost anything can go in the two arms (the same in both arms), and the result it the same kind of entity - in this case, bridi-tails. So the corrected version is: .i ri jo'u gi snuji lei sovda joi tamca le burna'a gi ckafi > > Thank you for your patience... > amanda je'e.uisai If you don't ask, you don't learn; and answering such questions helps me to learn too. co'omi'e kolin