[lojban-beginners] Re: du'e preti

["Brett Williams" <mungojelly@gmail.com>]

On 4/14/08, Minimiscience <minimiscience@gmail.com> wrote:
> I have completed the very excellent *Lojban for Beginners* and looked through
>  parts of the Lojban Reference Grammar for elaboration on several topics;
>  however, I still have quite a number of questions about the Lojban language:

.i'i fi'i
 
These are very interesting questions!  What follows are only my opinions.  Some questions in Lojban are difficult, and others are unresolved.  I have no particular authority. :)
 

>    1. If all of the places of a <bridi> have been filled, and said <bridi> is
>       nested inside another <bridi> (e.g., in an abstraction), is it still
>       necessary to explicitly terminate the nested <bridi> if it is followed by
>       a <sumti> of the main <bridi>?

Yes.  The grammar structures nested bridi without regard to their semantic content or place structure.
 
Note that in most cases it isn't even the nested bridi which is most sensible to terminate.  It's usually more clear to terminate the abstraction than the bridi it contains.  That is:
 
do simlu lo ka mencre vau mi
do simlu lo ka mencre kei mi
do simlu lo ka mencre ku mi

do simlu lo ka (mencre vau) mi
do simlu lo (ka mencre kei) mi
do simlu (lo ka mencre ku) mi
 
"You seem smart to me."  Any of the terminators works grammatically, so we use whichever is clearest.  "ku" could be terminating a sumti within the embedded bridi, and it's less clear whether "vau" is terminating the main bridi or an embedded bridi, so in this case it's usual to see "kei".  "kei" tells the listener specifically, "I'm done with whatever abstraction I was in the middle of, let's go back to the next higher level of bridi to complete that."
 

>    2. How should overlapping places be interpreted?  E.g., in "<fe ti klama fa
>       mi ta>," does <ta> occupy the second or third place of <klama>?  Is this
>       sentence even grammatically correct?

It's grammatically correct but somewhat discouraged.  It's my understanding that the "ta" goes into the same position as the "ti", officially.  But if that's what you mean to say, and you'd like to be understood, you would be wiser to add an additional "fe", or much better yet to take care of all of your "fe"s at once, like normal, with ".e" or some other proper connector. :)
 

>    3. When joining two <sumti> with <joi>, it doesn't actually seem necessary to
>       terminate the former one with <ku>.  The justification given in the LRG
>       (§14.14) only applies to poorly written parsers without look-ahead
>       capabilities, which yacc-generated parsers almost certainly are not.
>       Unless there exists a construct which can be either a <selbri> or a
>       <sumti> simultaneously, is there any good reason why <ku> should be
>       required here?

I don't really have an opinion on whether those constructs should be legal, and I don't know why the grammar is written that way.  I guess it seems to me that the confusion which might result is that "joi" is used both to join sumti into masses, and to join tanru elements into mixtures.  You can say either:
 
lo blanu joi xunre cu klama -- a blue-and-red-combined-thing comes
lo blanu ku joi lo xunre cu klama -- a blue thing and a red thing considered together come
 
I suppose that a "ku" would be useful for clarity in the second case whether or not it was required.

>    5. Why can't an elidable terminator be elided when followed by a vocative
>       phrase (LRG, §6.11)?  As the meaning of a vocative is independent of its
>       location and has no real bearing on the meaning of the surrounding
>       <bridi>, what difference does it make whether or not the elidable
>       terminator is present?

Vocative phrases and UI cmavo can both be put almost anywhere.  In the case of UI cmavo it can matter whether or not there is a terminator-- it's generally valid to put the UI whether or not the terminator is there, but it can have slight variations of meaning.  I'm not sure if anyone's thought to care about those variations as far as COI.  COI cmavo are rarely used with a bridi at all these days, in my experience.  There has been some experimentation with COI recently: One trend, which I personally enjoy, is spelling "coi" as "cccccccccccccoi".  There's also a trend of combining attitudinals with COI, like ".ui nai co'o" or ".ue coi", sometimes even with "zei" like ".iu zei fi'i".  
 

>    6. When a <sumti> has internal <sumti> and/or a relative clause attached to
>       it, does its terminating <ku> (if present) belong before or after the
>       internal <sumti> or clause?

By internal sumti you mean "be" "bei" "be'o" linked sumti?  Things linked with "be" are part of the sumti description, and come inside of the "ku".
 
NOI linked clauses can attach in several locations:  "the dog which is blue eats":

lo gerku poi blanu ku cu citka
lo gerku ku poi blanu cu citka
lo poi blanu ku'o gerku cu citka

A NOI phrase should be thought of as another bridi, which is being connected to this bridi via a connection point at a sumti.  There's various ways you can attach it to the sumti, and it's really neither inside nor outside of the sumti-- it's another bridi, grammatically wedged in, which happens to share a referent with the higher-level bridi.  In the case of "noi" the other bridi doesn't even have a semantic effect on the higher-level bridi-- it just happens to be hanging out there.

>    7. Can the terminator "<vau>" be applied to discursive <bridi> introduced
>       with "<sei>"?

I don't believe so.  "sei" introduces a bridi-tail, not a full bridi.  It's the same structure as after the word "lo": you have to bring any additional sumti into the embedded bridi with "be" "bei" "be'o".  I wasn't around when these design decisions were being made, but to me the rationale seems pretty obvious: In cases where you're more likely to need a complex phrase (nu, noi), you get to add sumti for free, but you often have to pay the price of a terminator.  In cases where you're more likely to want to move on after saying the selbri (sei, lo), you get to move on to another sumti of the higher-level bridi for free, but you need to pay a few syllables to tie in an extra sumti to the lower-level bridi.
 
There is the terminator "se'u" for "sei", which is rarely needed but could help provide clarity. :)
  

>    8. How do you state that someone said a string of non-grammatical words that
>       include "<le'u>" but not "<zo>" or "<zoi>"?

zo zoi can quote anything.   zo zoi is all-powerful.  zoi gugliplapanukik. I can say anything in here except gugliplapanu.. oops! .gugliplapanukik. cunso cmevla mupli
 
In reality most Lojbanists are not computers, and use a wide variety of quoting methods both formally and informally. There is a tradition for instance of quoting Lojban in curly quotes: {broda}  which I believe originated in trying to disambiguate while giving definitions of words:  {mi} is "me" and {do} is "you"   is less confusing than   "mi" is "me" and "do" is "you"
 
There was a fascinating discussion about experimental correction cmavo between eimi and dbrock on IRC recently, which used lo'u/le'u extensively in quotes about quotes and hurt my brain. (to .io sai doi re za'e tcecre jbopre toi)  Generally speaking, though, most such complicated conversations still happen using mostly English and {[(graphcial quotes)}}.
 

>    9. After uttering an empty ZOI or LOhU construct, if one were to erase it
>       with <si>s, would it be necessary to utter a <si> for the empty quoted
>       text?

I'm going to say [pe'iru'e] no.  "si" erases words, and an empty space, while semantically meaningful, is not a word.
 
In practice, if you're erasing more than a simple word or two, I suggest saying "sa". :)
 

>   10. What pro-<sumti> should be used to refer to the speaker and only a few of
>       many listeners?

Tricky.  I would go with "mi'a".  Like if I'm talking to la .pav. and la .rel. and la .cib., and calling all three of them "do", and I'm using a pro-sumti for something that I'm doing along with la .cib., I think that "mi'a" would make sense because the "other than mi or do" referent of the "mi'a" 
is la .cib. goi cy., who is different as a referent from the current "do" even though cy. is a part of that "do".
 
In any even vaguely complicated situation I suggest introducing absolute clarity with "goi".  Lojban's assignable prosumti are far more versatile and stable than pronouns.  You can assign a prosumti to a tricky concept like "myself and those of you listening who like 
cucumbers", just throw that up into zo xy. or zo zy., and then have it at hand later-- how much later depending on how memorable you can make it to your listener-- so you can simply say "well, I think that zy. will like the salad today..." 
 

>   11. Can <finti> be used to say that one invented a physical device (which
>       would seem to contrast with its other translations)?  If not, what word
>       should be used?  If yes, how can one distinguish invention of a physical
>       device from composition of literature, music, et cetera?

I would say that [lo'e se finti] (the sort of thing that goes into the second place of "finti") is an idea.  It could be an idea about how to make a 
physical object, or an idea about what notes to play, or 
what words to put into a poem.  It would make sense to distinguish different kinds of invention by different lujvo built on "finti", and indeed we have been given a good word-ending rafsi "fi'i" for such a purpose.  dairfi'i, dacti zei finti, object invent, could perhaps mean to invent an object.  ca'arfi'i, cabra zei finti, device invent, could perhaps mean to invent a device.
 

>   14. I am quite certain that I once read somewhere that the spaces in Lojban
>       text are unnecessary and that words can be unambiguously determined from
>       the text without them; however, I cannot find where I read this (if ever),
>       and the issues raised in the next three questions seem to go against that
>       claim.  Hence, just to be clear, are the spaces in Lojbanic text optional
>       & unnecessary?  Do the claim that Lojban sounds can be broken up into
>       words in exactly one way and the audio-visual isomorphism of the language
>       together imply that written Lojban does not change when the spaces are
>       removed?  How does the answer to this interact with the below three
>       queries?

You need either spaces or accents in order for written text to be unambiguous.  You could alternately think of accents as being the "real" quality which must be marked, and note that if you put spaces between the words you can place the accents simply by putting them on the penultimate syllables.
 
You can write unambiguous Lojban text by capitalizing accented vowels:  "miprAmila.lOjban.noimElbibAnguzi'enoipubAngula.lOkadin."  This style is popularly associated with a Lojbanist who I believe is named "l0kadin". 
 

>   15. According to the Lojban Reference Grammar, chapter 4, section 9, item 5,
>       any monosyllabic <cmavo> before a <brivla> must be followed by a pause,
>       yet this would require an inordinate amount of pauses to be inserted into
>       text.  By "pause," does the LRG mean a <denpa bu>, in which case why has
>       no one been following this rule, or does it mean a space, in which case
>       how does this interact with Lojban's claims of audio-visual isomorphism
>       and that sounds can be divided into words in only one way?

This rule applies only if the cmavo in question is stressed.  In normal speech, cmavo are not stressed, but it's possible you might want to stress a cmavo for emphasis (in conjunction with "ba'e", pe'u) or clarity.  So if you stress a cmavo, you should also pause afterwards-- this helps with the emphasis, as well. :)
 

>   16. Section 17.4 of the LRG states that the spaces in <denpa bu> and <slaka
>       bu> are mandatory to avoid confusion with <fu'ivla>, yet the rules for
>       <fu'ivla> explicitly prohibit compounds of <gismu> & <cmavo>.  Why is the
>       former restriction there?

This has to do with accent, which the spaces indirectly represent. :)  "dEnpabu" and "slAkabu" are a gismu followed by a cmavo, while "denpAbu" and "slakAbu" are (unused) fu'ivla.
 

>  18. When the LRG states that all <brivla> are emphasised on the penultimate
>       syllable, is it merely repeating a consequence of the pronunciation rules,
>       or is it indirectly requiring that the penultimate syllable of a <brivla>
>       not contain a syllabic consonant or "y"?  If the former, wouldn't it be
>       simpler for it to just state that <brivla> must have at least two
>       syllables?

Brivla are the only words which are required to hold stress.  You can stress the first syllable of two-syllable cmavo if you want; it doesn't make any difference.  "y" is ignored for the purpose of locating stress in brivla.

 

>   19. Exactly what changes to the language have been made since the release of
>       *The Complete Lojban Language*/the Lojban Reference Grammar?  Does there
>       exist a completely up-to-date compilation of all of the rules of Lojban
>       grammar (that is at least more user-friendly than that horrendous tiki)?

Lojban is a living language.  Changes are made to it every day and then forgotten.  Many conventions are not officially recorded anywhere.  There is a bureaucratic structure which makes official rulings about the language; it does not represent everything that is happening in the community.
 
Here's the basics, off the top of my head.
 
xorlo.  That is not a gismu. :)  That is an English word which is a combination of "xorxes" and "lo", as I understand it.  It is supported by many Lojbanists and opposed by others, and has some vague half-official status the details of which escape me.  To the extent that anyone has any idea what exactly they mean when they say "le" and "lo" these days, what many people have in mind is xorlo, not the official rules.  The largest practical change is that people say "lo re xrula", the two flowers, and they don't intend to assert that there exist only two flowers in the universe, which used to be the idea.  xorlo is just part of a large ongoing dialogue about the meaning of "le" and "lo".
 
The Dot Side.  Those on the Dot Side believe that instead of forbidding the strings "la" "lai" and "doi" from names and allowing a name to follow those strings without a pause, names should be allowed to contain "la" etc and a pause should be required before all names.  The Dot Side is not nearly as established as xorlo, but is presently quite popular.
 
Those are the only large heretical movements that I'm familiar with.  Fill me in anyone if there's any others; I think it's fascinating.
 
There are also many experimental cmavo, and these days many in active use.  The most common are shorthand ways of quoting non-lojban names and words.  There are plenty of good experimental cmavo available and there are plenty of experimental jbopre who will give your ideas a go if you think of something interesting.  It takes a long time for things to become official, but in the meantime many of us accomplish unofficial communications. ;)

 

> I have more questions, but they can be dealt with later.  For now, the answers
>  to these will be just fine.

ki'e te preti
 
I think that questions are as useful as answers.  Keep asking anything that's unclear to you & you'll help us all to clarify our language and our explanations.
 

>  mu'omi'e la'o gy. Minimiscience .gy.
 
ke'u fi'i ninpre .i ma do lojbo cmene 
 
mu'o mi'e .bret.