criticism of lojban needed

[Gleki Arxokuna]

the wikipedia article about lojban might need a short list of criticism of lojban with links (e.g. to posts in this mailing list). Balanced criticism actually makes languages more popular, so it's advisable to make such a list.

Here are some thoughts:

1. gismu are hard to learn by everyone (as the algorithm mutilates the initial form of words)

2. za'a/ga'a/zgana/zga is hard to remember (would be better to have one root instead of 4)

3. abundance of kafkylerfu

4. still no machine translation

ke'u, links needed, the list might need more items

[TR NS]

LOL. Maybe there are other conlang lists that would be better to ask?

My thoughts:

* Somewhat high phonological density --nearly every possible CV and CV'V cmavo is used.

* gismu has no taxonomic (a priori) design at all

* sounds a bit like an Apache Polish American speaking atonal Chinese

Ok. That last one is subjective, but my point is simply that it is unlikely to rank high in a "best sounding" poll.

[John E Clifford]

What is the point here?  Listing problems is not a good way to sell a product.  Better to find a positive spin on these (the vaunted derivation of the gismus is the best that can be done with most of the points above -- it's equally bad for everyone and is based on some sort of connection to real languages).  Rather than saying there is no machine translation, point out that it is the only language (well, you know what I mean) that has a real possibility of a machine translation (actually, it would be nice to see a machine program to convert Lojban into actual formulae of logic, the supposed underlying goal, after all.)  Rather than complaining about the number of forms, stress the ease of figuring out new terms.  Rather than complaining about the density (not just for cmavo), stress the simple resolution of strings (and the richness of vocabulary of various non-central sorts -- carefully avoiding the significant duplications at the core).  Avoid aesthetics altogether (the oldest comparison is with Albanian of the Italian variety -- I don't know whether that is Gheg or Tosk, Apache makes the present comparison feel not quite right since Lojban is not all that gutteral or nasal).

(What is a fart letter?)

--
You received this message because you are subscribed to the Google Groups "lojban" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lojban+un...@googlegroups.com.
To post to this group, send email to loj...@googlegroups.com.
Visit this group at http://groups.google.com/group/lojban.
For more options, visit https://groups.google.com/d/optout.

[Gleki Arxokuna]

Wikipedia always mentions critics to any movies, as for languages then to Esperanto.

Good critics is also a selling point since otherwise people might get an impression that we are trying to ignore criticism.

I forgot that the most well known criticism is xkcd's comic. May be we should mention it in the first place. it is easily disproved ofc.

[Robin Lee Powell]

On Mon, Sep 01, 2014 at 12:07:01PM +0400, Gleki Arxokuna wrote:
> the wikipedia article about lojban might need a short list of criticism of
> lojban with links (e.g. to posts in this mailing list). Balanced criticism
> actually makes languages more popular, so it's advisable to make such a
> list.
>
> Here are some thoughts:

My troubles actually speaking the thing on a daily basis:

1. The total lack of idiom

2. The incredible self-similarity of the gismu makes keeping them
straight extremely difficult for me in practice (i.e. when speaking
at full speed).

3. *WAY* too many c/s/j sounds; so many accidental tounge twisters

--
http://intelligence.org/ : Our last, best hope for a fantastic future.
.i ko na cpedu lo nu stidi vau loi jbopre .i dafsku lu na go'i li'u .e
lu go'i li'u .i ji'a go'i lu na'e go'i li'u .e lu go'i na'i li'u .e
lu no'e go'i li'u .e lu to'e go'i li'u .e lu lo mamta be do cu sofybakni li'u

[Gaetano Cajetan]

Tongue twisters are fun if you ask me. ;-)

[Gleki Arxokuna]

2014-09-01 22:56 GMT+04:00 Robin Lee Powell <rlpo...@digitalkingdom.org>:

On Mon, Sep 01, 2014 at 12:07:01PM +0400, Gleki Arxokuna wrote:
> the wikipedia article about lojban might need a short list of criticism of
> lojban with links (e.g. to posts in this mailing list). Balanced criticism
> actually makes languages more popular, so it's advisable to make such a
> list.
>
> Here are some thoughts:

My troubles actually speaking the thing on a daily basis:

1.  The total lack of idiom

2.  The incredible self-similarity of the gismu makes keeping them
straight extremely difficult for me in practice (i.e. when speaking
at full speed).

3.  *WAY* too many c/s/j sounds; so many accidental tounge twisters

Tongue twisters can be applied to 2. as well, right?

In general I think your three points are more important than other suggestions.

Could you publish them on a wiki or on your website so that we can link to it from Wikipedia?

 

--
http://intelligence.org/ :  Our last, best hope for a fantastic future.
.i ko na cpedu lo nu stidi vau loi jbopre .i dafsku lu na go'i li'u .e
lu go'i li'u .i ji'a go'i lu na'e go'i li'u .e lu go'i na'i li'u .e
lu no'e go'i li'u .e lu to'e go'i li'u .e lu lo mamta be do cu sofybakni li'u

[ianek]

My two major complaints:

* it's hard to learn and remember all places of some gismu and lujvo, and their order
* terminators are hard to parse by humans when they're nested in several layers (vau kei ku vau vau)

mu' mi'e ianek

[ianek]

By the way, in order to write about critism of Lojban on Wikipedia with accordance to their rules, you need a source to cite.
http://en.wikipedia.org/wiki/Wikipedia:Verifiability
http://en.wikipedia.org/wiki/Wikipedia:What_Wikipedia_is_not
http://en.wikipedia.org/wiki/Wikipedia:No_original_research

[wixn...@gmail.com]

I do not like the source languages even though this would cause Lojban to be slightly more unlearnable; two-thirds of the source languages are from the Indo-European language family, and this underrepresents other languages, such as the Turkic and Uralic languages, for instance, Turkish, Azerbaijani, Finnish, Estonian and Hungarianz

[Wuzzy]

Am Mon, 1 Sep 2014 06:06:25 -0700 (PDT)
schrieb TR NS <tran...@gmail.com>:

> * gismu has no taxonomic (a priori) design at all

This doesn’t have to be a bad thing. The taxonomic approach has also
its own problems. For example, a taxonomic approach has the danger of
implying to much, making the vocabulary less flexible.

Lojban’s approach, making each word’s definition practically
independent from each other, makes it more flexible flexible.

So it is arguable if it is actually a bad thing to use no taxonomic
design.

> * sounds a bit like an Apache Polish American speaking atonal Chinese

LOL! Even if that would be true, I don’t see how this is even an argument?

[Wuzzy]

My personal (!) criticism on Lojban:

MAIN CONCERNS:
Concerning the CLL:
- The morphology is only vaguely defined, and it shows thanks to
various disputes about wheather some word X is correct Lojban.
- The morphology is not formalized at all. It is not part of the YACC
or any other grammar; the YACC ends at the word level.
- The alphabet is not formalized. The CLL assumes from the
reader to know of the latin alphabet (which is arguably
the case if you can read the CLL, as the CLL is only in English now.
But what if it gets translated into, let’s say, Japanese?)
- The alphabet is only briefly defined, and somewhat clumsily. The list
of letters is lacking the capital letters, although those are used in
Lojban.

Lojban itself:
- Lojban is actually not really a finished language:
- See CLL criticism
- There are still no dictionary-quality definitions for all
structure words (cmavo). I think this is a task that should
really get done at some day.
- The vocabulary is still rather small in my opinion and there
is still no published dictionary. I recognize that jbovlaste
<http://jbovlaste.lojban.org/> is the de facto dictionary
and that it grows every day, and I think it is a great tool.
But still Lojban’s vocabulary needs to be much larger. I
personally help on this task occassionally.

MINOR CONCERNS:
Concerning the semantics:
- Operator precedence is completely undefined, intentionally . I
recognize the justification for it, but the CLL leaves us again in
the dark how to negotiate an accepted operator precedence then.

Concerning answers to questions:
- An answer to a “mo” question (for example: “do mo prenu” = (roughly)
“What kind of person are you?”) is allowed to consist of just a single
predicate word (brivla), for example: “gleki”. But what if the
answerer did mean “gleki” as an independent assertion, NOT as an
answer to a previous question? So the example sentence “gleki” could
mean “I am a happy person” (as answer to the question) OR “Someone is
happy about something” (as an independent sentence which is
unrelated to the question).
- Even worse, there is not even a structure word to mark a sentence
explicitly as an answer or a non-answer.






Disclaimer: I am a Lojbanist myself somehow and I am probably biased as
hell. Nethertheless, I have some criticisms accumulated in the past
months. But please don’t blindly believe what I said above, I may be
wrong on some points.

And besides, I wonder if a mailing list would be an acceptable source
for Wikipedians, especially it it consists mostly of insiders. But on
the other hand I don’t really care about the “image” of Lojban, as long
as people stick to the FACTS.

[Wuzzy]

> I forgot that the most well known criticism is xkcd's comic. May be we
> should mention it in the first place. it is easily disproved ofc.

I am not even sure if Randall did it mean that way. I think he was
kind of ironic. The tooltip says: “zo'o ta jitfa .i e'o xu do pendo mi”

For reference, here’s the link:
https://xkcd.com/191/

[Gleki Arxokuna]

2014-09-02 23:55 GMT+04:00 <wixn...@gmail.com>:

I do not like the source languages even though this would cause Lojban to be slightly more unlearnable; two-thirds of the source languages are from the Indo-European language family, and this underrepresents other languages, such as the Turkic and Uralic languages, for instance, Turkish, Azerbaijani, Finnish, Estonian and Hungarianz

They are not underrepresented.

Note that gismu are based not just on 6 arbitrary languages but on how this or that phoneme or their combination is spread all over the world.

The languages you mention would just get such a small weight  that wouldn't affect the form of gismu.

I tested an algorithm with 4 - 12 major languages and found that first 4 languages is not enough, 12 languages is too many (although of course you can use them all to score new gismu to prove or disprove my claim).

[And Rosta]

On Monday, September 1, 2014 4:07:03 AM UTC-4, la gleki wrote:
> the wikipedia article about lojban might need a short list of
> criticism of lojban with links (e.g. to posts in this mailing list).
> Balanced criticism actually makes languages more popular, so it's
> advisable to make such a list.

Lojban's greatest success is this:

1. The founders of Lojban set themselves the absolutely overriding goal of creating a version of Loglan that is stable and has a community of users. This goal was achieved.

My main criticisms of Loglan/Lojban are:

2. Even if being a Whorfian experiment -- as Loglan not very credibly purported to be -- were of academic interest, the experiment design was so poor as to render it an utter failure as an experiment.

3. Even relative to the compartively simple task of creating a logical language, Lojban does an exceptionally poor job. There are no formal rules that map a sentence's phonological form to its logical form, and to a large but very slowly diminishing extent there are no informal rules that do that either. There are ways to unambiguously encode logical forms in Lojban sentences, but these are clunky, verbose and unergonomic (and therefore largely unused); and anybody giving the problem twenty minutes' thought could have come up with a better design than Lojban's.

4. The morphosyntax of Lojban is full of unnecessary baroque complexity -- the proliferations of allomorphy, word-classes, constructions, function words could also be very drastically simplified.

5. Goal (1) is not a very interesting goal: having a stable language with a community of users is interesting only if the language itself is worthy of being a stable language with a community of users. Criticisms (2) and, in my view, particularly (3) and (4) mean the language itself isn't worthy of being a stable language with a community of users. That is, there are no reasons why it is a Good Thing to learn and use Lojban; any attempt to persuade people otherwise would in my view by deluded or dishonest. (A "Good Thing" is something more than just "whatever floats your boat".)

I've written at length about (3) and (4) elsewhere, on Lojban and Conlang lists.

--And.

[John E Clifford]

Yup!  If your goal isn't monoparsing, you have no reason to be interested in Lojban/Loglan.  If your goal is monoparsing, Lojan/Loglan may be the only living option but its success is not proven and, even if it were, it does just about everything in the worst possible way.

Let's see how that line of objections, rather than ones to the cosmetics, can be met and turned into a positive discussion of Lojban.

--
You received this message because you are subscribed to the Google Groups "lojban" group.

To unsubscribe from this group and stop receiving emails from it, send an email to lojban+unsub...@googlegroups.com.

[v4hn]

On Wed, Sep 03, 2014 at 08:44:55AM -0700, 'John E Clifford' via lojban wrote:
> Let's see how that line of objections, rather than ones to the cosmetics, can be met and turned into a positive discussion of Lojban.

I'm sorry, I might have missed something here,
but "It's all crap and anyone could do better" is no valid criticism pe'i


mu'o

[Bob LeChevalier, President and Founder - LLG]

On 9/3/2014 11:44 AM, 'John E Clifford' via lojban wrote:
> Yup! If your goal isn't monoparsing, you have no reason to be
> interested in Lojban/Loglan. If your goal is monoparsing, Lojan/Loglan
> may be the only living option but its success is not proven and, even if
> it were, it does just about everything in the worst possible way.
>

> Let's see how that line of objections, rather than ones to the
> cosmetics, can be met and turned into a positive discussion of Lojban.

I think that there are a few more reasons, though most of them are not
phrased in academic terms. Certainly, people have BECOME interested in
Lojban for other reasons and have found their interest to be sustained,
even as they learned that Lojban isn't hardly a perfect design.

Objections to Lojban should be separated into critiques of the language
design as a fulfillment of its academic goals, and critiques of the
design pertaining to its learnability and/or aesthetics.

Since the goal is to produce something that can be linked to by a wiki
article, I suggest creating a page in the Lojban wiki that addresses
these criticisms and organizes them in a list (with rebuttals as
appropriate).

lojbab

[John E Clifford]

Well, that is a summary of nearly 60 years of discussion.  It is a valid criticism, but one that would usually affect one's view only if more detailed points were offered (we need a good bibliography of all that discussion).  Note also that it is not said to be all crap -- it does do a large part of what it set out to do -- only that it hasn't provably done all that it set out to do.  Nor is it said that anyone could do better.  Some have done some parts better, most who have looked carefully have seen that certain things are unnecessarily complex or even off course -- suggesting that better is possible, and there have been many suggested improvements never worked out in detail (and officially not to be even discussed -- though they have been, of course).  In short, someone -- or, more likely, some group -- could do better, since the Lojbanist group is committed to not doing so.  

[selpa'i]

la .van. cu cusku

If the goal is to make a language that unambiguously encodes logic, then
that's relatively easy to achieve. See xorban for one model that doesn't
add too much unnecessary baggage. Lojban does achieves it too to some
degree, but it has so much extras that most of the language is still
undefined. We cannot easily convert Lojban to logic due to that.

If the goal is to make a language that is supposed to be used by people
in normal everyday discourse, then Lojban does a lot better than xorban
(the latter being rather heavy on the mental stack), but even Lojban can
be difficult to parse when there is a bit of nesting, as someone else
noted.

If the goal is to make a logical language that is beautiful or good for
poetry, then it becomes more difficult to measure the extent to which it
succeeded. Rhyming works relatively well in Lojban. You can write poems
and even rap music. Xorban can't really do rhymes, because every word
has the same ending (-V(kV)*).

Then you can also look at flexibility. Here Xorban does the worst of all
the loglangs I know. Lojban and Gua\spi are equally flexible, Lojban
probably being a bit better.

And what about simplicity? Here Gua\spi and Xorban are the clear
winners, closely followed by Toaq Dzu, and far in the distance comes
Lojban. Just compare the sizes of their grammars to get a rough idea.

I don't have enough information about Livagian, but it, too, seems too
difficult for normal human beings.

And you can have a number of other goals. Verbosity for example. Lojban
is extremely verbose (partly because all the predicates have at least
two syllables). Xorban is verbose, but much more efficient in its
encoding of logic, and is the only language that I know which perfectly
handles co-filled argument places (though this is also what makes it so
hard to speak). Gua\spi seems to fare slightly better than Lojban, but
it has its clumsy parts. Toaq Dzu to my knowledge is the most succint
loglang.

I could write more, but this should be enough to show how hard it is to
say whether or not a language succeeded.

mi'e la selpa'i mu'o

[John E Clifford]

There are other reasons advertised to be interested in Lojban and some of those reasons have a certain legitmacy.  The point here is just that, for any reason other than monoparsing, there are other conlangs that answer to that interest and are much less complex than Lojban, so that interest could be pursued more efficiently elsewhere.  That is a mere practical point and does not answer to the emotional or aesthetic appeal of Lojban, so the claim is stated a little to broadly (but it was a summary, after all).  

I suppose the first objection to your way of dealing with objections is to point out that Lojban's academic (or any other for that matter) goals have never been explicitly spelled out (and, indeed, attempts to do so have been explicitly rejected), so learnability and aesthetic considerations (what would those be, anyhow) may very well be among those goals, even major ones, whatever others might think.  However, I will stipulate that one academic goal is orthoparsing -- monoparsing that always gives the right parse.  Lojban/Logan is unique among widely discussed conlangs in having this goal (and most alternatives are overtly descended from some phase in L/L history).  It claims to have achieved monoparsing but the next step has never been explored or even well-formulated, that the correct parse accurately reflects the intended proposition.  Proving that would presumably take the form of an algorithm for converting a parse into a unique formula.  Beyond that, it is not clear what to put as goals: SWH is largely defunct (or reduced to the most rivial of claims, about which L/L clearly has nothing to say).  With it, both cultural neutrality and improving your thinking go as targets, except in some diffuse way that would apply equally to learning Chinese or differential equations.    As for learnability, there has never been an actual test of the claims about the complicated system of primitive terms and the anecdotal evidence is at best a mixed bag (every positive matched with an equal negative).  For the rest, the sheer bulk of particles of minutely different meanings (and in a packed word space) bodes ill for ease of learning (the lack of competent text books has also always been a problem).  Aesthetics is just too sloppy to be worth talking about; subjective problems are personal and, if someone doesn't like something, he can just avoid it.

[And Rosta]

This is an important point. If you restrict the set of candidate languages only to living ones (i.e. with speaker community), then despite its many faults Lojban is patently the best choice if you want what you aptly term "monoparsing".

More generally, most of the PR claims for Lojban are valid if and only if the set of candidate languages is restricted to living languages. (Which restriction is not unreasonable.)

--And.  

To unsubscribe from this group and stop receiving emails from it, send an email to lojban+un...@googlegroups.com.

[John E Clifford]

"living", i.e., being reasonably well-documented and having a community of reasonably competent users, may be an overly narrow class, depending on your aims. People seriously looking for what Lojban purports to offer might be willing to to look for something with a good start on a description and a very small active group and then build on that (cf.  Esperanto in the face of Volapuk).  The alternatives at the moment are to stick with Lojban in the face of its promise not to change and so to remain deeply flawed or to start from scratch or close to it, though with a the data from LoCCan and its offshoots -- and probably others -- to build on.  Is there no less well-known conlang that has not yet developed a clearly wrong turn but is developed enough -- and open enough -- to form a reasonable platform for development?  I don't know the present field well enough to have an answer to that.  Those I do know (all less than thoroughly) have nice points and apparent (to me) deep flaws, but are not yet so fixed that correction is impossible.  But, on the other hand, I am not inclined (nor able) to wait another 60 years for the sort of language envisioned to come along and the record shows that getting the sort of grunt work that creating such a language will require makes the 60 years from a rough start seem optimistic. So, on a third hand, perhaps getting Lojban to change even before it has a complete record of how bad it is is the best option -- though not a very promising one in the present climate.

[Bob LeChevalier, President and Founder - LLG]

On 9/3/2014 12:30 PM, selpa'i wrote:
> la .van. cu cusku

> If the goal is to make a language that unambiguously encodes logic, then
> that's relatively easy to achieve. See xorban for one model that doesn't
> add too much unnecessary baggage. Lojban does achieves it too to some
> degree, but it has so much extras that most of the language is still
> undefined. We cannot easily convert Lojban to logic due to that.

I need to note here that if the goal is merely to encode logic, then
Lojban can probably be called "successful" in that I believe that
anything expressible in logical notation can probably be represented in
Mex (the operators needed for any given notation have not been defined,
but Mex is defined so as to allow innumerable sets of operators as well
as precedences).

So far as I know, no one in the community is interested in such a narrow
goal.

> And what about simplicity? Here Gua\spi and Xorban are the clear
> winners, closely followed by Toaq Dzu, and far in the distance comes
> Lojban. Just compare the sizes of their grammars to get a rough idea.

Of course part of the problem is that very few such languages have been
USED to the extent that Lojban has. Feature growth has come from usage.

lojbab

[John E Clifford]

Bob's reply is somewhat disingenuous.  MEX can reproduce a logical formula exactly, I'm sure, but the goal is to produce *sentences* whose logical structure are transparent both at a glance and provably after analysis.  And still be a functioning human language.  Loglan is apparently the latter, the question remains whether it is the former.  There is, by the way, no doubt that this is and has been the goal form at least a significant portion of Lojbanist and Loglanist since the start (albeit somewhat vaguely worded in the beginning).

[John E Clifford]

Attached is a pdf version of a start to such a list.  Additions and corrections eagerly sought.  

[Pierre Abbat]

On Saturday, September 06, 2014 07:09:03 'John E Clifford' via lojban wrote:
> Attached is a pdf version of a start to such a list. Additions and
> corrections eagerly sought.

Correction: "dl" and "tl" are permitted in Lojban if the two consonants are in
different syllables, but not if they are both in the onset of the same
syllable. E.g. "fatlutfa'i", "badle'o". "dl" is a permitted initial in
Russian; I don't know for sure about "tl".

Irregularities in the list of permitted combinations and permitted initials:
"ms" is permitted but "mz" is not; "cm", "cn", and "jm" are permitted initials
but "jn" is not; similarly "zn".

Some pairs of gismu differ by exchanging two letters and are similar in
meaning: ractu/ratcu, stuzi/stizu, garna/grana. Other groups of gismu differ in
one letter and are similar in meaning: carce/marce/karce, culno/mulno.

Some sets of cmavo are phonetically related (zi, za, zu); others are not
(du'a, be'a, vu'a, ne'u). The ko'a-series and fo'a-series are phonetically
related but should not be, as within each series, they differ in the unstressed
vowel, making it easy to mistake which sumti one is referring to.

The symmetry of the logical conjunction matrix is broken by the sentence
separator.

Pierre
--
La sal en el mar es más que en la sangre.
Le sel dans la mer est plus que dans le sang.

[Jorge Llambías]

On Sat, Sep 6, 2014 at 7:58 PM, Pierre Abbat <ph...@bezitopo.org> wrote: 

Correction: "dl" and "tl" are permitted in Lojban if the two consonants are in
different syllables, but not if they are both in the onset of the same
syllable. E.g. "fatlutfa'i", "badle'o". "dl" is a permitted initial in
Russian; I don't know for sure about "tl".

Random comment: If I remember correctly, every permissible CC initial pair is represented in gismu, but unfortunately not every permissible medial is. There are several gismu with tl (catlu, patlu, nutli, ...) but I think there are none with dl.

It would be interesting to know how many and which valid medial CC are not represented in gismu.

mu'o mi'e xorxes

[mukti]

It would be interesting to know how many and which valid medial CC are not represented in gismu.

I count 189 valid medial consonant clusters, 29 of which do not apear in any of the 1,010 canonical CVCCV gismu:

• bg
  
• bz
  
• dl
  
• dm
  
• fj
  
• fm
  
• fp
  
• fx
  
• gm
  
• gz
  
• jf
  
• jg
  
• jk
  
• jp
  
• jt
  
• jx
  
• kj
  
• lz
  
• mv
  
• pf
  
• pj
  
• px
  
• sf
  
• tj
  
• xj
  
• xm
  
• xp
  
• xt
  
• zv 

Here are the eight most frequent medial CC, i.e. the ones which each appear in 20 or more gismu:

• nt 29
  
• nr 28
  
• ns 25
  
• nc 24
• nd 24
• rn 24
• nj 22
  
• nl 20
  

Check my math?

mi'e la mukti mu'o

[Gleki Arxokuna]

except that voiceless+voiced can't lamji.

{vedli} is now more or less popular.

 

Here are the eight most frequent medial CC, i.e. the ones which each appear in 20 or more gismu:

• nt 29
  
• nr 28
  
• ns 25
  
• nc 24
• nd 24
• rn 24
• nj 22
  
• nl 20
  

Check my math?

mi'e la mukti mu'o

--

You received this message because you are subscribed to the Google Groups "lojban" group.

To unsubscribe from this group and stop receiving emails from it, send an email to lojban+un...@googlegroups.com.

[mukti]

except that voiceless+voiced can't lamji.

 

The program I used to generate the clusters counted "j" as a sibilant, but not as voiced. If I add "j" to the voiced consonants, 10 unused clusters go away, leaving 19 unused out of 179 total medial clusters:

• bg
  
• bz
  
• dl
  
• dm
  

• fm
  
• fp
  
• fx
  
• gm
  
• gz
  

• jg
  
• lz
  
• mv
  
• pf
  
• px
  
• sf
  
• xm
  
• xp
  
• xt
• zv 

{vedli} is now more or less popular.

I only checked against the 1342 gismu in the baselined list.

 

[Jorge Llambías]

On Sun, Sep 7, 2014 at 2:42 AM, mukti <shun...@gmail.com> wrote:

 

The program I used to generate the clusters counted "j" as a sibilant, but not as voiced. If I add "j" to the voiced consonants, 10 unused clusters go away, leaving 19 unused out of 179 total medial clusters:

• bg
  
• bz
  
• dl
  
• dm
  
• fm
  
• fp
  
• fx
  
• gm
  
• gz
  
• jg
  
• lz
  
• mv
  
• pf
  
• px
  
• sf
  
• xm
  
• xp
  
• xt
• zv 

Interesting, thanks. Three of them (jg, sf and zv) are valid initials, so that means these three are the only ones that occur as initial CC (jgari, sfasa, zvati) but not as medial CC.

[Philip Newton]

On 7 September 2014 00:58, Pierre Abbat <ph...@bezitopo.org> wrote:
> "dl" is a permitted initial in Russian; I don't know for sure about "tl".

It is also permitted, e.g. "tlenie" (decay).

mu'o mi'e .filip.

[And Rosta]

Bob LeChevalier, President and Founder - LLG, On 04/09/2014 00:43:

> On 9/3/2014 12:30 PM, selpa'i wrote:
>> If the goal is to make a language that unambiguously encodes logic, then
>> that's relatively easy to achieve. See xorban for one model that doesn't
>> add too much unnecessary baggage. Lojban does achieves it too to some
>> degree, but it has so much extras that most of the language is still
>> undefined. We cannot easily convert Lojban to logic due to that.
>
> I need to note here that if the goal is merely to encode logic, then
> Lojban can probably be called "successful" in that I believe that
> anything expressible in logical notation can probably be represented
> in Mex (the operators needed for any given notation have not been
> defined, but Mex is defined so as to allow innumerable sets of
> operators as well as precedences).
>
> So far as I know, no one in the community is interested in such a
> narrow goal.

For lots of people in the community that goal is the bare minimum and then success is measured by how ergonomically the logic is encoded. Lojban can encode logic, but does so extraordinarily badly. And as Selpa'i notes, most of Lojban doesn't unambiguusly encode logic.

>> And what about simplicity? Here Gua\spi and Xorban are the clear
>> winners, closely followed by Toaq Dzu, and far in the distance comes
>> Lojban. Just compare the sizes of their grammars to get a rough idea.
>
> Of course part of the problem is that very few such languages have
> been USED to the extent that Lojban has.

Part of what problem? The problem of Lojban being so unnecessarily complex? Lojban is not more complex because it has been used. It is more complex because its creators and leaders never cared much about simplicity.

--And.