--
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to loj...@googlegroups.com.
To unsubscribe from this group, send email to lojban+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.
Capitalization: meh. {la} and {.i} and {ni'o} are better than capitalization in my book.
Soulless: I couldn't disagree more. Lojban is lively; it's full of life and zest. Attitudinals can be dropped in anywhere to spice up a sentence; flexibility of syntax keeps things mixed up plenty. English loses some of its *grammatical* soul because it has so many restrictions that pop up mid-sentence. It doesn't conform rigidly to rules but if you don't conform reasonably to a rather extensive set of rules, you get a sentence that is at best awkward and at worst severely ambiguous. ("I used one of the things that Joe proved completely incorrectly" is a real world example, which actually offended "Joe" until I explained.)
Unemotional: again, no. The lack of ambiguity in Lojban is in its syntax and grammar; a phoneme stream parses in exactly zero or one ways. It is not necessarily *read* in exactly zero or one ways. This is a fundamental distinction between *ambiguity* and *vagueness*. Not all meaning must be specified in Lojban, but what meaning is specified cannot be interpreted entirely differently from how it was intended by the speaker. Hence Lojban can (and very very often is) vague, but cannot be ambiguous. I think your love of ambiguity is probably actually a love of vagueness.
Whether we need it or not: No one ever claimed that we needed it. It's perhaps been claimed that it can be useful, but I'd say the only thing which is outright universally asserted is that its students enjoy learning it (which is of course circular, since few would study a conlang for very long if they weren't interested.)
Hebrew and Arabic aren't capitalized; both have letters which take different
forms at the end of a word, which English doesn't. Arabic is important enough
that much of Lojban vocabulary is derived from it, usually mixed with the
other source languages. Georgian (kartuli, not what they speak in Atlanta)
used to be capitalized, but the modern script is monocase.
> I think English is awesome. All the other languages make quite a bit
> of sense, but English, no, English is like the badass language. It
> doesn't conform to rules. It makes no sense at all. Like, for
> example, a while ago I discovered that an alternate spelling for
> "Medieval" is "Mediaeval." How awesome is that?!? Seriously, that's
> so cool, when you spell it like that, it seems like you're actually
> some mediaeval knight or something who spells things weird.
In Spanish, words ending in "-ma" of Greek origin are usually masculine. The
Greek words are neuter, and neuter usually merged with masculine (in some
words, like "folium", the plural was interpreted as a feminine singular).
Exceptions and apparent exceptions are:
la paloma (dove) - not Greek
la broma - Greek means "food" but the Spanish meaning is totally different
la coma (punctuation), el coma (unconsciousness) - both from Greek neuters
la diadema (diadem) - Greek is neuter and means the same, so I have no idea
why!
Spanish spelling has been more stable than French, though they both have
academies, which English does not. There are ways of writing French words
that look mediaeval, such as "fresne" (modern "frêne"). The spelling "fresne"
is archaic, but the name "Dufrêne" is still commonly spelled "Dufresne", as
the name "Ashe" in English still has the final e.
Lojban has less than a century of history. Don't expect mediævialitye from it.
It does, though, have lots of words with more than one equivalent form, e.g.:
citrai=ci'onrai=citytraji=3 others
fiprgado=finprgado
There are also at least two words with variants: one is "trixexo/trixexu", and
I forget what the other is.
Pierre
--
The Black Garden on the Mountain is not on the Black Mountain.
--
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to loj...@googlegroups.com.
To unsubscribe from this group, send email to lojban+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.
So don't use it. Why are you here?
--
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to loj...@googlegroups.com.
To unsubscribe from this group, send email to lojban+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.
Or someone trying too hard to be funny :) Didn't work that well, oh well.
--
Ecce Jezuch
"I think I'll lie here for a while.
Then, after that...
I think I'll lie here for a while." - Garfield
----------------------------------------------------------------
Skoda Octavia z 2009 roku juz za 11 tys. zl
http://linkint.pl/f2a36
(just answered to see if you guys like the fu'ivla {tro'olo}. Any
other suggestion?)
Unless we are all Misters Spock here, it is very likely you already
noticed that verbal ambiguity is a boon, at least in some situations :
some well designed sentence can be as a rock that hits two or three
goals at the same time, and it also allows us to imply meaning to
somebody else without being too direct or even, offensive.
Ambiguity is good !
Of course, and Lojban uses it *all the time*.
Where did you get this bizarre idea that Lojban is free of semantic
ambiguity?
-Robin
(btw, ambiguity exists in other ways that the semantic one)
{.ui klama}
"Yay! Going!", approx.
Perhaps said at the beginning of a road trip? Who is
going/travelling? Where to? Why is the speaker happy, exactly? we
don't know.
> (btw, ambiguity exists in other ways that the semantic one)
Of course. If there's a particular kind of ambiguity you have in
mind, feel free to give an English example.
-Robin
You can have ambiguity in all of these.
Incidentally, syntax is itself connected to lexical matters and
morphology, and these are connected to phonology (some theories mix
them in a whole), so you could also find amibguities in those.
The example you gave is anaphoric ambiguity. Sure Lojban has
anaphoric ambiguity, its logical-predicate-structure allows it. Yet
I tend to think a *lexical* ambiguity would be more rarely produced,
-- if ever.
--esc
I've seen ambiguity used in English in nefarious ways a lot; so
perhaps in some of the cases you are talking about losing the
capacity for ambiguity the benefits outweigh the costs. Your
argument may also have some relevance as to whether Lojban will
succeed as an international auxiliary language -- Esperanto
probably has the features you want (though, I admit that I
don't know much about it) --, but I am interested in language
as a means of expression; if nothing else, the non-ambiguity of
Lojban forces one to think more clearly; in my case, it also
allows to make precise distinctions that are had to make in
only a few words in English, even if I "cheat" in English by
using underlining to indicate the proper groups to be lexed
together. And one should fairly notice if they read my
writings that my use of English punctuation would have most
English teachers pulling their hair out. Lojban makes it much
simpler to express my more complex thoughts clearly.
EDIT: should fairly notice ->
should fairly quickly notice
I like to say that Lojban is never ambiguous, but often vague.
(The intended distinction being that you can't have multiple possible meanings without including the region between them; the vagueness is among meanings which form a relatively natural category, as opposed to ambiguity like the three unrelated parses of "Time flies like an arrow".)
--
Kevin Reid <http://switchb.org/kpreid/>
Yet, I don't think there is purely lexical ambiguity, as there is in natlangs
See also Kevin reid's reply. Ambiguity is discrete (e.g. quantifier scope ambiguities). Vagueness is continuous. (this is in terminology as used in linguistics.) Vagueness is inescapable and often desirable. Lojban is not ambiguous, and that is once of its chief attractions, since the range of occasions when we actually want time-flies-like-an-arrow or someone-loves-everyone ambiguity (e.g. poetry) is much smaller than the range of occasions when we want to avoid it.
On 27 Sep 2011 10:31, "Robin Lee Powell" <rlpo...@digitalkingdom.org> wrote:
On Tue, Sep 27, 2011 at 11:26:23AM +0200, Escape Landsome wrote:
>
> Unless we are all Misters Spoc...
Of course, and Lojban uses it *all the time*.
Where did you get this bizarre idea that Lojban is free of semantic
ambiguity?
-Robin
--
You received this message because you are subscribed to the Google Groups "lojban" group.
To po...
Are you sure ?
--
You received this message because you are subscribed to the Google Groups "lojban" group.
I'm sure for myself and for most people; and even in poetry I more often find myself, when writing it, as opposed to reading it, striving to forestall ambiguity or misreading of ambiguous phrases than to capitalize on it. Of course, someone might perfectly legitimately prefer a language that is ambiguous, but given that we have 6000 ambiguous languages and 1 unambiguous one (Lojban), it seems a bit unreasonable to criticize Lojban for providing what the other 6000 don't.
--And.
> --
> You received this message because you are subscribed to the Google Groups "lojban" group.
Tanru aren't ambiguous; they're merely vague. When instances of genuine ambiguity have been found, there's been consensus among the logical language lojbanists (as opposed to Usage Decides lojbanists) that there needs to be an interpretation rule that gets rid of it.
On 27 Sep 2011 18:46, "John E. Clifford" <kali9...@yahoo.com> wrote:
Remember that Lojban claims to be ambiguity- free in only two respects: every Lojban utterance can be uniquely decomposed into words, whose syntactic type is fixed, and every sentence that passes the official grammar has only one parse tree. This not a trivial claim, since most languages, including much simpler conlangs, do not have either of these properties. But when we move on to semantics, this is not enough to guarantee unique readings. Lojban is clear on transformation involving conjunction relations between two or more sentences with common elements, but not so much on deep structure below the sentence level. In particular, tanru are opaque, ambiguous, though other, less common, structure are as well.
Sent from my iPad
On Sep 27, 2011, at 11:20, And Rosta <and....@gmail.com> wrote:
> Escape Landsome, On 27/09/2011...
> --
> You received this message because you are subscribed to the Google Groups "lojban" group.
> ...
--
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to loj...@googlegroups.com.
To unsubscribe from this group, send email to lojban+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.
I've just noticed G.groups new face, nice!Now, that you mentioned it... Lojban has a very big problem with its morpho-semantics; even Arabic is more precise there.I've been working on an Arabic translation for the Lojban dictionary, but it's taking too long and I have school, my Arabic-orthography also got rejected... All in all, I'm willing to participate in a big part-time chunk to a project I've been considering for some time...How would the Lojban community accept a merging between Lojban's syntax and Ithkuil's morpho-semantics?
I'll participate as much I can.If you don't know Ithkuil here's a link.
--
You received this message because you are subscribed to the Google Groups "lojban" group.
To view this discussion on the web visit https://groups.google.com/d/msg/lojban/-/Lus2nvgDgeoJ.
pau la kres. mo do
If I ask "Who is Chris?", I get an identity when what I am
asking for is a relationship; ie, friend, friend's child,
child's friend, friend of friend (note that "child" here is to
be interpreted as immediate descendant, not any non-adult).
However, asking the question "How is Chris related to you?", I
get the answer "He is not" because I am being mis-interpreted
as asking about a _familial_ relationship.
So you got any bright ideas for that?; this is most decidedly a
case where the ambiguity of English is something I want to
_avoid_.
BTW, I am open to suggestions from anyone regarding the problem
I have discussed; I am just particularly being pointedly
critical of the poster I referred to in the first sentence.
> To the one who thought English is the best thing in the world:
> Can you give me a dynamic equivalent translation in to English
> of a question that I often want to ask (though with different
> first sumti) to English-speakers?
>
> pau la kres. mo do
>
> If I ask "Who is Chris?", I get an identity when what I am
> asking for is a relationship; ie, friend, friend's child,
> child's friend, friend of friend (note that "child" here is to
> be interpreted as immediate descendant, not any non-adult).
> However, asking the question "How is Chris related to you?", I
> get the answer "He is not" because I am being mis-interpreted
> as asking about a _familial_ relationship.
"Who is Chris to you?"
--
You mean the Arabic orthography you proposed? Rejected by whom?
By none... That's the problem! And on that, I found that while Arabic letters give Lojban a pleasant look, it's not fast enough in writing since Lojban uses separate vowels while Arabic is an abjad.
You mean the Arabic orthography you proposed? Rejected by whom?
--
You received this message because you are subscribed to the Google Groups "lojban" group.
To view this discussion on the web visit https://groups.google.com/d/msg/lojban/-/Dlnc5-wi2BgJ.
By none... That's the problem! And on that, I found that while Arabic letters give Lojban a pleasant look, it's not fast enough in writing since Lojban uses separate vowels while Arabic is an abjad.
You mean the Arabic orthography you proposed? Rejected by whom?
lo ni'aripycrida cu mintu lo crida noi cnita xabju lo cripu
--
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to loj...@googlegroups.com.
To unsubscribe from this group, send email to lojban+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.