[lojban-beginners] Re: A Newbie's First Impressions

[Steve Pomeroy <steve@staticfree.info>]

On Tue, Apr 22, 2003 at 11:37:19AM -0700, Travis Garris wrote:
> I've run across Lojban before, and I like the idea
> behind it.  I've just recently started learning it.  I
> suspect that there are lots of programmers in the
> Lojban community, and that we are probably all drawn
> by the same thing.  I'm just another programmer.

I'm quite fond of the language too, and am also a programmer. It's a nifty
idea being able to communicate in a regular grammar'd language.

> First off, I have to say that learning a language in
> silence is quite hard.  Yes, I don't know anyone who
> ...

True. Though it's far from someone else to talk to, I've found that the
text-to-speech is good at helping sound out things:
http://www.lojban.org/wiki/index.php/Lojban%20diphone%20speech%20synthesizer

Of course, if you're actually interested in poking at this, you'll have to
poke mightily hard at festival to make it go, but it's worth it. (the
project is still in its alpha stages. nothing has been compiled/packaged up
for mass consumption just yet)

> One of my problems in this area is the following words
> used as examples in "Logban for Beginners":  nis. and
> porc.  I would expect "nis." to be pronounced as
> "knee".  It is supposed to be Nice, France.  J'ai
> ?tudi? fran?ais for duex ans, and I thought that Nice
> was pronounced as "niece".  How to represent that in
> Logban, I have no idea.

As was stated before, the "s" is not silent. 

> This leads into my next observation:  why does a
> language that prides itself on being unabmigious have
> a Zen-like approach to verb tense.  To say "mi klama"
> in no way tells you when that happens.  I can
> understand allowing such cases to exist, but the
> beginner's guide paints a picture that this neutral
> case is cultrually preffered. So far this baffles me.

Ah, you confuse syntactic ambiguity with semantic ambiguity. When
people say that lojban is unambiguous, they generally mean that the
grammar is unambiguous. One sentance has one and only one meaning. You
don't get sentences as you do in english: "You work in theory" where
pauses and intonation can turn a simple statement of fact into a
serious insult.

Lojban allows ambiguity in the sense that you don't always provide all the
information about something up-front. In your example, depending on
context, "mi klama" could mean "i went" or "i will go". This, though, is
irreleveant as it never specificly means any of those; it only indicates
that something actually went/will go/is going and that something is you.

> There are a few other things that seem out of place. 
> I like that each number uses each of the five vowels
> in turn, but they use apparently random consonants. 
> And I find myself pronouncing "xa" and "ze" very
> similarly.

hm, I'm not sure I see how you are pronouncing those. I pronounce them:

xa as in "hot dog" if you were clearing your throat at the same time as
saying "hot"
 
ze as in zephyr

> While conversational Lojban (atleast elementary) can
> be littered with articles, I love the power in fa, fe,
> fi, fo, fu.  Any sentence can be turned into a
> procedure or function call.  This is what appeals to
> programmers.

Well, in a way every sentence /is/ a procedure. Even without those
place markers. Your example below is a good one - the "function" call is
the selbri ("dunda"). What you're missing, though, is that you don't /need/
to start with the selbri in order to parse it. In fact, there's a parser
that can handle all of the lojban sentences you've written so far. Look up
"jbofi'e" if you care to poke at it. 

> "dunda fa la djan. le cukta la klaudias."
> 
> This may not be cultrually sound in Lojban, it is a
> grammatical challenge in English.  You run the risk of
> confusing any parser with the verb conjugation and
> prepositions required.  When all sentences are put in
> this passive, verb-first format, a computer could very
> easily parse Lojban text (sorry for not using bridi,
> selbri, or other Lojban-specific terms, I'm not
> confident yet).

Parsing English just sucks. Ask... well, anyone who knows what a parser
is. Lojban, being so regular that there's a BNF grammar for it, is very
easy to parse. See the above. 

> I do have a question.  I may be getting ahead of
> myself, but the lessons so far haven't specifically
> spelled this out, so I wasn't sure how it was done.
> 
> "mi dunda fi la klaudias. fe xo ma"
> 
> Can I use "xo" and "ma" together, and does their order
> matter ("ma xo")?  And before anyone yells foul, I
> haven't been exposed to any Lojban culture outside of
> the beginner's guide, so saying "I gave Claudia how
> many what?" works to direct the listener's attention
> better than "I gave how many what to Claudia?"

You can use them together. You'd probably respond: 

pano lo mlatu

and if not, then you obviously don't love Claudia enough. 

Generaly, treat the question words as though you're asking someone to
"fill-in the blank"

Things get strange when you start using multiple xu's in the same
sentence, but that's apparently legal.

> Travis Garris
> Durham, NC, USA

-Steve Pomeroy
http://staticfree.info/