There are 7 messages in this issue.
Topics in this digest:
1a. Re: Related to the recent discussion about counting the number of po
From: George Corley
1b. Re: Related to the recent discussion about counting the number of po
From: Matthew George
1c. Re: Related to the recent discussion about counting the number of po
From: George Corley
1d. Re: Related to the recent discussion about counting the number of po
From: MorphemeAddict
1e. Re: Related to the recent discussion about counting the number of po
From: Logan Kearsley
1f. Re: Related to the recent discussion about counting the number of po
From: George Corley
2a. Re: CHAT: the Ithkuil of programming languages?
From: Logan Kearsley
Messages
________________________________________________________________________
1a. Re: Related to the recent discussion about counting the number of po
Posted by: "George Corley" [email protected]
Date: Tue Feb 26, 2013 12:55 pm ((PST))
On Tue, Feb 26, 2013 at 1:35 PM, Logan Kearsley <[email protected]>wrote:
>
> I would argue that there is really no such thing as "fake"
> information; concepts may be very close together, and maybe the author
> of a particular tweet wouldn't care about the distinction between
> them, but true perfect synonyms are incredibly rare, and there's
> always *some* reason why one word or one structure was picked over
> another one.
I would think that there is _usually_ a reason that one construction is
chosen over another, or at least that linguists should consider whether
there is a reason. But the claim that such a reason _always_ exists might
be overbroad. As an empirical claim, you'd essentially have to defend
yourself every time someone comes up with two sentences with the same
meaning. Besides that, considerations other than content could pressure
the choice of one structure over the other. Brevity comes to mind, while
we're speaking of Twitter.
Messages in this topic (12)
________________________________________________________________________
1b. Re: Related to the recent discussion about counting the number of po
Posted by: "Matthew George" [email protected]
Date: Tue Feb 26, 2013 1:01 pm ((PST))
Words are like snowflakes - there's nothing preventing them from being
perfectly identical, it's just extremely implausible.
Consider all of the associations words can convey: not just a definition,
but sets of definitions. Consider puns, cognates, and similarities within
a language. Consider rhyme - whatever that means in any particular
tongue. Think of all of the complications of usage and meaning from one
culture to another. Is it possible to duplicate all of these things for a
particular word? Sure - it's just not common.
Now, in specific situations, there may be cases where one language/culture
has a word that *can* be perfectly translated into another particular one.
But that's the exception. And it doesn't necessarily hold for even the
most common words - what's the Russian equivalent of the English 'the'?
The more obscure and technical the word, I suspect, the more likely that a
perfect translation becomes. It's the heavily-used ones that will change
away from each other most rapidly.
Matt G.
Messages in this topic (12)
________________________________________________________________________
1c. Re: Related to the recent discussion about counting the number of po
Posted by: "George Corley" [email protected]
Date: Tue Feb 26, 2013 1:12 pm ((PST))
On Tue, Feb 26, 2013 at 3:01 PM, Matthew George <[email protected]> wrote:
> Words are like snowflakes - there's nothing preventing them from being
> perfectly identical, it's just extremely implausible.
>
> Consider all of the associations words can convey: not just a definition,
> but sets of definitions. Consider puns, cognates, and similarities within
> a language. Consider rhyme - whatever that means in any particular
> tongue. Think of all of the complications of usage and meaning from one
> culture to another. Is it possible to duplicate all of these things for a
> particular word? Sure - it's just not common.
>
I would argue that rhymes and homophony are less relevant. In fact, rhyme
would be another example of a pressure other than meaning acting on
selection of a given word or structure.
Also, would you consider syntactic structures to be just as snowflake-like
as words? Certainly the choice of one structure over another can convey
subtle meanings at times, but we can also decompose structures, whereas
mophologically-simple lexemes can't be decomposed into their meanings (at
least, not in any formal way).
> Now, in specific situations, there may be cases where one language/culture
> has a word that *can* be perfectly translated into another particular one.
> But that's the exception. And it doesn't necessarily hold for even the
> most common words - what's the Russian equivalent of the English 'the'?
> The more obscure and technical the word, I suspect, the more likely that a
> perfect translation becomes. It's the heavily-used ones that will change
> away from each other most rapidly.
>
Moving back to content -- does "the" really contain that much information?
It's really just pointing out something about the discourse. That's
information, but not necessarily information about the world. Can we count
it as equivalent to a content word like "dog"? I'm not so sure.
Messages in this topic (12)
________________________________________________________________________
1d. Re: Related to the recent discussion about counting the number of po
Posted by: "MorphemeAddict" [email protected]
Date: Tue Feb 26, 2013 2:30 pm ((PST))
On Tue, Feb 26, 2013 at 4:12 PM, George Corley <[email protected]> wrote:
> On Tue, Feb 26, 2013 at 3:01 PM, Matthew George <[email protected]>
> wrote:
>
> > Words are like snowflakes - there's nothing preventing them from being
> > perfectly identical, it's just extremely implausible.
> >
> > Consider all of the associations words can convey: not just a
> definition,
> > but sets of definitions. Consider puns, cognates, and similarities
> within
> > a language. Consider rhyme - whatever that means in any particular
> > tongue. Think of all of the complications of usage and meaning from one
> > culture to another. Is it possible to duplicate all of these things for
> a
> > particular word? Sure - it's just not common.
> >
>
> I would argue that rhymes and homophony are less relevant. In fact, rhyme
> would be another example of a pressure other than meaning acting on
> selection of a given word or structure.
>
> Also, would you consider syntactic structures to be just as snowflake-like
> as words? Certainly the choice of one structure over another can convey
> subtle meanings at times, but we can also decompose structures, whereas
> mophologically-simple lexemes can't be decomposed into their meanings (at
> least, not in any formal way).
>
>
> > Now, in specific situations, there may be cases where one
> language/culture
> > has a word that *can* be perfectly translated into another particular
> one.
> > But that's the exception. And it doesn't necessarily hold for even the
> > most common words - what's the Russian equivalent of the English 'the'?
> > The more obscure and technical the word, I suspect, the more likely that
> a
> > perfect translation becomes. It's the heavily-used ones that will change
> > away from each other most rapidly.
> >
>
> Moving back to content -- does "the" really contain that much information?
> It's really just pointing out something about the discourse. That's
> information, but not necessarily information about the world. Can we count
> it as equivalent to a content word like "dog"? I'm not so sure.
>
I would say, of course not. "The" can be eliminated from a text with
minimal change to the meaning.
stevo
Messages in this topic (12)
________________________________________________________________________
1e. Re: Related to the recent discussion about counting the number of po
Posted by: "Logan Kearsley" [email protected]
Date: Tue Feb 26, 2013 2:55 pm ((PST))
On 26 February 2013 13:55, George Corley <[email protected]> wrote:
> On Tue, Feb 26, 2013 at 1:35 PM, Logan Kearsley <[email protected]>wrote:
>
>> I would argue that there is really no such thing as "fake"
>> information; concepts may be very close together, and maybe the author
>> of a particular tweet wouldn't care about the distinction between
>> them, but true perfect synonyms are incredibly rare, and there's
>> always *some* reason why one word or one structure was picked over
>> another one.
>
> I would think that there is _usually_ a reason that one construction is
> chosen over another, or at least that linguists should consider whether
> there is a reason. But the claim that such a reason _always_ exists might
> be overbroad. As an empirical claim, you'd essentially have to defend
> yourself every time someone comes up with two sentences with the same
> meaning.
No, you don't. You just have to defend it if someone finds two
different sentences that they believe to have the same meaning.
Natural production of two sentences that turn out to be genuinely
identical in meaning is entirely different from intentionally
producing two different sentences that could both be interpreted with
the same pre-decided meaning.
> Besides that, considerations other than content could pressure
> the choice of one structure over the other. Brevity comes to mind, while
> we're speaking of Twitter.
That just changes what the relevant interpretations of different
sentences would be, not the fact that they are different. If one word
is usually picked in preference to another for reasons of brevity, it
is then potentially significant if the other word is nevertheless used
at some point.
There is a cutoff for significance below which a reasonable person
probably wouldn't care, and different sentences could be considered to
have identical practical meanings in any given situation, but at the
very least the choice of one form over another provides some
information about the speaker/writer's neurological state.
Then consider that the magnitude of differences in interpreted meaning
could be wildly different for different people who have different
associations for the various expressive options, so sentences whose
differences in meaning are below the threshold of significance for you
might not be for me (or for the original writer), so if we're talking
about measuring different meanings that could be communicated, it only
really makes sense to measure those with reference to a single person
at a time. From which I derive my conclusion that it's really only
useful to talk about information content, rather than more nebulous
'concepts'.
-l.
Messages in this topic (12)
________________________________________________________________________
1f. Re: Related to the recent discussion about counting the number of po
Posted by: "George Corley" [email protected]
Date: Tue Feb 26, 2013 3:25 pm ((PST))
On Tue, Feb 26, 2013 at 4:55 PM, Logan Kearsley <[email protected]>wrote:
> On 26 February 2013 13:55, George Corley <[email protected]> wrote:
> > On Tue, Feb 26, 2013 at 1:35 PM, Logan Kearsley <[email protected]
> >wrote:
> >
> >> I would argue that there is really no such thing as "fake"
> >> information; concepts may be very close together, and maybe the author
> >> of a particular tweet wouldn't care about the distinction between
> >> them, but true perfect synonyms are incredibly rare, and there's
> >> always *some* reason why one word or one structure was picked over
> >> another one.
> >
> > I would think that there is _usually_ a reason that one construction is
> > chosen over another, or at least that linguists should consider whether
> > there is a reason. But the claim that such a reason _always_ exists
> might
> > be overbroad. As an empirical claim, you'd essentially have to defend
> > yourself every time someone comes up with two sentences with the same
> > meaning.
>
> No, you don't. You just have to defend it if someone finds two
> different sentences that they believe to have the same meaning.
> Natural production of two sentences that turn out to be genuinely
> identical in meaning is entirely different from intentionally
> producing two different sentences that could both be interpreted with
> the same pre-decided meaning.
There are technical frameworks for determining whether two sentences mean
the same thing, stating that if they have the same truth conditions they
are therefore equivalent. Of course, that often ignores extra pragmatic
phenomena, which leads to a whole other argument on whether pragmatic
considerations are part of the language system or not (I think that some of
them may be, contrary to what my semantics professor is teaching me, but
not all).
Also, am I correct in guessing that you assume that "in the wild" data is
the only valid data, or at least superior to linguist-presented sentences
that native speakers judge to be grammatical/ semantically identical? I
think "in the wild" data is more important than some modern linguistics
seems to want, but there are advantages to making up sentences and asking
native speakers if they are correct.
> > Besides that, considerations other than content could pressure
> > the choice of one structure over the other. Brevity comes to mind, while
> > we're speaking of Twitter.
>
> That just changes what the relevant interpretations of different
> sentences would be, not the fact that they are different. If one word
> is usually picked in preference to another for reasons of brevity, it
> is then potentially significant if the other word is nevertheless used
> at some point.
>
For individual words it is harder to make an argument, but I could see
someone choosing a longer _construction_ if it is more redundant. And yes,
redundancy is desirable in certain situations (such as noisy environments).
> There is a cutoff for significance below which a reasonable person
> probably wouldn't care, and different sentences could be considered to
> have identical practical meanings in any given situation, but at the
> very least the choice of one form over another provides some
> information about the speaker/writer's neurological state.
>
There are certainly differences. But how is information about the writer's
neurological state relevant to the meaning of the sentence, at least
insofar as it causes a minor difference in the structure that we can't find
any other meaning for?
> Then consider that the magnitude of differences in interpreted meaning
> could be wildly different for different people who have different
> associations for the various expressive options, so sentences whose
> differences in meaning are below the threshold of significance for you
> might not be for me (or for the original writer), so if we're talking
> about measuring different meanings that could be communicated, it only
> really makes sense to measure those with reference to a single person
> at a time. From which I derive my conclusion that it's really only
> useful to talk about information content, rather than more nebulous
> 'concepts'.
I have agreed that concepts need to be left out of it, but for different
reasons. If we're not sure what, exactly, a concept is in the first place,
then we can't really build any kind of theory around it.
I would want to see how wildly people's "significance thresholds" differ
before I accept that we can only judge this on the basis of a single
individual's idiolect. It's not terribly often, as far as I know, that I
say something and am so wholly misunderstood by another native English
speaker that I must conclude that our meanings for a given sentence are
fundamentally out of sync. Artists and authors often accuse critics of
misunderstanding their work or not getting their actual meaning, but those
are people engaged in deliberately expressing meaning through indirect
means, with subjective judgments about what should have been "obvious" to
the reader.
Messages in this topic (12)
________________________________________________________________________
________________________________________________________________________
2a. Re: CHAT: the Ithkuil of programming languages?
Posted by: "Logan Kearsley" [email protected]
Date: Tue Feb 26, 2013 2:57 pm ((PST))
On 26 February 2013 13:05, Daniel Burgener <[email protected]> wrote:
>>
>> Well, I don't think we can have a conversation about this without
>> mentioning Fith as the FORTH of conlangs.
>> I had thought that Lojban might be comparable to Prolog, but now I'm
>> thinking its more of a Haskell.
>>
>
> Agree on the Haskell. My first thought was "What would Haskell be?" but
> then I read your post before actually trying to come up with something and
> you seem to have a perfect idea.
>
> How about Ido as C++? They're both reforms on one of the most widely-used
> languages.
Is Ido ridiculously more grammatically complex than Esperanto? I think
that's a requirement for a true C++ analog.
-l.
Messages in this topic (5)
------------------------------------------------------------------------
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/conlang/
<*> Your email settings:
Digest Email | Traditional
<*> To change settings online go to:
http://groups.yahoo.com/group/conlang/join
(Yahoo! ID required)
<*> To change settings via email:
[email protected]
[email protected]
<*> To unsubscribe from this group, send an email to:
[email protected]
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
------------------------------------------------------------------------
