>From: email@example.com [mailto:firstname.lastname@example.org]On
>Behalf Of Lucien Carroll
>Hi Glenn and Peter,
> If I can butt in:
Of Course, that is what this list is all about--a free for all! ;-)
>The fact that semantics are conventional or context-bound makes
>quantification difficult, but it is not inherently impossible. Even if
>(or even though) lexical semantics is prototype-based rather than
>featureset-based, we're still looking at features, merely through the
>channel of a kind of fuzzy logic. I'm curious about your example of
>"puckle". If you weren't using it as support for your argument, I would
>assume that "Puckle is a clearly defined word with no imprecision" was
Of course you wouldn't assume it had any meaning. Indeed, for all most
people in the US know, I might have made up the entire idea of Doric
english. I didn't however, but I had never heard of it until I came here.
It is a dialect of Aberdeenshire. There is another one for the Orkney
Islands. I know no words from it, but was told of it's existence while
visiting there last month.
The key feature of the word is its imprecision.
>For lexical semantics it is so hard to quantify the semantic content
>because lexemes are typically such complex creatures. But even though we
>cannot under current semantic theory assign a number value to the
>content of "ruguo wo shou de bu dui", it is apparent that the clause by
>itself has less semantic content than the full question. Subjectivizing
>the clause may give the content more "weight" or "personal significance"
>than the clause read, but it will not add to its lexical or pragmatic
>significance. Being oblivious to the conventions of mandarin, on the
>other hand, does not remove the semantic content of the string of
>characters, because outside yourself there exists a convention that
>determines its meaning; there is a natural function of the string of
>words represented. Being unable to identify the significance doesn't
>remove its significance.
There is a structure which determines its meaning, but it is an agreement
between lots of people. It is internal to them but NOT inherent in the
sounds or inherent in the letters. It is a mental state, like my thoughts
about my boss or the thoughts of my employees about me. I can't quantify
their mental states (nor do I really want to know what they think about me).
The problem, as I see it, is that in order to know something exists, you
must be able to recognize it. When I put out those sequences and have asked
people who claim that it exists, they can never recognize in which sequence
it exists. This is analogous to the claim that there is some sort of
'biological' information in molecular sequences, but they can't tell you
what it is or that it exists a priori. I think I have a perfect right to
doubt the existence of this supposedly objective 'semantic' information if
no one can even recognize it when it is put before them.
>> Hearing German means nothing to me because I don't know the language. I
>> can't even tell if someone using a gutteral language is really speaking
>> German. I can have an idea that they are, but that doesn't mean
>> are. Thus I can't OBJECTIVELY determine meaning without being in on the
>> private agreement about what sounds mean what.
>> The concept is useless, empty and misleading. It does nothing
>for us other
>> than make us feel like we are really being scientific when in fact we
>If I understand correctly, semantic information was the only kind of
>information people talked about until Shannon came along. That alone
>should tell you this is not a useless concept.
Language is an agreement that a particular sound REFERS to a particular
object. It is a grandiose table of connections which ties the sound to the
referant. The sound itself contains NO semantic information. The table, in
the mind of the individual, contains the information of which you speak.
CHinese provides some really interesting examples of this (the only
technical flaw is that I will omit the tones and they might not be the same.
But to the western ear, at first one doesn't hear the tonation anyway).
ee shou ge is a song.
ee ge shou is a hand.
In the first, shou is a measure word like a PRIDE of lions or a CONSPIRACY
of crows or a FLOCK of geese. In the second, shou means hand and ge is the
measure word in the second. In the first, ge is the word song.
Now, if Adrian corrects me about the tonation, I know that similar examples
can be found with absolutely no worry about the tonation.
There is absolutely NO information semantically in the word shou or ge. The
meaning is determined ONLY in relation to other words. To the western ear
gan bing and bing gan sound similar but they aren't at all.
>I don't, however, think that there is anything incredible about the
>accumulation of semantic information. It is interesting to me that it is
>so, but it appears to me to be the natural result of a functioning
>process that we get forms that bear the imprint of what works.
There is a correlation with biology and language that I believe holds (won't
defend it really hard right now). That is we see that any sound or sequence
of characters can be used to refer to a particular object. This is the
equivalent in biology where MANY gazillions of different sequences will
perform the very same function. The only difference is that in language, any
sound will do. In biology there are some restrictions.
for lots of creation/evolution information
personal stories of struggle
This archive was generated by hypermail 2b29 : Thu May 09 2002 - 14:56:32 EDT