Humpty Dumpty

From: clayton stirling bartholomew (
Date: Tue Mar 17 1998 - 16:35:26 EST

Mari Broman Olsen Wrote:

Words DO have meaning outside of context, else we could all be Humpty

Twice today I have been accused of propagating the views of Humpty Dumpty on
lexical semantics. After some thought, I have decided that I am not really in
agreement with Dr. Dumpty and I would like to point out why.

If my memory serves me correctly Dr. Dumpty claimed that a word could mean
whatever he wanted it to mean. This, I believe was the substance of his
remark. Dr. Dumpty was claiming that he could arbitrarily extend the range of
a word's semantic domain by using it in a new way. While there is some merit
in carefully considering this claim, I think that the claim is also fraught
with problems. One of the problems is that Dr. Dumpty is essentially denying
the existence of a semantic domain. By making the possible meanings of a
given lexical token unlimited he has essentially abolished the notion of
semantic domains.

I am not really in agreement with Dr. Dumpty when I say a word has no meaning
without a context. I use the the notion of semantic domains extensively. I use
them both for lexical tokens and for grammatical forms. I think that a
language user chooses a word for a given utterance because they have a
knowledge of it's domain and know that one particular meaning within this
domain will serve their purpose. When the language user puts this word into an
utterance the lexical token becomes instanciated with a meaning from the
semantic domain. Until the the lexical token is placed into an utterance it
remains uninstantiated. This is what I mean by no meaning without a context.

So Dr. Dumpty is being rather arbitrary in claiming he can make a word mean
whatever he wants. His claim is quite divorced form my statement that words
only take on a specific semantic value when the are place into a context.

The following is some verbiage along the same lines that I sent in a private

clayton stirling bartholomew wrote:

> My understanding is that words outside of a context are uninstanciated
> variables that draw their meaning from a pool of possible meanings (semantic
> domain) when they are placed in a context. I am not denying the existence of a
> semantic domain associated with a word. But the word's actual semantic value
> only becomes instantiated when it is placed in a context. I hold this same
> view for grammatical forms. It is the instantiated/uninstantiated state of the
> lexical token that causes some many problems to people that think that a word
> actually "means something" independent of actual use.

I actually think Dr. Dumpty's claim is worth some consideration when one is
studying the issue of language change but he has quite thrown out the baby
with the bath when it comes to lexical semantics. Dr. Dumpty, having
obviously studied under Ferdinand de Saussure, understands the arbitrary
relationship between the signifier and the signified. However his thinking,
his methods have become unsound.

Clayton Stirling Bartholomew
Three Tree Point
P.O. Box 255 Seahurst WA 98062

****************************************** "In the interior you will no doubt meet Mr. Kurtz." - Heart of Darkness, Joseph Conrad ******************************************

This archive was generated by hypermail 2.1.4 : Sat Apr 20 2002 - 15:39:14 EDT