[b-greek] RE: Lexical Fallacies

From: Iver Larsen (iver_larsen@sil.org)
Date: Thu Feb 07 2002 - 02:17:00 EST

> Lexical Fallacies
> There is a rather common fallacy employed in lexical semantics which has
> surfaced once again on this list in the last week. I don't know a name for
> the fallacy but it is stated something like "we find no attested
> examples of
> that word used with that sense . . . ." The hidden flaw in this
> argument is
> that you cannot find new meanings of a known word as long as you are using
> this logic. The knew meanings are impossible since the are not "attested."
> So the new meaning which is staring you directly in the face is
> ruled out as
> impossible, even though the context in which the word is found makes this
> new meaning highly probable. The employment of this "we find no attested
> examples . . . " leads lexical analysis to an abrupt dead end
> when the first
> "attested example" is the object of the dispute. In other words this logic
> is a trap.
> I have been using LEH* extensively while working Codex Panopolitanus
> (1Enoch) and have found a number of rare words in this codex which show up
> only once in the LXX and are marked as Neol. (Neologism). These are often
> compound words made up of elements that can easily be identified
> but you end
> up being forced to use a certain amount of etymology in combination with a
> study of the context to determine the probable meaning of these
> words. There
> were not attested examples of these words or their meanings before they
> appeared in the LXX. Most scholars have no problem accepting the
> existence
> of "new meaning" when it comes to a neologism but they have a BIG problem
> accepting "new meaning" when it comes to known words.
> I have also run across more than one word which is found in the LXX and
> elsewhere but in no other known document is it "attested" in the
> sense which
> is found in C. Panopolitanus. Both new words (Neologisms) and
> new meanings
> of known words are regularly encountered in the process of
> exploring ancient
> documents. I did some work in Agamemnon (Aesch.) a few years ago and came
> across a number of known words used with unknown meanings.
> If the lexical analyst isn't prepared to use the context as the
> determining
> factor in word meaning, he will never discover new meanings to
> known words.
> People who are suffering from this syndrome need to browse in Louw & Nida
> and mark all the semantically isolated low level domains, i.e., words
> attested with a given sense only once. This little exercise should help to
> release one from the death grip of the "no attested examples" fallacy.
> Clay

It seems that you are saying: "For neologisms we have very little basis for
knowing the range of meaning of the word in question but have to rely
heavily on guesses from context as well as other means, such as related
words from the same root. Therefore, the same applies to words that are well
attested and found in many contexts." I doubt that you would agree that this
is what you are saying, since in my summary the fallacy of the argument
should be easy to see. But this is how I interpret what you are trying to

Ludwig Wittgenstein introduced a new aspect to lexical semantics in the last
century when he said "A word is defined by its use".
The normal procedure in lexical semantics is that you study all the
occurrences you can find of the word in as large a corpus as possible. Then
you classify them on the basis of context into groups that become senses.
These senses may be very different, and in translation into another
language, different words will probably have to be used. In other cases, the
senses are close, just a small difference in nuance. For these cases, you
may or may not be able to use the same translation into another language.

If a word is common, the likelihood of discovering a sense which is outside
the so far attested or described senses, is much smaller than if a word is
found only two or three times in a big corpus. Therefore, when we deal with
a common word, the discovery of new senses is not impossible, but the
context needs to be such that all the already attested senses cannot apply.
This new sense must be forced on the word by the context, including the
non-linguistic context. Irony, for instance, will often use a shared
knowledge outside the text itself to help communicate a different meaning
than the normal sense. In most cases, though, there are clues in the co-text
that suggest ironic intentions.

It is also very important to understand that a word come to its context
"with an open mind". By this I mean that a word usually does not have one
clearly defined specific sense, but has a range of possible senses. The
range may be narrow or very wide. Which sense is relevant in a particular
context depends on the words it co-occurs with. Some words are very broad,
like the Hebrew word for "lie down/sleep" and its counterparts in Greek and
English. I use slashes to indicate that two words in English are needed to
approximate the range of senses of the underlying Hebrew and corresponding
Greek words. The "open-minded" word gets a specific sense by the words it
co-occurs with, so that the result is a narrowing of a wide range of
potential senses to a much more specific sense which is relevant in this
particular context. That is why "lie/sleep (in bed)" has a different sense
from "lie/sleep with (a woman)" and different again from "lie/sleep (in the
grave)". If we are unaware of this aspect of lexical semantics and
communication, we are likely to be misled in our exegesis, as has been
pointed out recently on this list.

In order to avoid lexical fallacies and questionable exegesis, it is also
important to understand how human communication works. If a word is
well-known, the reader will expect that when it is used, it will be used in
one of the senses that fall within its normal range. The reader will use the
context to pick which of these potential senses is relevant for the context
at hand. If the intended sense is outside the established normal range of
sense, it is the context alone that has to force that new sense to be
understood in this particular context. That is why a new sense for a word
with a well established range of sense can only be postulated if the context
very strongly supports, or rather requires, such a new sense.

I hope this helps to clarify the issue for those who or not very much
familiar with lexical semantics.

Iver Larsen

B-Greek home page: http://metalab.unc.edu/bgreek
You are currently subscribed to b-greek as: [jwrobie@mindspring.com]
To unsubscribe, forward this message to leave-b-greek-327Q@franklin.oit.unc.edu
To subscribe, send a message to subscribe-b-greek@franklin.oit.unc.edu

This archive was generated by hypermail 2.1.4 : Sat Apr 20 2002 - 15:37:18 EDT