[Lexicog] palm trees

Wed May 20 18:18:35 UTC 2009

Greg Mellow wrote:

"there are at least three possible ways of ordering the semantic domain
list"

My goal in compiling the DDP list of domains was your first option-to find
those groupings that would most likely be universal and average of
ethno-classification systems. I was guided in this by numerous lists of
domains that had been developed for a variety of languages. I also held two
workshops for two Bantu languages in which I asked them to organize 1000
words by domain. The words were written on cards and shuffled so they would
be in random order. The second workshop went a little further and started
collecting words in each of the domains the participants had set up. When I
analyzed the results of these workshops, I was startled to find that almost
all of their domains corresponded to my English domains. There wasn't enough
time in the workshops to do nearly enough classifying or collecting. The
speakers of the two languages were almost all educated. So I was left with
the suspicion that (1) the speakers were influenced by their education and
(2) there was probably a lot that had been missed and would have come out
had we continued.

The longer I have worked on DDP, the more I have become convinced that my
English bias was affecting the list to a greater and greater degree. This
effect is probably greater on the major categories and the
sub-sub-sub-categories. The middle categories are far more likely to be
universal. Anyway, this is why I want to do more research on a
non-Indo-European language whose speakers are unlikely to have been heavily
influenced by western education. I want to do enough detail work to learn
the basis for the domains the speakers set up. The goal is to "crawl inside
their minds" and view the world from their perspective.

As Greg notes, each of the three possible ways of ordering a list of
semantic domains has its own benefits. Ideally we would like to describe the
emic domains of the vernacular language because that is part of the lexicon.
The mental lexicon is not an alphabetical listing of lexemes. That is a
convenience for the dictionary user. The mental lexicon is a rich network
that operates on many levels. Discovering the natural emic groupings is just
one task that we could pursue. However I believe it would take a huge amount
of work to develop an emic list of domains. But that is one thing that I am
hoping to find out. Can I develop a simple, easy method for developing a
list of emic domains? Right now the only easy option we have is to use a
standardized list of domains like DDP. And DDP is the only list that I am
aware of that is anywhere close to being exhaustive and detailed enough to
classify all the words of a language. DDP is a compilation of about 25
different lists of domains plus a lot that I added myself.

So until I or someone else comes up with something better, I recommend that
you use DDP to collect words and then start researching your emic domains to
see where they are different from the DDP domains. When you find a
difference, modify the DDP list.

Ron Moe

  _____  

From: lexicographylist at yahoogroups.com
[mailto:lexicographylist at yahoogroups.com] On Behalf Of Greg and Heather
Mellow
Sent: Wednesday, May 20, 2009 12:58 AM
To: lexicographylist at yahoogroups.com
Subject: Re: [Lexicog] palm trees

Hi Wayne,

I am sorry if my reply seems inane, but one purpose of a semantic taxonomy
is to produce a list of words grouped by semantic domain, to help find
similar words. If the reader wants to explore the plant names of a
vernacular, it is useful to have the plant names grouped in a list, and to
have the list divided into various sub-groupings.

Having read the LDL today, it seems there are at least three possible ways
of ordering the semantic domain list. (1) There may be a universal order,
which would reflect a kind of average of individual ethno-taxonomic systems.
This has the advantage of being intuitive and easily used by everyone. (2)
You could use an English ethno-taxonomy, which has the benefit of being
intuitive and easily used by an English reader. (3) You could use the
ethno-taxonomy of the vernacular language, which has the advantage of being
a insightful window on the vernacular language and thought system, but the
disadvantage of possibly being obscure to an outsider. I was wondering which
of the three options the DDP taxonomy was aiming at, and whether the DDP
taxonomy needs to be tweaked for each language, or deliberately kept
unchanged.

I have just learned that in the Austronesian language I work with, coconut
palms, sago palms, pandanus trees and betelnut palms are all not considered
'trees'. There is no superordinate term to refer to them, as you rightly
suggested might be the case.

Regards, Greg

----- Original Message ----- 

From: Wayne Leman <mailto:wayne_leman at sil.org>  

To: lexicographylist@ <mailto:lexicographylist at yahoogroups.com>
yahoogroups.com 

Sent: Wednesday, May 20, 2009 2:09 AM

Subject: [Lexicog] palm trees

Greg, if you are attempting to reflect the lexicon of a language, what is
the purpose of using a semantic taxonomy that does not reflect the lexical
relationships of that language?

It seems to me that using an "external" taxonomy to aid English readers
creates a distorted view of the lexical relations within the language, which
includes taxonomic relationships.

By the way, it is not necessary to have superordinate category
classification words in order for a people to have the concept of a semantic
grouping, although it definitely helps. Not every concept that people have
is lexicalized, including in English.

Wayne
-----
Ninilchik Russian dictionary online:
http://ninilchik. <http://ninilchik.noadsfree.com> noadsfree.com

--------

Hi Ron,

I was not thinking of using a vernacular classification because the
vernacular I am studying actually seems to have few classification words of
levels that I can find. Also, I want my semantic domain list to reflect a
likely folk classification of English readers because it will mostly be
English readers who access the (English) sematic domain list. I know you do
not want an English folk classification. Sorry.

<snip>

Regards, Greg

No virus found in this incoming message.
Checked by AVG - www.avg.com
Version: 8.5.339 / Virus Database: 270.12.35/2124 - Release Date: 05/20/09
06:22:00

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lexicography/attachments/20090520/8d4a3738/attachment.htm>