[Lexicog] Re: lexical relation for boar - sow?

Mon Feb 27 22:05:26 UTC 2012

I've been working with lexical relations and semantic domains for over ten
years, analyzing the relationship between words, classifying them into
domains and lexical sets, and puzzling over odd words. The concept of
lexical relations is a very useful one, but I've found that it is also one
that can be misleading.

Lexical relations are used to indicate the relationship of words in a
dictionary. We use them in cross-references. It is nice to indicate the
relationship more precisely than just a plain cross-reference:

boar n. male pig. complementary antonym: sow

(or)

boar n. male pig. gender pair: sow

(or)

boar n. male pig. female: sow

(or)

boar n. male pig. cf. sow

(or)

boar n. male pig. see: sow 'female pig'

The question for me is what is most helpful for the user. For a "popular"
dictionary I would prefer the last example above as a way of indicating
lexical relation cross-references. As linguists we may like our technical
terminology. But I don't think they are always so helpful to our users.
FieldWorks enables us to do it all. FieldWorks enables you to indicate the
technical term for the lexical relation in your database and export it for
publication. FieldWorks also enables you to indicate the technical term in
the database, but export the dictionary with a generic "cf." or "see" with
or without a gloss for the cross-referenced lexeme. You can produce any of
the formats above or even:

boar n. male pig. complementary antonym: sow 'female pig'

I might choose such a format in a technical bilingual dictionary.

Lexical relations are also used to analyze derivational affixes. The
root:derivative relationship is often, but not always, a lexical relation.
Being able to label the lexical relation helps in describing the morphology
and glossing/defining derivational affixes.

Lexical relations are also used to help in analyzing semantic patterns. For
instance Robert Hedinger and Richard Gravina mention "complementary
antonym". Richard mentioned the test "If not X then Y." If "complementary
antonym" is a viable lexical relation, then there should be patterns of
syntactic and semantic behavior that are shared by all members of the
lexical relation. Similarly "gradable antonyms", such as big:small, should
fit in frames such as "very (big)" and "neither big nor small, but in
between" (compare: *very male; *neither male nor female, but in between).
Examples of the "converse" relation fit the frame "John sold the bicycle to
David," and (with the same referential meaning) "David bought the bicycle
from John." For a lexical relation to be valid there should be a set of
frames and tests that can be applied to all examples of the lexical
relation. So if you are going to develop a list of lexical relations for
your language, you should develop your own list of tests and frames for each
lexical relation that is valid for the language. You can't just borrow a
standard list, although a standard list is a nice place to start.

One of the problems with lexical relations is that many pairs of words are
obviously related, but the relationship is not one of the standard lexical
relations (e.g. rain:umbrella, rain:rainbow). So our standard lists of
lexical relations are insufficient to describe all the relationships between
lexemes in the mental lexicon. That doesn't mean the concept is invalid or
unhelpful. It's merely inadequate.

Ron Moe

  _____  

From: lexicographylist at yahoogroups.com
[mailto:lexicographylist at yahoogroups.com] On Behalf Of lengosi
Sent: Monday, February 27, 2012 1:18 PM
To: lexicographylist at yahoogroups.com
Subject: [Lexicog] Re: lexical relation for boar - sow?

Thanks for the explanation of different kinds of antonyms, Richard. I found
that helpful. I guess it works to a certain extent for the pair boar and
sow. But on the other hand: if not sow then {boar | barrow (neutered male) |
gilt (female that has not given birth)}... There seem to be too many options
for them to be complementary antonyms (which you described as a 'pair'
relationship). As I think of it 'gender pair' also seems to be limited in
this regard. Ron, you're right--this is a mess! Is 'counterpart' too loose
to describe their relationship? It seems to have the advantage of not being
limited to a pair.

Paul

--- In lexicographylist at yahoogroups.com
<mailto:lexicographylist%40yahoogroups.com> , "Richard Gravina"
<Richard-Sue_Gravina at ...> wrote:
>
> Complementary antonyms express oppositions where the denial of one member
of the pair implies the assertion of the other member. If not X then Y. Some
examples in English are: alive and dead, male and female, open and shut,
relinquish and retain. If a person is not dead they are alive; if a person
is not male they are female, and so on. Other types of antonyms are Gradable
antonyms (big and small) and Converse (buy and sell).
> 
> 
> Boar and sow fit well as Complementary antonyms, as Robert says.
> 
> Richard
> 
> From: Robert Hedinger 
> Sent: Saturday, February 25, 2012 3:58 PM
> To: lexicographylist at yahoogroups.com
<mailto:lexicographylist%40yahoogroups.com>  
> Subject: Re: [Lexicog] lexical relation for boar - sow?
> 
> 
> 
> Some linguists call these Complementary antonyms.
> 
> Robert
> 
> From: lengosi 
> Sent: Saturday, 25 February, 2012 5:29
> To: lexicographylist at yahoogroups.com
<mailto:lexicographylist%40yahoogroups.com>  
> Subject: [Lexicog] lexical relation for boar - sow?
> 
> 
> Is there a widely accepted lexical relation to describe male-female pairs
of animals (e.g., boar and sow)? They seem to be (co)hyponyms of some sort;
probably not antonyms. Specifics of a Generic (pig)?
> 
> Well, I'll make an end of showing my ignorance. ;-) Thanks for any
guidance,
> 
> Paul
>

  _____  

No virus found in this message.
Checked by AVG - www.avg.com
Version: 2012.0.1913 / Virus Database: 2114/4835 - Release Date: 02/27/12

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lexicography/attachments/20120227/9be697e7/attachment.htm>