Trask terms etc

[Harold Koch]

----------------------------Original message----------------------------
1. Borrowing terminology.
The discipline is badly in need of some standardised terms, other than
"cognate", for (a) words related by borrowing, and (b) words related by
either cognation or borrowing, that is, a generic term that encompasses
"cognate" and whatever (a) should be called. I hope Larry makes some
recommendations in this area.
 
In comparing the vocabulary of two (or more) languages that are both
genetically related and liable to have borrowed from one another (this is
a common situation in Australia), one tries to gather together all the
words that are "related" in some sense-- possibly called "tentative (b)s",
then sort out the "cognates" from the "loans"-- i.e. (a)s--, then work out
the direction of the "loans". In the last step there is perhaps a further
problem with the term "loanword"; can it be used symmetrically?
 
To illustrate the problem-- If I decide that Warlpiri wapirti and Kaytetye
apeyte (both 'pencil yam') are related by borrowing, can I refer to both of
them as "loanwords" before I have decided on the direction of borrowing? If
I decide that the direction was from Kaytetye to Warlpiri, I can say that
Warlpiri wapirti is a loanword from Kaytetye, but can I say that Kaytetye
apeyte is a loanword into Warlpiri, or is there another available term? I
am aware of Crowley's proposal to use "copy" instead of "loanword", but can
this term be used symmetrically?
 
You can see that I think the working comparativist needs terms for each of
the following: (1) a word or pair of words related either by
cognation/inheritance or borrowin- a cover term (2) a word or pair of words
related by cognation /inheritance-- here "cognate" is established, (3) a
word or pair of words related by borrowing, without specification of the
direction of borrowing, (4) a word related to another by the fact that the
former is a copy of the latter-- this seems to be the focal meaning of
"loanword" (5) a word related to another by the fact that former is the
original from which the latter is a copy.
 
2. Borrowing symbols
It would be useful to try to institutionalise some symbols for specifying
borrowing in etymological entries. I have suggested and been using the
following (in contrast to > and < for inheritance):
 
A <- B: A is borrowed from /is a copy of B
A -> B: A is borrowed to become B / is copied as B
A <-> B: A and B are in a borrowing relationship of indeterminate direction.
 
Cf. H. Koch 1983, Etymology and dictionary-making for Australian languages
(with examples from Kaytej). In Peter Austin (ed.), Papers in Australian
Linguisitcs No. 15: Australian Aboriginal lexicography (Pacific Linguistics
A-66) Canberra: Dept of Linguistics, Research School of Pacific Studies,
Australian National University; 149-173.
 
3. Symbolisation of proportional analogy
Should the various formalisms that have been used to represent proportional
analogy be given in the symbol guide, or can they all be assumed to be
understood from some other discipline?
A quick check in HL textbooks yields at least these three:
1) sow : sows = cow : x (Bloomfield)
2) stone:stones::hand:X (Arlotto)
3) stikker/stak = nikker/X (Anttila)
 
4. Ghost words
I believe the term Blaine Erickson (7 June) is seeking is traditionally
called a "ghost word" (or "mot phantome" in French).
 
Harold Koch, Senior Lecturer
Department of Linguistics
Faculty of Arts
The Australian National University
Canberra ACT 0200 Australia
 
Telephone:  (02) 6249 3203 (direct) / ..3026 (messages)
(overseas) 61 2 6249 3203
Fax:  (02) 6 279 8214
(overseas) 61 2 6279 8214
email:  Harold.Koch at anu.edu.au