archaic forms

[Richard Rhodes]

Monica,
	I think Becky is pretty much right in the way she divides 
things up, but I think I'd say there are two parameters that 
shouldn't be confused. 1) formality -- formal/neutral/casual,  and 2) 
obsolescence -- obsolete/archaic/current. We tend to get confused 
because many archaic words end up being maintained only in formal 
contexts, but, in fact, you can find all combinations.
	The problem is, as Becky points out, that some words that 
were neutral, or even casual, migrate to being formal and you need a 
way to capture that.
	But this is only analogous to the problem of representing 
semantic shift. Words used to mean things they no longer mean. For 
example, "put something by" used to mean to reject, but now it means 
to save. So Shakespeare wrote in Julius Ceasar: "he put it by" 
(referring to the crown offered to Caesar), meaning he rejected it.
	What all this means is that the properties at issue are not 
simply properties of the word, they are part of a relation (in the 
database sense).

lexeme - time period - formality - obsolescence - meaning

Most of the time most of the information in this relation is, from 
the point of view of dictionary users, redundant. Most entries in our 
dictionaries cover all available time periods, are neutral in 
formality, and current in obsolescence. So we enter only the lexeme 
and the meaning and think we've got what we need. But when we get to 
the tough cases like the one under discussion that's were the problem 
arises. It stems from trying to make our databases look like the 
simple cases, and patch them when troubles arise.

Normalizing a data structure will reveal relations like that above. 
That's what I was trying to get at in the dictionary session when I 
said the computer folks are not doing their job. They should be 
insisting that we work out all the hardest cases to see what the 
logical relations are so they can be designed in at the outset, then 
problems like this one would be solved at the design stage and not 
patched 10,000 entries later.

Rich

At 12:01 PM -0600 11/3/05, Monica Macaulay wrote:
>Posoh fellow Algonquian dictionary compilers...
>
>As promised, I'm going to throw out a topic for discussion. This has 
>arisen in the context of our Menominee dictionary. We of course have 
>used Bloomfield's Lexicon as a base, and sometimes he marks things 
>as "archaic." We put a checkbox into our database and checked it 
>when he made that notation. However, it dawned on me that there was 
>another, similar situation, which was when we asked our speakers 
>about a word from the Lexicon, and they didn't know it. We were 
>putting that in our notes field as "not familiar." So I started 
>wondering if we should check "archaic" for those, or have a new 
>checkbox, or what. It seemed important to me to have a record of 
>what Bloomfield found to be archaic in the 1920s and to keep that 
>distinct from what the speakers today are not familiar with. I think 
>there's a difference between words that were archaic in the 1920s 
>when there was still a viable community of speakers, and words which 
>are unknown today, when there are very few speakers and the language 
>is in a severe state of attrition. Furthermore, there is definite 
>register compression (to use Ives' term), with the elders constantly 
>saying "oh, that's that High Menominee, which we didn't learn." We 
>started a project-internal discussion of it and Becky Shields wrote 
>this long message laying out a huge list of possibilities (which 
>I'll paste in below), and now we're stumped about how much to 
>include. So I thought I'd ask the list, and see if anyone else has 
>considered this issue. Here's Becky's message:
>
>Thanks for the discussion - I think I see the complexity of the 
>issue more clearly now. And I now agree that it would be useful to 
>distinguish between various types of "unknown" words.
>
>Here's the way I see it now (sorry if I'm repeating a lot of what 
>you just said - just trying to systematize it in a way that makes 
>sense to me):
>
>There are at least two different issues:
>
>1) there are colloquial and formal registers, and the formal 
>register may contain archaic words which (presumably) used to be in 
>the colloquial register in the past, but now survive only in 
>prayers, storytelling, etc. Speakers familiar with the formal 
>register know these words, but do not use them in everyday 
>conversation.
>2) there are words that have totally fallen out of use (in all 
>registers). These are "unknown" words.
>
>In addition, the data we are analyzing come from two distinct time periods:
>A) 1920's
>B) contemporary (1980's-present, if we include data gathered by Tim 
>Guile, etc.?)
>
>In principle, there could be distinct sets of both archaic and 
>unknown words at each time period. So there are four possibilities:
>1A) archaic words in the 1920s - words LB found in use only in the 
>formal register, presumably the ones he calls "archaic" in the 
>lexicon
>2A) unknown words in the 1920s - words from earlier sources (like 
>Hoffman?) that LB's speakers did not know. If LB didn't do this type 
>of elicitation, we may not be aware of any of these.
>1B) contemporary archaic words - words that speakers currently know, 
>but use only in formal registers like prayers. Lavina and Marie may 
>not know this register much at all, but perhaps some male speakers 
>(Joe Beaver?) do, or perhaps there are some prayers or stories on 
>the pre-2000 tapes. I have definitely heard Marie and Sarah say 
>about certain words that they recognized them only from some prayer, 
>so this set certainly has some members.
>2B) contemporary unknown words - words not in use by contemporary 
>speakers. As you point out, we have a very limited set of 
>contemporary speakers, even if tapes from the 80s and 90s are 
>included. This limits our data set, and in addition contemporary 
>speakers probably have a smaller vocabulary than speakers in the 
>20s, due to the moribund status of the language. This is just a fact 
>of life though, and I don't know what we can do about it, other than 
>be very very sad! We are certainly obligated to report what we 
>actually observe, and not what we wish we had observed. And 
>definitely not try to pretend we observed what LB observed. 
>Accurately documenting language attrition could also be very useful.
>
>The sets obviously might overlap - so words in 1A may now be in 1B, 
>2B, or even have become known (although this seems unlikely). So 
>actually for any given word there are multiple possibilities. These 
>seem the most likely combinations to me:
>- 1A-1B (used to be archaic - still archaic)
>- 1A-2B (used to be archaic - now unknown)
>- 2A-2B (used to be unknown - still unknown)
>- 1B (used to be colloquial - now archaic)
>- 2B (used to be colloquial - now unknown)
>- (and of course the unmarked case - used to be colloquial - still colloquial)
>
>Maybe that's more detail than anybody needs or wants, I don't know. 
>J But as you guys pointed out, this info could be useful to 
>historical/comparative linguists, and community members interested 
>in bringing back older words, and also people studying language 
>death.
>
>-becky
>
>So we would be very interested to hear what y'all have to say on the matter.
>
>
>Thanks!
>
>
>Monica Macaulay
>
>Department of Linguistics
>
>1168 Van Hise Hall
>
>1220 Linden Drive
>
>Madison, WI 53706
>
>phone (608) 262-2292
>
>fax (608) 265-3193
>
><http://ling.wisc.edu/~macaulay/monica.html>http://ling.wisc.edu/~macaulay/monica.html


-- 
******************************************************************

  Richard A. Rhodes
  Department of Linguistics
  University of California
  Berkeley, CA 94720-2650
  Voice (510) 643-7325
  FAX (510) 643-5688

******************************************************************
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/algonqdict/attachments/20051103/415058ea/attachment.htm>