6.168 Comparative Method, Speech Error

[The Linguist List]

----------------------------------------------------------------------
LINGUIST List:  Vol-6-168. Tue 07 Feb 1995. ISSN: 1068-4875. Lines: 137
 
Subject: 6.168 Comparative Method, Speech Error
 
Moderators: Anthony Rodrigues Aristar: Texas A&M U. <aristar at tam2000.tamu.edu>
            Helen Dry: Eastern Michigan U. <hdry at emunix.emich.edu>
 
Asst. Editors: Ron Reck <rreck at emunix.emich.edu>
               Ann Dizdar <dizdar at tam2000.tamu.edu>
               Ljuba Veselinova <lveselin at emunix.emich.edu>
 
-------------------------Directory-------------------------------------
 
1)
Date: Thu, 2 Feb 1995 10:01:30 -0500
From: "Paul Purdom" (pwp at cs.indiana.edu)
Subject: Re:  6.134 Comparative Method
 
2)
Date:         Wed, 01 Feb 95 23:42:23 EST
From: Larry Horn (LHORN at yalevm.ycc.yale.edu)
Subject:      Re: 6.135 Qs: Alliteration, Portuguese grammar, MLA, Barney Frank
 
-------------------------Messages--------------------------------------
1)
Date: Thu, 2 Feb 1995 10:01:30 -0500
From: "Paul Purdom" (pwp at cs.indiana.edu)
Subject: Re:  6.134 Comparative Method
 
It is interesting to compare how biologist compare DNA sequences with how
linguists compare langauges.
 
When comparing DNA sequences the first step is to align the sequences so
that after a few insertions and deletions are made the i-th position in
each sequence is appropriate to compare. In language comparison there is
no need for such a step, although if you strech a little perhaps the business
of deciding which of several words that have essentially the same meaning is
appropriate to use in comparison with other languages is somewhat the same.
On a more general level I guess you could say that in both fields a little
preprocessing of the data is done to be sure (or at least increase the
chances) that compariable items are being compared with each other. Biologists
know that there are some methodical questions concerning this first step,
but it does not lead to near the concern and debate that the corresponding step
does for linguists.
 
The second step for biologists is to consider various trees that could
possibly link the various organisms. (In princeple they consider every such
tree, but they have techniques for ruling out certain trees as being the
best solution, and sometimes they skip considering some trees that are
extremely unlikely to be the best solution even though they do not have a
proof to justify leaving out that particular tree). For each tree they
consider what DNA sequence for the intermediate organisms (those associated
with the intermediate nodes in the tree, which are now extinct). For a
particular tree that labing for intermediate organisms which requires the
least number of changes is consider the most likely labelling. (This
labelling will not unique. To take a simplified example, if one child of a
node is labelled with A and the other with C and we have no information about
the parent of the node, then the node will be labelled with the set
{A,C}. Anyway, there is a cost for any particular labelling measured by the
number of changes implied by the labelling. (In the simplified example, there
is a cost of 1; whether the node is labelled with A or C, there will be one
change between the node and its children.) There is a fast algorithm to find
the best labelling for a particular tree. The tree with the least number of
changes is taken to be the best solution. The is no quick algorithm to always
find which tree is best.
 
The thing to notice about this is that the biologists analyze their entire
data set. The final answer does not depend on pairwise comparisons. From
mathematical studies they know that consider the entire data set at once
gives an answer that is morely to reflect the true situation. There is a
method that some biologists use called neighbor joining, which does use
pairwise comparison, but most people in the field believe that its only
advantage is that it leads to a quick calculation rather than to the best
answer. (There are a few people who would disagree.) It would be extremely
useful for people who are interested in comparing languages to read about the
methods that biologists use for comparing DNA sequences. The problems are
quite similar even if it is more difficult to obtain compariable data in the
language case.
 
--------------------------------------------------------------------------
2)
Date:         Wed, 01 Feb 95 23:42:23 EST
From: Larry Horn (LHORN at yalevm.ycc.yale.edu)
Subject:      Re: 6.135 Qs: Alliteration, Portuguese grammar, MLA, Barney Frank
 
Since Marge Jackman, in her posting "Barney Frank/Fag", brings up the issue of
Dick Armey's purported speech error, perhaps it is worth cross-posting my
recent outil message here.   Responses on that list tend to view the "error"
as a Freudian slip par excellence and/or an accidental public exposure of a
familiar private (in-group) epithet, but in any case NOT a phonological tip of
the slung.  Responses from experts in that field are especially welcome.
______________________________
[outil posting of 1/30/95, slightly altered:]
 
If, as Max Weinreich used to say, a language is simply a dialect with an army
and a navy, it's worth looking at one particular feature of the Armey dialect.
I refer of course to House Majority Leader Dick Armey's recent "slip of the
tongue" in which he referred to Rep. Barney Frank (D-Mass), who our new
majority would call "an admitted homosexual", as Barney Fag.  In response to
an outraged New York Times editorial (1/29/95) castigating him for "bigotry
aforethought", Armey wrote a letter published in today's Times (1/30/95),
explaining that it was all an innocent mistake:
 
   'In saying that I did not want to "listen to Barney Frank haranguing me",
    I blended the two words (Frank and harangue, which I pronounce with a
    hard "g") in a way that made it sound as if I was using a slur.  I
    immediately corrected myself and moved on with my comments.'
Mr. Armey (R-Texas) went on to blame "the Washington media" for turning
his stumble into a national story and a lead editorial.  "No wonder", he
concluded, "public officials feel the need to speak only in scripted
sound bites".  Now while our collective heart must go out to the unfortunate
congressman, who happens to be the second most powerful figure in the House,
I was wondering if anyone on the list knows enough about speech errors
--from either a Fromkinian or Freudian perspective--to be in a position
to evaluate his explanation.  I think it's important to challenge this line,
before we're treated to a steady stream of the same, a la...
 
   Speaker of the House Newt Gingrich (R-GA) explained that he had not
   intended any personal aspersions.  'While I may have characterized the
   First Lady as a bitch and a slut, I categorically reject the politically
   motivated attempts to impugn my motives.  In seeking to call attention
   to her status as "Bill's rich wife" and to her personal and
   professional qualities as "a slender attorney", I inadvertently
   blended "Bill's" and "rich" in the former case and "slender" and
   "attorney" (which I pronounce with an initial schwa) in the latter.
   I call on the press to apologize for their manipulation of an
   honest error.  Nor is it fair to quote me as claiming that our historic
   Contract with America is "supported by every constituency in the U.S. but
   niggers and kikes".  In my natural excitement over our glorious
   program for our land, I meant to except only "nihilist muggers"
   (I favor a short [i] as the traditional value for the first vowel of
   the adjective) and "killer tykes".  The elitist liberal press will stop at
   nothing to twist my deeply felt cry of the heart into a supposed slur that
   I certainly never intended, so help me Dodd.'
 
--------------------------------------------------------------------------
LINGUIST List: Vol-6-168.