Darwin-L Message Log 7:31 (March 1994)

Academic Discussion on the History and Theory of the Historical Sciences

This is one message from the Archives of Darwin-L (1993–1997), a professional discussion group on the history and theory of the historical sciences.

Note: Additional publications on evolution and the historical sciences by the Darwin-L list owner are available on SSRN.

---

<7:31>From sally@pogo.isp.pitt.edu  Sat Mar 12 07:07:22 1994

To: darwin-l@ukanaix.cc.ukans.edu
Subject: Re: Structural and historical linguistics
Date: Sat, 12 Mar 94 08:07:21 -0500
From: Sally Thomason <sally@pogo.isp.pitt.edu>

   Bob O'Hara asks what it is about Saussure's historical inference
about the vanished "laryngeals" that made it a structural inference.
Of course Saussure's proposal wasn't made in an intellectual vaccum:
the regularity hypothesis for sound change (the idea that if sound x
changes to y in context A in one word, x will change to y in context
A in all other words) is also, in an important sense, structural.
What was different about Saussure's proposal was that he (a) figured
out the structure of the set of alternations, and then (b) realized
that the structure would be a lot neater, more regular, if there had
been a set of consonants in the proto-lg. that were unattested in
*any* of the then-known daughter languages.  That is: he used the
structure of the system to construct a hypothesis about previously
existing sounds, rather than following the usual procedure of his
day in taking only the attested sounds and reconstructing a best-guess
proto-lg. sound from that set.

   The system in question had to do with vowel alternations in roots
in various morphological contexts (i.e. word structures).  English
verb pairs like sit/set, drink/drench, and fall/fell [as in "to fell
a tree"] are descended from one such alternation; the first member
of each pair -- at least those pairs that have an ancient etymology,
like sit/set -- goes back to a PIE word with a vowel *e; the second
member had *o in PIE.  Other alternations resemble those found in
English sing/sang/sung.  And so forth.  The patterns could be
established by comparing the attested languages (not surprisingly,
if even modern English retains traces of some of the alternations!),
but there were a LOT of exceptions, and for these there was no
adequate explanation at all before Saussure published his monograph.
Instead of taking the patterns AND exceptions as a given, he took
the patterns, the overall structure, as a given, and then considered
how the exceptions might have arisen from such a structure.

   What Saussure was doing was what is now called Internal
Reconstruction: he started with the parent language, Proto-Indo-
European (that's what PIE stands for -- sorry, should've said
that earlier), in its form (as believed in Saussure's time) shortly
before it diverged into the various daughter languages, and drew
conclusions about its earlier structure, basing his conclusions
entirely on the then-current reconstructions of PIE structure.
(One could then talk about "early PIE" -- before the "laryngeal"
consonants disappeared -- and "late PIE", shortly before the
break-up into daughter languages.)  (The reason for the assumption
that the "laryngeals" were lost before late PIE -- again, based
on what was known in Saussure's time, minus Hittite -- was that
the consonants disappeared with exactly the same effects in all
the daughter languages, as far as one could tell then; so the
simplest hypothesis was that the consonants were lost before the
break-up into separate languages.)

   And finally, a coda to Bob's question about whether he could
have done the same thing without a "structuralist component"  --
or, at least, without seriously structural(ist) thinking: I doubt
it.  Without the concept of the overall structure of the system,
and a belief that the system must have had a truly regular structure
(at least at an earlier time), it is unlikely, I think, that
Saussure would have thought to posit something so weird (in his day)
as a whole set of unattested consonants.  It's true, of course, that
this was before he began the work that was to earn him the title
of Father of Structural Linguistics.  But I think his future work
is clearly foreshadowed in his 1879 Me'moire sur le syste`me primitif
des voyelles dans les langue indo-europe'ennes.

  Sally Thomason
  sally@pogo.isp.pitt.edu

---

Your Amazon purchases help support this website. Thank you!