Date: Thu, 21 May 1998 14:18:29 -0400
From: Larry Horn
Subject: Re: Law/Principle of Least Effort

At 12:06 PM -0500 5/20/98, Gerald Cohen wrote:
> Once again, I am grateful for Larry Horn's messages on the Law/Princple
>of Least Effort. Here now are some thoughts in regard to his second
>message (5/19/98).
> The overall picture I get is of an alleged law/principle with so many
>exceptions that one wonders by what justification its existence as a
>law/principle is advanced.
At the risk of further alienating listees with no interest in these
not-particularly-dialectal issues, let me push on. I'm not sure what OTHER
explanation can be offered for the wide range of cases generally attributed
to the workings of some form of the law/principle of least effort, ranging
from assimilation, elision, and coarticulation in phonetics and phonology
to acronymy, initialism, and truncation in the lexicon to a variety of
"fast speech phenomena" in parole.

>Larry writes:
> "In many cases, least effort is violated in parole while maintained
>in langue."

And vice versa, as the reference to fast speech phenomena indicates:
"Whyncha go?", "Looks like it's gonna rain", "--Jeet jet? --No, jew?" and
so on.

> If this statement is true, it seems to shake the Law/Principle
>severely. After all, individuals speak individually; and if their
>individual speech isn't subject to the Law/Principle of Least Effort, then
>how can one agree with Martinet (_Elements_, p.167), when he writes:
> "...Here [G. Cohen: i.e, in communication] as elsewhere, human
>behavior is subject to the law of least effort, according to which man
>gives of himself only so much as is necessary to attain the end he has in

Nobody to my knowledge has claimed that "individual speech isn't subject to
the Law/Principle of Least Effort". My point was that quite often
different correlates show up in langue from those that show up in parole.
If I can't remember for the moment a particular lexical item and use a
periphrastic construction instead (as a child uses "plant-man" or
"tooth-guy" before she incorporates "gardener" or "dentist" into her
lexicon), that's an asymmetry between langue and parole, just as in the
deletion of certain predictable subjects ("Looks like it'll rain", "Think
the rain'll hurt the rhubarb?") in parole where the rules of langue require
a subject here. In fact, the "null-subject parameter" dictating that
English and French MUST have a subject in indicative sentences while
Spanish, Italian and Japanese needn't (and will, in fact, dictate the
elimination of predictable subjects via least effort) is another case where
the operation of least effort will vary in accordance with other
linguistic, cognitive, and behavioral principles.

> As for French "pleonastic ne" (where "pleonastic" is synonymous with
>"extraneous"), I fail to see how this "ne" increases the semantic content
>at all.

As I tried to show in my last note, the current stage in which the (formal)
language demands ne...pas is a stage on the way from a preverbal negation
to a preverbal negation with postverbal reinforcer (necessitated, as
Jespersen argues, by the least-effort tendency to reduce the phonological
realization of the proclitic) to a postverbal negation with preverbal
"forclusif", as the French grammarians call it, to the stage in which
postverbal negation stands alone, the preverbal marker now being otiose.
Claiming that the ne...pas stage vitiates the least effort principle
strikes me as analogous to claiming that a snapshot of Michael Jordan at
the apogee of a jump (or the existence of helium balloons) vitiates the law
of gravity. Indeed, the ongoing loss of the now vestigial _ne_ in
colloquial French (parallelling its extinction in late Middle English) is
an indication that semantically and grammatically empty markers do indeed
tend to disappear.

> So, with the Law/Principle having holes like a sieve, I would ask:
>What evidence do we have that this Law/Principle really exists? Martinet,
>_ibid._, in one same paragraph refers to this feature as a "law" and a
>"tendency." Which is it?
Since it doesn't operate unchecked, I'm not sure what follows from the
distinction. Is there a law of self-preservation, or is it a tendency?
Does the fact that people sky-dive and bungee-jump, commit suicide, or go
off to war refute the existence of this law or tendency? Does the fact
that birds, airplanes, and rockets fly indicate that the law of gravity has
holes like a sieve and should be abandoned?

>Simply because a few linguists have assumed that
>a given law or principle exists does not mean that its existence should be
>accepted without question in later generations.

Among those seeking to establish functional motivation for structures,
constraints, and effects of language change (lexical, grammatical and
phonological), I'd say it's hardly (just) "a few linguists" who assume,
either ex- or implicitly, some version of the least effort principle. And
since analogous principles of economy, as Zipf stressed, operate elsewhere
in human behavior, why should we assume that language alone is exempt?
Nobody (to my knowledge) claims that the principle (or tendency) to
economize effort explains EVERYthing (in language or elsewhere), but it
doesn't follow that it explains NOthing, especially in the light of both
common sense and the difficulty of formulating a non-ad hoc alternative to
account for what it encompasses (cf. again assimilation, anaphoric
reduction, acronymy, truncation, contraction, coarticulation, etc., as well
as various other phonological, lexical, and pragmatic phenomena attributed
to least effort or to its interaction with other linguistic and social

> Incidentally, here are two items I have written relevant to syntactic
>blending and the Principle of Least Effort:
>1) "Contributions To The Study of Blending," in: _Etymology and Linguistic
>Principles, vol. l: Pursuit of Linguistic Insight_ (ed., Gerald Leonard
>Cohen; Rolla, Missouri; published by the editor), pp. 81-94, esp. pp.90-91.
>2) _Syntactic Blends in English Parole_ (= _Forum Anglicum_, vol. 15).
>Frankfurt am Main: Peter Lang. l78 pp., with bibliography, where mention is
>made of Martinet and Zipf).
>--Gerald Cohen

Thanks, Jerry; I'll look for them. (I'll send you under separate cover an
attachment of a relatively recent paper of mine on economy and redundancy,
not that we're likely to sway each other.)