<Language> Lass: Part 3 of 3: Electronic Review

[H. Mark Hubey]

<><><><><><><><><><><><>--This is the Language List--<><><><><><><><><><><><><>

Lass, Historical Linguistics, 1998

The comments inside braces {...} are probably all mine and were
meant for memory jogging and markers.

-------------------Part 3 of 3 ---------------------------
P. 298  (cycles)
Long-term cycles of change are familiar from the histories of many (I
would suspect most) languages.  In these cases particular items or
categorical structures appear, disappear (through loss or
transformation) then we appear, and such alterations can go on for
indefinite periods (this is beautifully laid out in von der Gabelentz
1891). These cycles are interesting because there's clearly no
particular necessity that either the states represented should appear at
all.  If for instance a line which develops a particular contrast, then
loses it, then develops of again, it is obvious that neither the state
with it nore the one without it is preferable, since the languages seem
to get along perfectly well in both.

{phonological change i.e. Brownian motion--> Gaussian, Einstein--> most
don't change..}

P. 299

Cycles of course may pose was severe problems for reconstruction since
one has to know where in the cycle what is to know what direction was
moving in. what they are also interesting in themselves: why should such
(quasi)-periodic attractors exist at all? 1 possible answer is that so
many linguists info, are in essence simply the manifestations of one of
a pair of binary choices, so that in effect if a change is to occur at
all in a particular domain, it will virtually always have to involve
exiting one state and entering the only other one there is. Many
linguistic states therefore can only be construed as points on the limit
cycle, and the cyclicity will always be capable of  emerging to the
historians field of view.  It least in principle; whether in fact
depends on the period of the attractor.  Some may have such long periods
at one cannot tell them from point attractors, or even stable conditions
that existed forever: a language state may be so long-lasting that it
looks like stasis, and may or may not be. It seems that we have no way
of telling.  There are however certain good candidates for limit cycles,
as there are for point attractors.  The significant point is that
identifying regimes which are likely to the cyclical demystifies
cyclicity: there is no need for anything occcult or mysterious, given a
system which is unstable to begin with.

{examples of cycling of vowels, short long etc.}

P. 300

One particularly interesting kind attractor, somewhere between a cycle
in point, his or have called the center of gravity phenomenon. A center
of gravity is a kind of limit cycle of relatively long period, in which
the system, however much may wander about in the interim, continually
returns to particular values for some parameter, or favors recurrence of
particular output types or processes. { more examples}

P. 301

That is, language is often display, either for the whole of their
histories or for specific periods, apparent preferences for certain
kinds of change, or certain developmental patterns... in a variety, in
which groups of changes appear to move in a certain direction, or
towards a goal, has been familiar and problematical at least since the
1920s, under the rubric of drift, laterly also conspiracy (Sapir, etc).
This is the kind of change that can be construed as autocatalysis.

In a drift sequence of classical kind, what appears to happen from a
macro perspective is that (some subsystem of) a language tends to move
in a particular direction, so that the end point of an evolutionary
sequence is a fairly major state change (e.g. a language starts out
phonemic and ends up without it, starts out OV and ends up VO).  To this
has been interpreted (mistakenly) as a kind of orthoGenesis or directed
evolution (notoriously by Lass, which I regret...) {examples given}

P. 302

Any possible language state (at micro or macro levels of resolution) can
become any other;....  Much changes ran them or semi random tinkering,
or play with options available in the category space (6.4). Both thereof
within histories possible some directions on larger scales, which appear
on the finer scale as individual grammaticalizations (e.g. noun
>clitic>...) or on a larger scale as conspiracy or drift.

>From most important point that ought to emerge from this discussion is
that the landscape or dynamical system image makes any form of teleology
superfluous.!!!!!  It is the combination of initial conditions and the
interaction with abstract control parameter is that determines the shape
all systems wondering us through its phase space.  This has been very
well put by Cohen & Stewart (1994:31).

	A dynamic does not necessarily imply a purpose.  Darwinian 		evolution
as a dynamic, but organisms do not seek to evolve.
	The existence of attractors does not implied that dynamical
	systems are goal seekers: on the contrary, they are
	goal-finders, which only recognize what the
	goal is when they have found it.


P. 303

Language histories (especially those primarily embodied in written
corpora, which are the easiest to look at) often appeared to show long
period of stasis, followed by bursts of innovatory activity... both the
stasis and a burst of activity may be real...

A history off a language-states is of course only as far as our sampling
intervals: the bigger they are, the > degree of manifest change; and or
conversely, the more likely the apparence of stasis.

P. 304
what we have been such cases is pseudo-stasis due to post settlement
homogenization.  On the other hand, there is genuine stasis, even in
non-literary languages, anyone may be called variational stasis.

P. 305
Otto Jesperson once remarked (1992:418) that "transformation is
something we can understand, while a creation out of nothing can never
comprehend up by the human understanding"...

In principle, there was seem to two basic types of possible innovation:
transformation of inherited material, and innovation from zero or
invention or relative vs. radical or absolute innovation....  You've
seen that languages (ceteris paribus, as always) innovate preferential
he by utilizing (including transforming) existing material; as a second
recourse they borrow; as a third they just might (in certain domains
anyhow) invent in. there seems to be a preference hierarchy (which also
serve as a guy to likelihood in reconstruction, as directional pathways
do): phonological change > morphosyntactic change by analogy or
reanalysis > morphosyntactic borrowing > absolute invention.  Relative
or transformative innovation is generally amenable to reconstruction (if
not always of the strict kind).  Absolute innovation involves even more
than the emergence: it involves the forging ex vacuo of new material.
Thus it should in principle be possible to find cases, even in the
histories of well-known languages, where there is no ancestors but zero.
The argument for this type of innovation is not empirical, but
transcendental: but it is one worth following a, as it has not, as far
as I know, been pushed to its limits.

It's why should innovation from zero the possible, or why should we even
think of it?  There is a historical argument, which, interestingly, is
the same whether one believes in monogenesis (like the flakier adherents
of Protoworld) or polygenesis (perhaps the safer option) for human
language. Is clear that all the characters of proto-world or where you
care to call it, or on a poly genetic view, the characters all the
original proto languages for each (established) family, must at some
point represents absolute innovation.

That is, while languages as (say) the type of software or capacity may
be mono genetic, individual languages and therefore families are the
results of polygenesis, the software operating on and/or creating
material to fulfill the capacity.  It doesn't matter in this sends
whether it happened once or any indefinite number of times. This is
probably, and typically speaking, the non-issue, since the time-depths
involved are so vast that we have no technologies for exploring them.
There is no way to find out when the noun phrase was invented....

English is separated from its putative ultimate parent
Proto-Indo-European by any reasonable estimates some 8000 years; yet
there is not one piece of inflexion or generational material that
English has now or ever had that cannot be traced back to the parent
language.

p. 314--biology, rna, imperfect copying etc

P. 316
Exaptation... This "missing term in the science of form" as they called
it (Gould and Vrba) denotes the co-optation doing evolution of
structures originally developed for other purposes. Classical examples
are the co-op station of feathers....  Exaptation, that is, is
opportunistic: it is a kind of conceptual renovation, as it were, of
material that is already there, but either serving some other purpose,
or serving no purpose that all.  Thus perfectly good structures can be
exapted, as can junk of various kinds.

I am convinced (see further 7.6-7) that there is such a thing as a
theory of historically evolve systems, and other virtually any subsystem
to meet certain criteria is going show, that look like junk deposition.
In other words, human cultural evolution, (or of the evolution of human
cultural artifacts, which is almost virtually not the same thing), like
the evolution of biological systems, is based at least partly on
bricolage, cobbling, jerry-building, what everyone to call it; pieces
also systems are always falling off and if not lost out of recycled,
often in amazingly or general and clever ways...

P. 317
if a piece of juncture arise, for any reason, there are three things one
can do with it: pleaded as it is; get rid of it; or recycle it been used
for something else.  Languages like organisms display all three
strategies.


P. 319
It is important to distinguish exaptation from analogical and similar
processes, or abduction.

P. 325
people have tried to explain why it linguistic change should occur ever
since they first became aware of it.  There are two major some issues:
(a) why should any change that all occur? and (b) why should some
particular observed (type of) change have occurred?  Question at present
has no agreed on technical answer ...

P. 326
the usual focus in the literature has however been not on wide change
cars off unexplained particular changes or change types under larger
rubrics with some supposed theoretical warrant.  There seem to the two
polar positions on explicability. ... Explanation is a complex motion,
with many possible construals...  Taking into account in the source of
objects that had been called explanations in various sciences and
pseudo-or quasi sciences  there seem to the three basic or canonical
conceptual types:

1.  Causal explanations.

2.  Functional explanations.

3.  Genetic or historical explanations.

... we might add a fourth category, which could perhaps be seen as
subcase of two; rational explanations.

This category, under one guise or another, could be considered the type
for hermeneutic explications of various kinds: any attempt at the
explanation or exegesis extrapolating from, `the common experience of
being human' could eventually fall under this heading, since what
underwrites them all abduction based on something like empathy or
understanding, which in the end amount to a kind of a anamnesis.


P. 329
if a strong D-N type seems to the usual of the language change;
histories of contingent domain, there are no laws of the requisite
power, if there any kind that all.  Probabilistic explanations can be
applied in case of tendencies or likelihoods but they are peculiarly
unsatisfactory, in that they typically reduce at best to tautologies:
whatever is more likely to happen is more likely to happen, and our
surprise at the occurrence is to that extent diminished....  If
languages as historical objects are indeed systems with sensitive
dependence on initial conditions than micro-level prediction is
impossible, since the initial conditions are in principle either
unknowable or unspecified with the requisite degree of precision; though
after the fact explications of a quite delicate kind may be feasible....

We are then left with a seventh week explanations of different times,
involving such notions as function, reason, etc.....

A further variant is the invisible hand explanation in which people do
rational things but in the course of doing them achieve some unintended
goal.. The discussion above may seem rather limiting; probably reason is
that I'm trying (as I did somewhat all successfully in Lass 1980)
through distinguish between explanation is a technical term in academic
discourse, and some of its looser everyday uses....

p.311

Some writers, while still making a distinction between weak and strong
explanations, attempt to group all exegetical activities under the same
heading...(see Heine)..; he is willing to use the term explanation
whenever any of the following goals have been achieved:

1.  Description of a phenomenon is an example of a more general
phenomenon.

2.  Seeing facts in a wide context or in the larger pattern.

3.  Imparting organized knowledge, i.e. knowledge of relation between
various facts..

{see Futuyma 1995.. discussions of testability of evolutionary
hypothesis}

p. 332
it's now indicate that a half since I launched a full frontal attack and
what seemed to be the false tensions of most times explaining linguistic
change (Lass 1980).  That both in retrospect was a bigger could, course
depositors assault on all forms of explanation other than those fitting
D-N or covering law model...

P. 333
in fact the explanatory commencement of restored or sciences (if one can
called them that) is a special and rather different sort from that in
the natural sciences. As du Preez says of psychoanalysis, taking is a
paradigm case to the hermeneutic discipline:

	the explanation of any actual event is a long narrative
	about 	circumstances...
	it does not fit into the hypethetico-deductive format. The
	interface, it is  like history, which also has generalizations,
	is explanatory, and attempts  to discover the origins of events
	in circumstances, but cannot be predictive.

On the other hand (and this is relevant to the discussion that follows),
in psychoanalysis narratives may be falsifiable (in some details) as
theories in the natural sciences. But because of the complexity of the
narrative... because although many options on tone and incident one can
rarely falsify a narrative as a whole.  The notion of what might
constitute a falsification is crucial, as we will see below.


P. 334
To use Peirce's famous beanbag example, deduction induction and
abduction all involve the Triad of rule, case, and result, but inference
moves in different directions....

Abduction on the other hand involves inferring the case from rule and
result:

	Rule: all beans from this bag to our white.
	Result: these beans are white.
	Case: these beans are from this bag.

P. 335
there is a fundamental problem here: other peoples all abductions
(unlike their correct deductions) may fail to convince, since the nature
of the particular abduction depends a contingent attributes of the
abducer....  Abduction, says Sebeok (1983: 9) enables us to formulate a
general prediction, but with no warranty of a successful outcome; key
like Pierce does not seem to mind this, but there are strong objections,
from a number of points of view.


P. 336           (hermeneutic explanation)

I reject completely the view (especially in the historical context) that
Pateman outlines under the heading of the hermeneutic challenge (1987:
tenf)

	the concept of dependent existence of cultural objects implies 	
	that the proper scientific attitude towards them is the 	
	hermeneutic one in which we attempt to understand their meaning,
	and that explanation of them and all the relations of human agents  	to
them will be given in the vocabulary on beliefs, reasons, 		motives, and
purposes rather than in the make a mechanistic 	
	vocabulary of causes and laws.

At least I decree that clauses and the laws may be inappropriate, but
the others are even more so, at least in any reality other than
pragmatics.

Abandoning predictive or deductive explanation apparently leads us only
with untrustworthy post hoc strategies of explications.  This would not
in itself necessarily be a bad thing, if it were not for one troublesome
fact: because of the quirkiness of abduction(among other things), the
convincing force of any hermeneutic explanation depends on a number of
disparate and contingent variables; and not all of them seem to be
satisfactorily handled in any given instance, and may not in the genre
itself.  These include (a) the ingenuity of explainer; (B) the
interpretation's coherence with the rest of the explainee's putative
knowledgeable of the domain; and (c) most importantly to explainee's
willingness to accept the interpretation (which may as we will see
involve a good deal more than the satisfaction of simple logical
criteria, including aspect and temperament of practice). There is
something you didn't simply  dialogic about such explications, which is
not the case in the positivist mode; there if you accept the rules of
the game in the first place you have no problem with particular
instances.  Hermeneutic interpretations then are epistemically very
different from other types, and in that one has a choice not to accept,
since neither logical deduction nor any other kind of valid inference is
primarily at issue. All such explanations by permissive, not binding on
the explainee. (A good deductive explanation, even if it's boring, does
have the one excellent property that no sane explainee can reject it.)

P. 338
genetic linguistics, according to Anttila (1989: 23.2) deals of history
and human action; therefore paramedics is the only viable metatheory,
since the subject matter is maximally open systems in which mechanistic
prediction is impossible. So in history... piecing together a possible
past is guessing through indexes, where the laws involved somehow relate
to the common experience of being human, and results are traces of human
action.

Elsewhere Anttila says that

	reconstruction... means piecing together a possible chain of 	
	events, a state affairs, i.e. inferring the case.  The inference
involved is abduction (and induction...) not deduction, and the 	frame
this classical hermeneutic anamnesis (reenactment through
	interpretation and understanding) not natural science.

P. 339
I find impossible to make out what some of the terms that Anttila or
pateman use could possibly mean in the context of historical
linguistics.  How can one understand or get an intention behind or
discover the meaning or shift from ....

P. 340
The point of hermeneutics appears to be the discovery and formulation of
meaning. ... see Sir Medawar's elegant hatchet-job on psychoanalytic
theories.

P. 342
a classic and persistent example is the claim that optimal sound --
meaning correspondence is biunique, formulated by Anttila as the slogan
`one meaning one form'. The reasoning for ironing out of alternations..
is that `the mind shuns purposeless variety'...(MSPV principle)

P. 345
(new stuff... principle of isomorphism... regularization....  strict
one-to one symbolization is possible only in closed systems such as
formal (mathematical) languages...)

P. 346
change in principle, given the data we have, be random rather than
motivated, and the results would probably look pretty much the same....
Crucially, the notion of preference here is pseudo-statistical, no more
than universalization of an intuition....  Virtually all collections of
data often sort that I go of our tendentious, gathers to make specific
points and appearing only in the context of sketches of short store of
periods, or sense of examples designed to illustrate the operation of
MSPV....

Crucially, the notion of preference here is pseudo-statistical, no more
than the universalization of intuition.

P. 352
in a recent attempt to argue for the speaker-centred and
non-structuralist account of language change, in which Speaker
interaction is the primary locus and determinant, James Milroy (1992)
argues against the system based views of language. He suggests that if
languages are machines, ... then variation and   change ougth to be
dysfunctional, and thus not only inexplicable but counterpredicted.
((Suggested notion of system is excessively restrictive))..


P. 353    {systems}
relate complex biological systems, for instance, clearly are not, but
they do not lose any of their systematicity in this way.  It is only in
what might be called a folder adaptation as model of the organism... it
is not generally clear, as Gould has been arguing in detail for over a
decade that animals for instance, while certainly functional systems,
are often quite sloppy ones.  Parts are Jerry-built and cobbled together
out of other parts,....

{function and dysfunction}
for any complex system regular feature doesn't have to be either
functional or dysfunctional;...  It may affect the precisely the
non-functionality of such characters that is the primary enabler of
change (Lass 1990a)..

		{Redundancy and noise..}

It is in this light that we probably want to interpret linguistics
variability.  Languages are imperfectly in replicating systems and
therefore throw up  variants during replication; the fact of variation
itself is neutral.  If a variant were to be genuinely  dysfunctional in
some serious and compromising sense, the chances are that its survival
would be prejudiced.

So it's perfectly possible that both variation and change itself (as a
result) are neutral: even selection does not necessarily have to select
event which is better adapted. In any case, that are even in biology
modes off apparent selection that are not in the Darwinian sense
genuinely selective or adaptive (Kimura 1983). All of these
possibilities, given the much better understood the nature of variation
and change in organisms, need to be considered before any claim for
function can be made for variation or change.

A final point on systems.  Milroy's dichotomy between a system and a
means of communication is incoherent.  Once we allow for non- physical
systems, any means of communication that is to work as such most in fact
be a system; not only a system, but one that its share (within certain
bounds) by all users.


P. 355

In Lass(1980) idle to the general argument against functional
explanation, which was later attacked by ML Samuels, and others (SAmuels
1987)..

This argument largely concerned so-called homophonic clash and the
danger of merger neither of which I am willing to accept as rational
motivations for change, on precisely the same grounds with that I
rejected MSPV..

Let's consider a classical and family of functional explanation,
linguistics change (with apparently a more solid basis than MSPV, but
still essentially hermeneutic, in that it attempts to provide a reason
why some things should happen). ...  According to the neogrammarian
principle...(useful and largely correct) position as a basis, cases that
appear to violate the condition are interesting, and require
explanation.

P. 361
in summary, a conceptual basis for functional explanation is
underwritten by five in plausible assumptions:

1.  That exist functional and dysfunctional language states in the
normal of things, i.e. all language states are not equifunctional

2.  Speeches can have intuition is about the efficiency or a phonology,
etc. all their language for its communicational tasks (unless one is
true this can't be)

3.  Speakers can make comparisons between the present (structural) state
of language and some as yet unrealized one, and opt for one or the
other.

4.  Following from 3,speakers can have any global structural intuition
about their language.

5.  Speeches can change of the alignment on the basis called any toned
information in 1-4: hence change can be a motivated behavior.


These assumptions all involve the category mistake endemic in both the
generativist child -- as -- little -- linguist and hermeneutic
rational-agent models: confusing ordinary speakers for whom languages by
and  large in non-focal historical  given those special speakers for
whom making language focal is a professional concern.

P. 363

Curiously, then, individual list theory used actually neutralize the
individual proper, since they must make all individuals in affect
interchangeable.  What ultimately happens is this: the larger picture of
historical change is interpreted as if it all occurred in an individual
speaker... this individual picture is than projected onto a collective
language history, which becomes sum of a set of individual acts so alike
that anyone can stand for the type.
(heat bath and thermodynamics)

P. 365

As Bickerton (19791:461) has pointed out, typical patterns we observe
see impossible ' unless something, somewhere is counting environments
and keeping a running score of percentages'.  And individual list
interpretation of demand that each speaker not only keep track of his
own scores, but proposal other members all the group, so that if one
speaker goes off of the rails, he can adjust his own school or to
maintain the average.  And sense he does this also with interlocutors
from outside the community, if you must be able to keep his scores right
even when there is no feedback from other community members.

Bickerton doesn't think we can attribute this kind of sophisticated
running competition to individual, nor do I.  But as he says" something"
must be adjusting individual behaviors to conform with certain norms.  I
am not at all short what this might be (or if there's anything better at
all other than some mathematical property of such systems of competing
variants of themselves); but it clearly is not a generalized
individual's behavior can be hermeneutically understood or re-enacted.

It is difficult to see how this kind of things can be accommodated
except in a view of language has been at least partly transcendental or
meta-personal, on system with which speakers interact, but which used in
some sense outside them, and extra mental reality. (field theory).  But
one conclusion seems to emerge fairly strongly from the preceding
arguments, which reflects badly on the hermeneutic challenge.  Rather
than introducing a truly human element into the explanation of
linguistics change, hermeneutic explications is able to function only as
a totalitarian uniformizing imposition, since it is based on the
problematic assumption that one man's mind can stand for the minds of
all others. It brings us back two days before Weinreich, Labov, and
Herzog introduced the notion of orderly heterogeneity.

P. 366	Agents: structure, pragmatics and invisible hands

p.367
that he is, there are too quite distinct time dimensions needed to talk
about language....  Change occurs over geological time, beyond the
capacity of human staff, since no active density consequences on his
actions....

This is the geological time dimension, where speeches are not conscience
all that rolled improper dating variation, and indeed can't be: the
analysis on thrift, says Sapir (161) is certain 'to be unconscious, or
rather unknown to the normal speaker'.

P. 370		A modest ontological proposal

I conclude that hermeneutic or functionalist explanatory strategies are
not very satisfactory. And this is sold because of a fundamental mistake
about the nature of what is to be explained.  This mistake is
considering language change to the something that speeches to rather
than something that happens above languages.

P. 373
compare this spread all language change to level respiratory virus in an
epidemic.  We want to the disease spread throw population typically in
the same kind of exponential curve that characterizes lexical diffusion.
{prob dens again}.

P. 374
more generally: there are many phenomenon and the world that are
properties of particular kinds of systems, rather than the entities that
happen to make up the systems. Exponential growth curves this triable by
a certain kind of logistic equation are typical among extraordinary
number of natural and cultural phenomenon although most diverse kinds,
in such diverse fields as population genetics, the study of
predator-prey relationships and epidemiology, and may well be simply a
universal property of the growth off variant populations on the certain
classes of conditions rather than all the particular objects including
those classes. (Fibonacci, Pi, Planck's constant) etc)..  The
pervasiveness of abstract system types in different domains makes it
dangerous to argue with the kind of specificity Milroy does that a
particular sort of propagation must necessarily take its origin from the
local nature of the population through which it moves, rather than from
the nature of a process type, which may be a piece of world structure as
its were, rather than an attribute of a particular kind of lower-order
object in the world. {nonsensical or ignorant}.

p.375 {about replication and genetics}

So in a quasi-species sequence spaces each individual genotype its
computable as being a certain distance (in terms of differing nucleotide
sequences) from each other one, and the spread of genotypes into a
sequence space has the form, as Eigen puts it, of a cloud. {prob density
again).

What happens if we construe all language as this kind of object, which
exists needed in any individual Morton collective to, but rather as an
area in an abstract, vastly complex, multidimensional phase space....
For one thing, we can start to talk about variation and change the same
way we do for other populations: all language is a population called
variants moving through time, and subject of selection...

p.378
any imperfectly replicating (i.e. evolving) system throws up a (random
or near random) scatter of variants or imperfect copies overtime....

{examples, drift, conspiracy?...vector, attractor....)

p.382
speciation and the growth of new populations typically involved what
population geneticists call bottlenecks; when any subset of a
polymorphous or variable population splits off form the ancestral stock,
it will carry with it only that subset of variant characters that it
possesses....  So there is a major sense in which the development of a
new language represents, paradoxically, a decrease in variation. ....
If the description given above is valid, then in fact, monogenesis (at
any level from the origin of language to that of a particular low-level
dialect cluster) virtually entails contingent properties of the
population being transmitted in bulk and amplified via bottlenecks.
Such properties often have long viability periods. For example the first
personal marker /-m-/ has its survived in Indo-European for it least
some 8-10,000 years.

P. 384
I am rather asking here under what guise it might be most profitable to
look at human languages as a system in time, and what kinds of
consiliences fallout from a particular view.

P. 385
Milroy and others have argued eloquently (and in part successfully) for
embedding the historical study of language in a social matrix:
Andersen,Anttila, Pateman, Keller and Ikonen have argued (to my mind
less successfully) for incorporating accounts of human meaning,
interpretation, action etc.

P. 388
nothing in linguistics anyhow really seems to be that original; the line
I've been pursuing his bowl in principle is not in detail... (Davies) in
this view, language change was seen, like geological change, to be the
result of powerful non-human forces, in which human goals and actions
and no part: the speaker no more controlled the vast movements of
linguistics change than the farmers interventions control the formation
of geological strata. Output was not of course structuralist but
historical, though its focus was on large scale autonomous systems; but
the tension between autonomous or  naturalized  and speaker-centred or
psychologistic (or actional) orientations is old and recurrent.  And the
end both outputs are probably complementary; given enough history, the
paucity of major choices makes all innovations reactionary.

Looking back on this chapter, and parts of the rest of the book, it
seems that a great deal of the discussion and not been so much about the
ostensible subject matter as about epistemological style.  In
particular, the polemic in this chapter has been about incommensurable
difference in approach to the subject (and some peripheral connections)
that seems in his own way as important as any substantive matter is
actually dealt with.

{dichotomy--physics etc}

I suppose what I find most objectionable about Anttila, Shapiro and
others of the romantic persuasion is that they lack sobriety; and they
are enthusiasts....

The fundamental error although it hermeneutic approach is that it
attempts to get inside something that because of its demands historical
extension may not have an inside at all....  It abrogates the scientists
primary responsibility: to free investigation and knowledge from human
emotional attitudes, to step back to the position of a spectator who is
not part of the world on the study.


References:

Anttila, Raimo, (1972) and introduction to restore and comparative
linguistics, Macmillan Company,New York

Baldi, P. (ed) 1990 Linguistic change and reconstruction methodology,
Berlin: Mouton de Gruyter.

Bomhard, A. 1990 A survey of the comparative phonology of the so-called
`Nostratic' languages, in Baldi 1990:331-58

Boyd & Richerson, (1992) How microevolutionary processes give rise to
history in Nitecki & Nitecki 1992:179-210.

Cohen & Stewart (1994) The Collapse of chaos: discovering simplicity in
a complex world, London: Penguin.

Kimura, M. (1983) The neutral theory of molecular evolution, Cambridge
University Press.

Lass, R. (1980) On explaining language change, Cambridge University
Press.

Medawar, P. (1984) Pluto's Republic, Oxford University Press

Milroy, J. (1992) Linguistic variation and change, Oxford:Blackwell.

Nichols, J. (1992) Language Diversity in Space and Time, University of
Chicago Press.

Samuels (1987)
--
Best Regards,
Mark
-==-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
hubeyh at montclair.edu =-=-=-= http://www.csam.montclair.edu/~hubey
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
<><><><><><><><><><><><><><>----Language----<><><><><><><><><><><><><><><>
Copyrights and "Fair Use":     http://www.templetions.com/brad//copyright.html
"This means that if you are doing things like comment on a copyrighted work, making fun of it,
teaching about it or researching it, you can make some limited use of the work without permission.
For example you can quote excerpts to show how poor the writing quality is. You can teach a
course about T.S. Eliot and quote lines from his poems to the class to do so. Some people think
fair use is a wholesale licence to copy if you don't charge or if you are in  education, and it isn't.
If you want to republish other stuff without permission and think you have  a fair use defence, you
should read the more detailed discussions of the subject you will find through the links above."