<Language> Lass: Part 2 of 3; Electronic Review

[H. Mark Hubey]

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

Lass, Historical Linguistics, 1998.

-------------Part 2 of 3 --------------------------------------
p. 215 		The Nature of Reconstruction

This chapter deals in more detail with phonological reconstruction, and
includes some consideration of the rather different kinds of problems
that may arise outside of phonology.

p.217 [an example}
So given correspondences of the type {p,f,0} we ought to make two
reconstructive assumptions (regardless of the attestation in our corpus}

1. A correspondance {C,0} where C is some non-glottal voiceless
consonant implies a pathway C > h > 0.

2. Therefore the zero-dialects went through an /h/ stage.

...A theoretical imperative generates occult complexity, which surfaces
as history.,,,In this case I invoe a fairly conventional lenition
hierarchy (Lass & Anderson 1975;ch.5) which can be illustrated with the
velars:

...

These are the guiding assumptions

1. Ceteris paribus, movement down the hierarchy is more likely than up.

2. The lower a segment is on the hierarchy, the more likely a further
descent (`inherent weakness', or zero as attractor)

3. Generally, changes tend to proceed one step at a time (assuming the
`steps' represent from one `universal category space' to another.

Historical evidence for these particular constraints is close to
overwhelming (Lass 1978:ch.6) but the type-residues left by historical
change in particular instances may look quite different. This is one of
the crucial historical/synchronic dividing lines: change can leave or
distorted or misleading traces in the data which require principled
sorting. What warrants this procedure ('principle overrides data') is
that at least in particular cases, and in enough of them, it has
considerable empirical backing.  Again and again, so often that we're
inclined to think of it as the norm, if we look hard enough we find
material that justifies the hidden sequences in particular cases. And
the more such successes we have, the more confident we can feel about
invoking the  techniques in cases with less information. Fortunately,
the micro-dialectology and micro-history of many languages will provide
the information we need; though when a family is represented by a small
number of distantly connected and  highly differentiated languages the
relative roles of principle and data are skewed in the direction of the
former.

p.221 		Quanta and phonetic gradualism: a few suggestions

I have been making a pitch for the venerable principle of gradualism
(nature non facit saltum). In biology it is associated with Darwin, and
is enshrined in the general preference in modern evolutionary theory for
`micro-mutation' as opposed to `saltation' as the source of major
apomorphies, even radical taxonomic distinguishers (Gould 1983,Ayala
1983)

It would be nice to have a hard theory of the size of the units or
`quanta' of phonological change...What is a permissible step in change
depends on  some antecedent theory of primitives. If the smallest
theoretically significant phonological primitive is the (binary)
distinctive feature (Jacobsonian, or Chomsky/Hallean, or whatever), then
obviously no changes can be any smaller than the size of such a feature.

p. 228	Projection again: conventions and justification

1. process naturalness
2. system naturalness
3. simplicity
4. phonetic legality
5. family consistency
6. oddity condition
7. portmaneau reconstruction


p.237
Many Uralic languages show a phenomenon called `consonant gradation', in
which particular consonants in certain syllable-configurations have
different `grades'....

p.238
But there are two apparent violations of the `one-step' principle of
descent down the hierarchy: /p/ > /v/ involves two stages, and /k/ > 0
traverses the entire hierarchy in one step. There is an apparent
difference between what must have been the historical situation, and its
synchronic residue...

p.270	Phonetic realism: the art of coarse transcription

{example given of pig...}
Coarsely, this means that if we met a Proto-Indo-European speaker we'd
expect that his word for `pig' would begin with some kind of voiceless
labial stop. A symbol like [p] in this case is a claim that a decent
phonetician would want to use IPA [p] in notating at least a good part
of the initial stretch of this utterance (and hence for the type of
which the utterance would furnish a token); and that there would be no
real temptation to use symbols like [t,k] on one dimension or [b,beta,w]
on another.


p.272
Mary Haas (1969:31) observes that `any language is an actual or
potential protolanguage'. This of course is trivially true: if a
language survives across even one generation it is the protolanguage for
the undoubtedly somewhat changed language of the next.

p. 277			Chap 6  Time and Change

p.278
What people think of as `changes' can be grouped under five major topoi,
each of which might have a classical or antique, or at least literary
slogan;

1. Change as loss

2.  Change as (neutral) change of state of flux

3. change as creatio ex nihilo; `In the beginning God created the Heaven
and earth.'

4. Change as degeneration: `Change and decay in all around I see.'

5. Change as progress:`Till we have built Jerusalem/In England's green
and pleasant land.'

Types 4 and 5 may just be ideologically colored versions of 1-3.

In the first, changes the serious ontological matter: one individual or
type or natural kind becomes another, there is a transsubstantiation or
trans-individuation....

In the second type, there is no state change, but merely `substitution'
of another (normally pre-existent or potentially pre-existent) state.
This one is adumbrated in Sweet (1900:ch3), magisterially laid out in
Hoenigswald(196) and carried to its apogee in Katicic (1970).

Sweet distinguishes what he calls `organic' change (phonetic change due
to mishearing, mis interpretation ,assimilation, gradual phonetic drift,
etc.) from external change e.g. analogy, and the like.

Katicic (1970:ch 3) on the other hand pushes the substitution idea to
its ultimate conclusion; even internal phonetic change is a 'change in
relations between languages and communities rather than a change in the
languages themselves'. Since all language varieties belong to an
atemporal (apparently quasi-platonic) universal variety set, all change
can be construed as substitution; all conceivable outcomes exist in
posse. A history doesn't show change at all, but rather a succession of
different languages, which can be seen as`one language with some
internal variety'. Unlike Sweet (or Hoenigswald), and like Saussure but
more so, Katicic manages to get rid of time and history completely;
history (in the week sense in which there could be said to be any at
all) is just permutation of what is essentially there already and always
has been.  Problems of filiation and the like are resolvable to
operations on the universal variety set that lead to `restricted
diversity'.

p.281
Historians, like all other linguists, are practitioners of the `Galilean
style' (Botha 1982); they work with categories of high abstraction and
idealization. But it's not always clear (maybe not even to all
historians themselves)  just how far the abstractions go; sometimes we
can learn interesting things by looking behind the idealizations, and
finding out what we lose by getting what we gain from them.

p.290                   Linguistic Time (Arrows and cycles)

The modern physical sciences now recognize both types of: reversible or
classical (Newtonian) time, and non-reversible or thermodynamic time.
Reversible time in physics is not normally construed as cyclic; though
specific reversibilities can lead to equilibrium, which may be.

p.291
Whereas under a thermodynamic regime (say you the current order of the
universe on the heat death interpretation), entropy increases in a close
system and leads to a static equilibrium, maximal disorder, etc.  This
has been a basic philosophical problem in physics since the
19th-century, when classical dynamics, in which time is reversible, came
into an apparent conflict with the irreversibility associated with the
second law of thermo dynamics, and the idea of entropy. ... the idea
that there is an immutable substrate or background to all or temporal
experience is compelling, since it imposes the special kind of order on
the universe; one could see the same kind of motivation in Katicic's
position as in Parmenides's or Newton's or Einstein's; if timing change
on illusions, the universe has a `ground' that blacks in the presence of
general line transformation.

Such visions animate not only larger scale philosophical or Cosmo
logical schemata, but may manifest more locally, in linguists attitudes
toward their own subject matter, particularly as it unfolds in time.
The issue of the direction and linguistic evolution has been of interest
since antiquity; as early as Plato's Cratylus the idea was broached that
if the first Namer named things by nature and not by convention, the
form/meaning fit has become (unidirectionally) less transparent.

p.292
During the 19th and 20th centuries, three more or less articulated views
of the shape of linguistic history have emerged, each for its radical
proponents the result of some kind of directional law.  These overall
metaphysical characterizations seem to be of two general kinds:


1.  	Uniform directionality..  There are three main types:

	(a) positive (Progressivism): languages evolve a particular optimizing
direction, becoming more efficient or simple or sophisticated or
whatever. ...

	(b) negative (decay): languages move from a perfect type towards some
less perfect one: e.g. Bopp (1833) from analytic into a synthetic, for
Jesperson the opposite.

	(c) non-evaluative: there simply are directions, either in actual
glottogenesis (from a primitive state) or in the evolution of languages,
but those do not necessarily have anything to do with quality (perhaps
Humboldt 1882).


2. Cyclicity.  Languages moves through life-cycles like organisms (cf
Davies 1987); they may have periods of youth, maturity, and senescence
(as in 1b), but recycle over and over, e.g. each great type comes around
again after language has passed through the others in some particular
series: e.g. isolating> agglutinative > inflecting/fusional, etc (von
der Gabelentz 1981), at a local rather than a global level Meillet
1912.)

Few scholars now would believe that any of these principal legislates
for language change overall: there are no global directionalities fixed
by natural law.  Individual histories (or parts of them) can be any one
of the above.  Though, especially in the work growing out of the
tradition of grammaticalization studies started by Meillet (1912) and
revived recently by Elizabeth Traugott, Bernd Heine and others (Traugott
& Heine 1991), Hopper & Traugott 1993), certain directions are
increasingly being singled out as major or overwhelming.


p.293
The history of any dynamic system can be mapped as a trajectory in a
multidimensional space (phase space) where each point in the space
represents a possible system state.  By a dynamic system, I mean any
evolving ensemble where variation of parameter setting produces a change
of state.  Under this (relatively standard) mathematical definition, not
only a mechanical or thermodynamic systems (e.g. a swinging pendulum,
convection in a needed fluid) dynamical, but so are evolving
populations, whether systems and even valued sets generated by
completely abstract equations, where changing the numerical values
satisfying some control parameter produces an evolution.  Such evolution
was maybe partly linear, or at least show continuous change, but may
then settle into other configurations....

Dynamical systems in general can be characterized as tending to move
towards regions in phase space called attractors: an attracted as a
region 'such that any point which starts nearby gets closer and closer
to it' (Stewart, 110).  In simple and rather loose terms, and attractor
is region into which a system tends to set up, and in which it remains
unless it is dislodged in some special way.  The most common or typical
attractors are single point attractors or sinks, and limit cycles.  The
precise mathematical definitions is not at issue here, since this is a
heuristic rather than mathematical discussion; what counts is the image
of an evolving system as a kind of flow in some n-dimensional space, and
the existence of regions in that space towards which the flow tends to
converge.

P. 294
Such imagery and terminology are very general, and apply to innumerable
evolving systems, both purely abstract and physical.  This kind of
language was originally developed for talking about quantifiable
mathematical systems, but that are (at least so far) non-quantifiable
systems that exhibit this same type of behavior, or at least have
properties similar enough so that we can  informally but appropriately
borrow the terminology.  The point of such borrowing is that
terminology's neutral with respect to content though the system; to put
another way, a general dynamical description is a syntax without
semantics.  Such a neutral expository language allows us to talk about
the shapes of historical developments without an ontological commitment,
and may lead us see things that we would not otherwise, or at least see
things differently.  The larger -- scale philosophical implications of
this point will be taken up in 7.6; for now I am interested mainly in
the utility of the notion of trajectories and related concepts for
talking about histories as trajectories in time.  Their function for
moment is defining types of temporal configurations that seem to repeat,
and serving as a source of generalized images for visualizing them as
trajectories.  Sinks and limit cycles are what might be called typical
or ordinary attractors.  But there is another type, appearing in system
after system, which is rather different properties.  Such a strange
attractive user region in phase -- space (typically found when a system
is far from equilibrium) in which the behavior on the system becomes
increasingly unpredictable and chaotic, and parameter values less and
less orderly, and less and less likely to repeat. But within such
attractors are often occur what are called windows of order in which
orderly phenomenon are apparently self generated out of the chaos...
That is (deterministicaly) chaotic systems can generate their own
order.  It is becoming increasingly clear, both in chaos theory and the
developments now often grouped under complexity theory (Lewin 1993) that
the edge of chaos regimes in all sorts of natural (and artificial)
systems in which self regulation and order are generated out of apparent
disorder (this is sometimes referred to as autopoiesis).  The evidence
for this kind of temporal trajectory isn't relevant for the historical
linguists, because RMON other things he suggests that there simply are
rather general system types that behave in certain ways, regardless of
what the systems are composed of, or who happens to be using them.

P. 295

Many evolutionary pathways in language change seem to lead to sinks... a
good example is the set of phenomenon now usually grouped under the
general heading of grammaticalization.  For instance (cf. Givon 1971,
Comrie 1980, Campbell 1990c) it seems that case markers typically (even
according to Givon and some others exclusively) evolve out of
grammaticalized free nouns, along the pathway, Noun > Postposition >
Clitic > Case-marker.  The step along this pathway seems irreversible or
nearly so; once a now has become a pulse position it can't become one
out again, a case market cannot detach itself than become a
postposition. {examples from Kannada, Estonian, Hungarian}

{note Sumerian noun-adjective could have given rise to agglutination}

P. 296

Developments so this kind can be construed as paths along a chreod
leading to a point attractor..


--
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."