Cladistic language concepts

bwald bwald at HUMnet.UCLA.EDU
Wed Aug 26 12:32:03 UTC 1998


----------------------------Original message----------------------------
Michael Ghiselin writes:
 
>               Surely there are adaptive changes in language through
>          time, at least in the sense of coinages and borrowings when
>          there are new conditions of existence.  The advent of new
>          modes of transportation obviously evokes a needed
>          vocabulary.
 
Linguists do not differentiate languages for complexity on the basis of
such things as how many names they have different kinds of machines (or
snow, cattle, etc.) because size of vocabulary is not seen as an *inherent*
property of the SYSTEM of any particular language.  More could be said
about this, including some problems, but linguists consider vocabulary size
a trivial basis for comparing the complexity of languages, and tend to
consider change in vocabulary (as it reflects change in what's recognised,
referred to and talked about) as a trivial kind of change.  Certainly such
things are not trivial as reflections of change in the social environment
in which the affected language is used, but they are trivial in terms of
the notions of linguistic system that linguists have and continue to focus
on.  Anyway, it is totally unclear what the biological analogue of a "word"
(or "lexical item" or "vocabulary") is.
 
He continues:
 
>               I would argue that the same kind of entropy exists with
>          respect to languages, and that if they do maintain their
>          organization it is due to something counteracting the
>          tendency to decay.  Consider the vocabulary as it gets
>          passed from parent to child.  The probability that every
>          single word will get transmitted has to be somewhat less
>          than one.  But as everybody knows, children and adults alike
>          coin words when they need them.
 
I don't know about the entropy notion in this context, but linguists are
indeed impressed by the existence of processes in all languages which allow
users to coin new words.  The processes vary from one language to another,
but all languages have them and can use them equally successfully (as far
as we can tell).  Total continuity of vocabulary is not important.
Productive/creative processes equal to the task of conveying and
recognising "meaning" is what impresses linguists about linguistic systems,
and where they can't find one language more complex than another, or more
adept in some other way.
 
Next,
>       A cell is often part of more than one organ
>          system, but I cannot think of any that are part of more than
>          one organism, although they can move from one organism to
>          another.
 
Now definition of an "organism" might matter.  Siamese twins of various
degrees of severity comes to mind.  If the "sharing" is slight, e.g., a
non-vital limb (if that happens) one might be inclined to recognise TWO
organisms which have partially melded by chance.  It might be cultural for
us to imagine that "a" person with two heads would have to be "two" people
with "one" body.  "twin" implies and is etymologically related to "two" (in
English), but the concept of "organism" belongs to a more specialised
expert system.
 
        Organisms are never part of more than one species,
>          but they can be part of more than one club or other
>          organization.
 
Although I used "species" as analogous to "historically continuous
language" (as implied by such linguistic terminology as GENETIC
relationship), there are other viewpoints.  On some level, the
universalists -- linguists who are more impressed by what languages have in
common (whatever that is) than how they differ -- might argue that ALL
languages belong to the same "species", a species ultimately determined by
the human species, and esp certain neurological structures in the brain
enabling language and, according to innatists, unique to the human species.
 
        A bilingual person may be said to participate
>          in more than one language, though I suppose nobody considers
>          such a person a part of either.
 
No.  That's a social issue.  Bilinguals can be parts of either or both
languages, depending on the social circumstances.  They are not necessarily
recognisable as bilinguals when they spread particular changes which have
germinated in their communities to monolingual communities.
 
>       The person's idiolect is
>          supposedly a part of a language, so the person would have
>          more than one idiolect.
 
Idiolect is a tricky word.  If it means the totality of an individual's
repertory, then monolinguals and bilinguals cannot be distinguished in this
way.  If it means anything else, monolinguals and bilinguals will both
still have more than one, because we can observe that people change the way
they talk depending on the circumstances.  Everyone is no doubt unique in
the complete details of their repertory.  Idiolect is not a user-friendly
term for historical linguists; it obscures what it means for a "language"
to "change", since it is associated with the individual, and the individual
can only be recognised for "change" by comparison with other individuals.
Generally, the "language" consists of what a socially (yes!) coherent group
of "idiolects" HAVE IN COMMON at some point in time.  That is never uniform
over what is called a (particular) "language", so "language" is always
somewhat variable and abstract at any point in time.  Nevertheless,
whatever they have in common (to the extent that they ALL do) can and does
ALSO change from one period of time to another.  Therefore, it is not
vacuous to abstract the "language" from the overlap of "idiolects", despite
the problem of irreducible variation, and still speak of change in
successive collections of "idiolects".
 
(I don't have much use for the term "idiolect", but many linguists accept
the concept/s for whatever reasons they have, so I'm not challenging it
here.)
 
>       But transfer from one idiolect to
>          another within the same person and hence across languages,
>          would not imply that there was just one idiolect or just one
>          language.
 
I already made a comment about "idiolect" above.  For the above passage I
would also point out the concept "transfer".  That is indeed the usual word
used for the process by which features are recognised to have originated in
one language and come into some varieties of another, and it is also often
associated with what individual bilinguals do.  However, we recognise
transfer from a comparison of different speakers of the "same" language
(for one or both of the languages involved), and it is most often the case
that the bilingual is not aware of doing "transfer", but only of applying
the same strategy (whatever it is) to both languages.  If all languages are
the "same" at some level (the universalist hypothesis), then "transfer"
simply consists of failing to distinguish what is true of all (or at least
the relevant) languages from what is specific to some particular languages.
We say analytically that A is transferring something from language X to
language Y, but the doer is often unaware that there is any difference
between X and Y TO BEGIN WITH.
 
(Bilingualism can be more complicated than this, but this seems sufficient
than now.)
 
>               The situation is rather like what we encounter when a
>          certain amount of gene flow occurs between populations
>          through hybridization.  But only partly.  To get something
>          like a bilingual person we would need organisms with two
>          independent genetic systems that can coexist and get
>          transmitted separately.  I can imagine that, but to my
>          knowledge there is no such thing in nature.
 
Again, it may be useful to compare languages with species, but then again
maybe it's not.  Maybe the best analogy is between languages and
sub-species/genera or whatever of a single species.  Maybe languages differ
from each other only like, say, different dogs (which vary anatomically
much more than different people), but not like dogs and paramecia, or
perhaps even dogs, bears and seals, etc.



More information about the Histling mailing list