the meaning of "genetic relationship"

[Isidore Dyen]

----------------------------Original message----------------------------
On Mon, 22 Jun 1998, bwald wrote:
 
> Isidore Dyen writes:
>
> >Theoretically (and by
> >definition) two languages are interrelated (related to each other) if they
> >separately continue what was once a unitary (but not necessarily uniform)
> >language.
>
> The distinction between "unitary" and "uniform" leads to problems, as we'll
> see.  For the moment, I suppose that "not uniform" is mentioned because all
> observable "languages" are of that kind.  At the same time, most of them
> are assumed to have evolved from a single "system".  Already some hedging
> is necessary since it is problematic that "Black English/Ebonics" or
> whatever you call it, is descended "unitarily" along with "the King's
> English" (and "Webster's"), and Sranan is so different that from most
> "Englishes" that the issue doesn't even arise in a useful way.
 
The assumption that languages are unitary is a matter of definition.
Perhaps you would prefer a new term like 'hololect' instead of the term
language; a hololect is unitary and composed of dialects. Dialects are
connected ot each other in a hololect through a chain of pairs of mutually
intelligible dialects. This is our or at least my conception of the
way the realm of the world's language can be viewed and i it is in this
sence that they are wholes. It does not matter to me whether you find this
interesting or not, but if I then assume that this type of realm has
persisted from the origination of languages through a succession of native
speakers, I can make inferences about the history of languages
including to some extent the shapes that appeared in their earlier
stages. Again it does not matter whether you consider this interesting.
Such assertions of interest are personal and not subject to proof in any
sense and are thus irrelevant.
Another assumption is that there are different hololects in our world.
Chinese and English are different hololects. Likewise English and German
are different hololects. The fact that from a practical point of view the
meaning of 'mutual intelligibility' is not refined enough to permit us to
determine in every single case whether it is present or absent is a matter
of a lack of scholarly interest, but its direct relation to the primary
function of language recommends it as a criterion for classifying
languages.
Another assumption is that one hololect can become two by the
disappearance of any connecting pair of mutually intelligible dialects.
These assumptions can be regarded as part of a set of definitions and
theories that permit us to deal with the past of languages. They are not
all the assumptions. More will appear.
 
 
> Dyen continues:
>
> >To demonstrate that it is likely that two languages continue the
> >same unitary language it is necessary to show that they exhibit systematic
> >correspondences, better called collateral correspondences to distinguish
> >them from lineal correspondences that a language shares with its earlier
> >stages. It is these separate lineal correspondences that form the
> >collateral correspondences that are used in the reconstruction of forms of
> >the original unitary language.
>
> Fine.  And collateral correspondences work for much of the BE and Sranan
> *lexicon* with other "English".  But the "whole" I referred to that Dyen
> quotes (see below) includes morphology, syntax etc, for which BE is
> problematic in some cases, and Sranan is more generally unlikely (and no
> one tries to derive Sranan syntax from Old English, let alone
> Indo-European).
>
> (NB.  I'm taking liberties with Sranan as "English" for the sake of Dyen's
> points on  genetic relationship, since no one considers Sranan "English"
> (as far as I know, certainly not the speakers, or those familiar with the
> language).  Maybe we should be discussing whether "Flemish" is "Dutch"?  Or
> whether Catalan and Provencal are the "same" language?  I'm anticipating
> Dyen's mutual intelligibility criterion for a "unitary" "language"
> (discussed below).
>
> Dyen goes on to later say:
>
> >...speaking about languages as 'wholes' is not loose talk.
>
> I said it was, with respect to the assumption of genetic relationship on
> the basis of partial reconstruction (usually of some lexical material),
> which is what he is responding to in context .  He immediately continues to
> say
>
> > A  language, technically as opposed to a dialect, is a bounded chain of
 pairs
> >of mutually intelligible dialects. It has a boundary that it shares with
> >each other language since none of its dialects is mutually intelligible
> >with any of theirs.
>
> The last statement is false.  There is no way to set a boundary to
> distinguish one "language" from *some* other on the basis of mutual
> intelligibility.  There is no "technical" sense of the word "language" that
> can do this in practice, i.e., that corresponds to something observable
> and/or, in some way, testable.  It is a vacuous attempt at a definition of
> "language" (in a "technical" sense).  Historical linguists do not concern
> themselves technically with the notion of "mutual intelligibility".  As
> soon as we have lack of "uniformity" we already have the possibility
> (indeed the virtual certainty) of "mutual unintelligbility" on some point
> or other.  That has nothing to do with whether we are dealing with two
> "languages" or two "dialects" of ONE language.  The claim made in the last
> statement is not helpful, as far as I can see.  Is it meant to apply to the
> difference between obviously distinct languages like "English" and
> "Chinese" and/or to distinct related branches like "Slavic" and "Germanic"?
> That is not a problem, and it is not related to the problems I raised
> above.
>
  See my comments above.
 
> Meanwhile, he continues:
>
> >In this sense it is a whole. What you speak of as
> >'genetically related' parts--with the implication that some parts are not
> >'genetically related'--are more commonly called 'cognate' or 'shared
> >inheritances'.
>
> Obviously "languages" descend from "wholes".  The problem is that their
> parts may descend from the parts of different "wholes", and to some extent
> they always do. Just how much of a "language" continues the "whole" of some
> earlier single language is determined by extensive research -- never
> completed, but sometimes overwhelmingly favoring one former "whole" over
> others.  My point was that you don't know how much of an observed language
> descends from some former single "whole" until you do the research.  The
> historical literature on syntax, and to a lesser extent on phonology, is
> loaded with suggestions about "borrowing" as motivation or actuation for
> this or that change.  That already presupposes (not always validly) that
> "inheritance" has already been established for relevant, though different,
> points.  "Relevant but different", now how does that work?  The most
> vulnerable assumption, I think Dyen would concede, is that if much
> vocabulary and even some morphology, is shared by two "languages" then they
> *must be* genetically related as WHOLES even if most of their syntaxes are
> historically unrelated.  (The "bad" literature, e.g., attempting to
> exclusively derive peculiarities of Afrikaans from random localised Dutch
> dialects, or BE from random localised British dialects, shows the dogmatic
> operation of such assumptions; they turn out to be historically
> problematic, and the least that can be said is that they show that whatever
> changes they are used to explain are presumably "possible" *internal*
> changes in the some variety of the "language" at issue.  Beyond that, a
> factual historical account of the evolution of the varieties in question
> remains problematic, once the methodological dogma is dismissed as
> misleading.)
 
 
If languages as hololects are assumed to be wholes, it follows that any
part of a language has descended from a hololect which was its earlier
stage. It is assumed that hololects do not mix. Under this assumption each
observed hololect is the endpoint of an infinite (i.e.
uninterrupted) sequence of stages originating in the first hololect (i.e.
the first nad only language) in the world. I sympathize with our
impatience with certain types of attribution, but this should only
motivate you to do better.
 >
> Next,
> >The remainders are composed of individual inheritances and
> >innovations, the latter including borrowings.
>
> No conceptual problem here.  Much practical problem.
>
> >The reason interrelated languages are treated as wholes is that each
> >represents a separate continuation of the original unitary language via a
> >succession of native speakers, their separation occurring at the moment
> >the last cross-pair of mutually intelligible speakers had vanished.
>
> As might be assumed from what I said above, this is not an interesting or
> even practical criterion for "language".  It only serves my suspicion that
> Dyen has major = unproblematic branching in a tree model in mind.  The
> changes themselves are what's most interesting in historical linguistics,
> and what one change has to do with another -- if anything -- often a
> difficult problem to solve.  This has nothing to do with the difference
> between "dialects" and "languages", I repeat here for emphasis.  Of course,
> loss of mutual intelligibility with time, presumably accumulation of
> changes, is also interesting, but has hardly been studied.  As I said
> before, it is particularly interesting to study within a SINGLE "language".
> Branching becomes problematic when one branch shares features mutually
> excluded between two other branches and both are innovative features.  Then
> we have a branching problem which is identical to the ubiquitous branching
> problems in classifying "mutually intelligible" dialects of a single
> "language".
>
> >It is an assumption of the comparative method that different languages do
> >not mix (under natural circumstances).
>
> That assumption is known not to be valid, notwithstanding the resistance
> that cannot resist asserting that mixture is "rare". (Since code-switching
> is extremely common, the assertion seems to claim that mixture does not
> arise from grammatically conditioned code-switching.)  BTW, without pencil
> and paper, mixing can only occur under natural circumstances.  One cannot
> intentionally spontaneously mix languages the way they have arisen in
> nature, e.g., Michif, Aleut-Russian, whatever, i.e., switch between
> languages on the basis of the grammatical category.
>
> Creoles are either aberrant
> >dialects if they are still part of a chain with other dialects or
> >different languages if they are not.
>
> What "chain"?  Is this the chain of mutual intelligibility?  If so, how
> does that have anything to do with the "unitary" origin of the "creole"
> within that chain?  Maybe Dyen is thinking of some kind of relatively
> radical "restructuring" which he might insist on viewing as "internal
> change" and thus continuation of a unitary "language".  Otherwise, what's
> the point?
>
The point is to maintain the nature ot the hololect. The term creole is
applied to some dialects that are, though aberrant still part of a large
whole by virtue of being connected to it by pairs of mutually intelligible
speakers whereas others are not.
 
> >In the latter case its first native
> >speaker was not mutually intelligible with any dialect of any of the
> >languages that contributed linguistic material to it.
>
> I'm not sure what the point of this consideration is.  Is it about
> "convergence", which I suppose from Dyen's perspective is cumulative
> borrowing from the base to the creole?  Again, in the case of Hawaiian
> Pidgin, i.e., an English-based creole, that is not the case any more than
> it is the case for all kinds of dialects of English regardless of their
> ancestry.  And in practical terms we must now realise that "mutual
> intelligibility" is a theoretical notion which has not been defined in
> Dyen's perspective.  It remains to be seen, for example, whether some
> English-speaking area in the Midwest US would find "Hawaiian Pidgin" or
> Glasgow (working class) English more difficult to understand.  And it does
> not strike me that the results would be relevant to Dyen's proposal.  Would
> they refute it if Glasgow was more difficult to understand than Hawaiian P?
> To take an extreme example, a makeshift pidgin based on English would
> probably be easier to understand for a relevant group of English-speakers
> than Glasgow English for comparable messages.  Yet, such a pidgin as a
> WHOLE hardly descends from (Old) English (indeed a makeshift pidgin is not
> a "whole" in the sense that any historical language is, or in any sense
> amenable to linguistic analysis), while, for the sake of argument, Glasgow
> English as a WHOLE does descend from (Old) English.
 
No. The point is this: a creole hololect must be mutually
unintellibligible at its start with those from which it draws its
linguistic matter, for if not, then it is mutually intelligible with at
least one and so a dialect of that one.
Creole hololects originate in situations in which a number of langages are
in competition and a pidgin develops for the convenience of all, but is
in the first place a second (non-native language) for all and not mutually
intelligible with any of the contributing languages.
 
A child born into such an environment might acquire that pidgin as his
first language, thus introducing it as a member of the native languages of
the world. We have no record of the appearance of such languages prior to
the period of colonialism.
 
> I appreciate Dyen's attempt to articulate the traditional assumptions of
> "genetic relationship", but I question the success of that attempt.  I find
> the part about "mutual intelligibility" least succesful, and unnecessary.
> -- Benji
>
See discussion above. Enjoy. ID.