What is Relatedness?

[Dr. John E. McLaughlin]

Steve Long wrote:

"I am simply trying to proceed one point at a time.  The only issue is how
you know that evidence of what was shared has not disappeared."

I've got a couple of points to make on this issue.  If two modern languages
had a common origin very long ago (past the 10,000 year boundary layer, for
example), then it's highly probable that they have lost or changed in
opposite directions so many common characteristics that their relationship
is no longer recognizable.  If we assume monogenesis, for example, then
Basque has modern relatives, but all evidence of the relationship is gone or
changed beyond recognition.  In other words, if two languages have changed
so much over a long period of time, then the evidence, as you rightly say,
IS lost or unrecognizable.  To tie that comment directly to the above quote,
in this case we CANNOT know anything about what has disappeared or even
whether or not it was there in the first place.

My second point, however, calls into question your assumptions that this
case must also apply to Indo-European in many respects.  The process you
mention (complete loss or unrecoverability) is a fixture of DEEP time depths
and prevents us from crossing the BOUNDARY.  Indo-European is a relatively
young family (compared to, for example, Australian, whose time depth is on
the order of 40,000 years and whose member languages have been subject to
areal borrowing and dissemination of features on a scale that isn't really
surpassed anywhere else [you should read R.M.W. Dixon, The Rise and Fall of
Languages]).  Let's look at English as an example of a "radically changing
language".  If we look at the lexicon of English, we find that well over
half of the total vocabulary is Italic or Hellenic in origin.  But if we
look at the 1000 most common words, then they are 80% inherited from
Proto-Germanic.  Within the morphology of the language, much of the
Proto-Germanic morphology is overwhelmed by derivational affixes borrowed
from Italic and Hellenic.  Yet, once again, when we look at the most common
morphology it is Germanic.  Our syntactic structures are also clearly
derived from Germanic structures.  But that is working with productive
evidence and is not to your point.

When a feature is "lost" in a language, it does not suddenly disappear, and
usually does not completely and totally disappear even after a long period
of time.  Take, for example, the strong verbs of English.  These reflect a
productive process in Germanic for forming past tense verbs by vowel
modification.  In English it is dying out.  It is not productive, but it is
not gone either.  We see the process in decay in many ways as it becomes
less common, but there is still substantial evidence for it.  (My wife and I
were playing Boggle tonight and she found the word 'dived'.  For the life of
me, I couldn't make the past participle of 'dive' sound at all right.  'I
have dove' is certainly ungrammatical, but 'I have dived' is equally
obnoxious to me.  I'm glad I don't enjoy swimming!)  As another example of
how "long lost" are actually not really lost, but only hiding, in
Proto-Numic (ca. 2000 BP) from the Great Basin of the US (Uto-Aztecan
stock), geminating the medial consonant of a verb stem (overwhelmingly CVCV)
indicated that the verb was durative in aspect.  Some of the daughter
languages have exactly the opposite meaning for gemination, i.e.,
momentaneous aspect, and some have lost the feature altogether.  However,
the key to determining that this feature is not "lost" in any of the
languages is that a very small (and dwindling) set of verbs in each of the
languages also mark plural object or subject by geminating the medial
consonant.  It is this internal reconstruction that allows us to see
evidence of older features of the grammar of a language even when that
feature has long since become totally unproductive.

It's been my experience after 20 years of looking at comparative data from
Native America (where there is no written tradition), that using internal
reconstruction can triple or even quadruple the amount of good phonological,
morphological and syntactic evidence for demonstrating relationships or
doing subgrouping.  The time depth of Proto-Indo-European is well within any
maximum limit to the effectiveness of identifying internal evidence of
formerly productive processes in the great majority of languages.

So then:

Can every piece of evidence for a relationship be lost or unidentifiably
changed?  Yes, but only at very great time depths (at least over 10,000
years).

Can vocabulary be borrowed to such an extent that evidence of a relationship
is lost?  Not that I have ever seen.  Even when a language borrows really
extensively from another language (as English has borrowed from Italic),
there is always a base level of commonly-used vocabulary that still retains
its ancestral lexicon.

Can morphology and syntax be borrowed to such an extent that evidence of a
relationship is lost?  Most likely not.  While languages in long-term
contact, such as the languages of the Northwest Coast of North America, can
exhibit a great many similarities in phonological, morphological, and
syntactic structure, there are always significant differences.  Internal
reconstruction can be used to identify structures characteristic of
ancestral relationships even when none of the structures are still
productive.  Many comparative linguists actually consider the results of
internal reconstruction to be better evidence for a relationship than
productive processes.

There may be a few features of Proto-Indo-European that are either lost or
unrecognizable by any means in some or all of the daughters, but they are
undoubtedly a small minority compared to the array of recognizable features.

John E. McLaughlin, Ph.D.
Assistant Professor
mclasutt at brigham.net

Program Director
Utah State University On-Line Linguistics
http://english.usu.edu/lingnet

English Department
3200 Old Main Hill
Utah State University
Logan, UT  84322-3200

(435) 797-2738 (voice)
(435) 797-3797 (fax)