🚀 go-pugleaf

In article <Jan.4.16.54.01.1993.22727@pilot.njin.net> hubey@pilot.njin.net (Hubey) writes:
>In article <1993Jan4.205115.8521@leland.Stanford.EDU> alderson@elaine46.Stanford.EDU (Rich Alderson) writes:
>
>> In article <1993Jan4.071434.27930@jyu.fi>, ryyti@jyu (Jarmo Ryyti) writes:
>> >Rich Alderson (alderson@cisco.com) wrote:
>
>> In the nineteenth century, one accepted methodology for positing relationship
>> among languages was morphological typology--the famous "inflecting" versus
>> "agglutinating" versus "isolating" spectrum.  However, it was shown *during the
>> nineteenth century* that this is *not* a reasonable methodology.
>> 
>
>
>Would you please explain the basis for the original confidence in
>the methodology and then the later loss of confidence ? What
>makes it unreasonable as a measure.
>
The morphological and syntactic typology of a language can change over
time (e.g. from highly inflected, loosely SOV Old English to fairly
isolating, strongly SVO Modern English).  Therefore it's not unusual to
find considerable typological variation among demonstrably related
languages:  isolating Chinese and agglutinative Tibeto-Burman (both
Sino-Tibetan), isolating Khmer and Vietnamese and agglutinative Munda
(both Austroasiatic), isolating (Modern) English, inflecting Russian,
and agglutinative Bengali (all Indo-European).  Since we don't find
a reliable correlation between typlogical profile and genetic
relationship in cases where we have independent knowledge of
relationship, there's obviously no reason to trust it when we don't.

Scott DeLancey                     delancey@darkwing.uoregon.edu
Department of Linguistics
University of Oregon