LINGUIST List 2.97

Wednesday, 27 Mar 1991

Disc: Fonts, IT, Lg families, Falsifiability

Editor for this issue: <>

Directory

1. , WordPerfect/LaserJet fonts
2. , IT (interliner processing software)
3. , Language Families
4. Herb Stahlke, RE: Language Families; Conferences
5. , Popper's View on Falsifiability

Does anybody know how to get phonetic symbols, and in general how
to generate special characters, in WordPerfect that would then
print on HP LaserJet (and hopefully that would appear on the screen)?

Is there anyone out there who has had experience in using this. It was de-
veloped by the Summer Institute of Linguistics and I have MS DOS version 1.1c.
Somehow I can't get the program to accept my data. I'd be happy to send more
details if you're willing to hear me out.

I am gratified by the response to my postings about hypothetical
language families, much of it in the form of messages directly
to m
y email address, but raising more general issues. Actually, the
one issue that has been raised so far, but by several people, is
the propriety of working on, say, Nostratic when Afroasiatic is
far from done, or on Sino-Caucasian when Sino-Tibetan remains
murky. This is the first substantive issue regarding this kind
of work that I have seen raised, so I would like to address it
briefly here.

The answer is simple. Indo-European was not reconstructed on
the basis of prior reconstructions of its various branches.
Indeed, to this day, we do not know what the correct branching
is for IE, for the most part. Balto-Slavic remains, for example,
a controversial hypothesis, and there is certainly no consensus
on what 2 (or 3) languages PIE broke up into when it first broke
up. Likewise, there is little consensus on the subclassification
of the much smaller Uto-Aztecan family, yet no one has ever said
that we would first have to settle those issues before postulating a
PUA language or even reconstructing it. Likewise, on a smaller scale,
we do not doubt the reality of English and the common descent of all
English dialects, yet certainly do not have a completely clear
cladogram (branching diagram) of said dialects. And, indeed, the
process of constructing these intermediate nodes can, in most cases,
only be started once the reality of the highest node is granted.
By the same token, incidentally, there is considerable unclarity
about how to classify the Nostratic branches. While there has
been some talk about a Proto-East-Nostratic, it is all very vague
and I think premature. 

Nor is there any great philosophy in all this, I suspect. It just
a result of accidents of history that some branchings leave much
clearer evidence of having occurred than others. And the clarity
of the evidence has nothing (or little) to do with the age of the
branchings.

 As an Africanist who got started during the Greenberg-Africanist wars
of the sixties, I'm puzzled by the complete absence of reference to G's
African linguistic classification. The arguments then and now parallel each
other closely. Twenty-five years after the publication of _The Languages of
Africa_ his work stands up well. The four major families--Afro-Asiatic,
Nilo-Saharan, Niger-Kordofanian, and Khoi-San--have not been broken up. There
has been some interesting reclassification within them, such as Bennett and
Sterk's work on South Central Niger-Congo and Bender's proposed Omotic branch
of Afro-Asiatic, but this is work that was facilitated by Greenberg's overall
classification. Not being an Americanist, I won't try to evaluate his work on
this hemisphere, but his track record is certainly impressive. I also have
seen very little discussion of his methodological contributions, just blunt
rejection of the methods. Clearly what he does is not traditional comparative
reconstruction and should not be evaluated as such. If anything, mass
comparison looks more like an archeological method than a linguistic method, if
such an analogy helps. The bottom line is that his methods worked in Africa.
I would be surprised to find that they did not work elsewhere.

Herb Stahlke
Ball State University

I was asked for a specific reference to Popper's own admission that
falsifiability does not ultimately work as a criterion for scientificity.
I refer to the index to his book, The logic of scientific discovery,
under Falsifiability, never final or conclusive, which refers us to
pp. 42 and 50. On p. 42 he says "...it is always possible to find
some way of evading falsification. ... It is even possible without
logical inconsistency to adopt the position of simply refusing to
acknowledge any falsifying experience whatsover. Admittedly,
scientists do not usually proceed in this way, but logically such
a procedure is possible." To be sure, he then proceeds to write
a whole book about how scientists DO proceed, but he never succeeds
in giving a water-tight definition, to my way of thinking. And even
if he did, it would not, given the passages cited, be merely a definition
in terms of falsifiability, but rather in terms of a specific empirical
method, as he calls it. So, my point is two-fold. To the extent that
he succeeds, Popper is not defining science MERELY in terms of
falsifiability. And, it is not at all clear that he ever manages
to define precisely what it is that scientists do NOT do, even though
they are logically entitled to do it. 
Specifically, on p. 82, Popper admits he has no definite definition
of the kinds of stratagems (as he calls them) that scientists must
avoid if falsification of false theories is to be possible. "The list
[he earlier gave such a list] makes no claim to completeness: it
must be left to the investigator ... to guard constantly against the
temptation to employ new ... stratagems...