Comparative Methodology

[Rich Alderson]

----------------------------Original message----------------------------
> Can you give an exact reference to where Bloomfield says this?

Actually, I can't.  I looked at the chapter on comparative method in _Language_
and at the introductory material in "Algonquian" (in _Linguistic Structures of
Native America_), and found nothing particular about the *number* of languages
used.

I had originally run across discussion of this in Mary Haas' _The Prehistory of
Languages_.  She did not explicitly ascribe a preference to Bloomfield.

As I recall, Isidore Dyen *did* explicitly ascribe this to Bloomfield in his
graduate course on comparative method at Yale.  But I don't recall whether he
ever made the statement in writing, either.

This may simply be part of the folklore of the discipline, as it turns out.

> there may be situations where only binary comparison is possible

Needs must, but the preference is of course for more data, as you note.

                                                                Rich Alderson