In defense of etymological speculation

[Douglas G. Wilson]

Fred Shapiro:

>Of course, Jerry's explanation ... is also, like Paul's, contradictory to
>the historical evidence I have uncovered.  Any explanation of "bulldyke"
>should address the seeming priority of "bulldyker."

I don't know whether anybody's reading this (maybe my name elicits an
auto-delete) ....

I did exactly and specifically assert priority of "bulldyker" in my recent
note in CoE and in my recent posts here. But note that "bulldyker" looks
like the basic noun even on the basis of what is in HDAS, for example, if
one looks closely. The 15-years-earlier citation from 1906 just
(significantly) strengthens this impression.

I am sure that Fred is not saying that one cannot present a tentative or
speculative etymology, because I have before me Fred's paper in "Verbatim"
26(2):17 (2001) in which he presents the etymology of "chad" = "punched-out
paper fragment" from Scots "chad" = "gravel". This etymology has not a
shred of textual evidence (AFAIK): but it seemed plausible enough for M-W
(ca. 1960), and for Shapiro (ca. 2001), and it seemed (as of 2001)
plausible enough for me too. Of course Fred says "seems to" and "may", as I
would have said also. This tentative position is correct and reasonable
IMHO. Etymology is not chemistry and reproducible experimental data cannot
be demanded. There are superficially absurd etymologies such as Mr.
Chadless and Card Hole Aggregate Debris (both of these now decisively
exploded, I believe) and there are plausible but unsubstantiated
etymologies such as the "gravel" one which can be tested against the
evidence as it trickles in. There may be some disagreement about what is
plausible in some cases.

I think Fred is saying (1) that one should search for evidence for one's
conjectures, and (2) that one should test each conjecture against all
available evidence rather than just tossing it out casually. Is this right?

Hello? Anybody reading?

-- Doug Wilson