In defense of etymological speculation

[Gerald Cohen]

    I've been doing etymological research for 30+ years and regard
speculation as appropriate when wrestling with a refractory problem.
When detectives approach a crime whose solution isn't readily
obvious, they engage in various speculation, most of which is no
doubt incorrect. When investigators deal with a mysterious
airplane-crash, they set up various hypotheses and then check their
plausibility.

    Etymological research is sometimes similar.
Errors/hypotheses/speculations of various sorts are an integral part
of this research. Of course it's nice to find the earliest
attestation of a term (e.g. for "shyster"--1843) and a direct
explanation of how a term originated (e.g., for cant "wire" = the
pickpocket who actually puts his hand in the victim's pocket or
purse). Those are cause for celebration.

     But before the conclusive evidence is found (if ever), mistakes
are often inevitable. The eminent scholar Roman Jakobson once
commented when speaking of research in general: "A bad idea leads to
a better idea; the absence of an idea leads to nothing."

     Now to "bulldyke" possibly arising from "bulldog-like." This is
only a possibility, clearly presented in a vein of speculating. We
know that bulldogs are aggressive, we know that
bulldykes/dykes/bulldaggers/etc. are described as aggressive. Maybe,
just maybe, the word "bulldog" is hidden in "bulldyke."
This suggestion may turn out to be wrong, but it isn't totally farfetched.

     And by my putting the suggestion in the public domain, interested
ads-l members may be motivated to keep an open eye for confirming or
refuting information. Maybe this search will correctly lead to a
totally new suggestion. Meanwhile, dictionaries can with complete
accuracy continue to write "Origin unknown" until an etymology with
good supporting evidence emerges.

Gerald Cohen