More information on my too quick typing. Mr. Kensuke Nanjo Associate Professor of Phonetics, St. Andrew's University kindly forwarded to me his message to John Wells, in which message he writes
I've read your October 9 blog and found Michael says "What surprises me is the unanimity with which descriptive dictionaries fail to report this common American form."But the pronunciation /eSu:/ (the stress is put on the second syllable) of "eschew" is recorded as the first variant in Merriam-Webster's Collegiate Dictionary, Eleventh Edition (2003). /ISu:/ is the second. Since I checked this dictionary when I transcribed "eschew", both /eSu:/ and /ISu:/ have also been recorded as the first and second American variants respectively in my forthcoming Taishukan's Genius English-Japanese Dictionary, Fourth Edition.
Mr Wells' responds, after finding the forms in question plus the possible pronunciation /eˈskjuː/:
I wonder what the evidence for this is. To be honest, it looks to me like an uncertain lexicographer covering every possible spelling pronunciation.
I've said before that I don't like Merriam-Webster. And it's unfair because they don't necessarily get things wrong -- I probably just haven't liked the implications of some of their categorisations and organization. (Had I used the plural to keep a true parallel construction the meaning of that sentence would change inappropriately.) But now I can't remember any solid complaints other than the thin etymology and thinner word history I used to find in the collegiate editions. They are often the odd dictionary, reporting forms and pronunciations that are not found in other sources. And I must see that as a virtue. Having moved away from seeing a good dictionary as proof of correctness, and towards it as a source of keen observation, this avant status wins my admiration -- as long as it's not abused.
With a remarkable turnaround time the M-W pronunciation editor, Joshua Guenter, writes to Mr Wells:
While I don't claim to be any more certain than any other lexicographer, I can tell you that the pronunciation of "eschew" as /ɛ'skju:/ in our Collegiate Dictionary is based purely on citations. All of these citations are from the United States. We also have some citations for the pronunciation /ɛ'sju:/, but not enough to warrant its inclusion in any of our dictionaries.
He adds after some exchanges
We have an enormous collection of data on 3x5 slips of paper. I don't know how many of these we have, but it must be in the millions. They date back to the 19th century, though the ones dealing with pronunciation mostly started with Edward Artin in the 1940s. Most of the editors spend time "reading and marking" magazines, newspapers, etc., for data. The pronunciation editors write down examples of speech from TV, radio, speeches, conversation, etc. This is what we base everything in our dictionaries on.
Mr Wells offers a starkly positive endorsement to the people working at M-W:
It confirms the opinion I had already formed: that in the absence of any more recent dedicated pronunciation dictionary for AmE than Kenyon & Knott (1953, now hopelessly outdated) the best source of information on word pronunciation in American English is the Merriam-Webster Collegiate. And it's the only free on-line dictionary that offers you sound clips.
[btw update: Merriam-Webster is not the only free online dictionary that offers sound clips. Bartleby.com, which provides an online American Heritage 4th edition also offers a .wav file for pronunciation.]
I'm embarrassed that I did not take the time to look up M-W's entry for the word. It's going on my resources panel in the sidebar. For a few day's I will highlight it. I'm offering my hand with head bowed. In fact I'm going back to one of my early posts and crossing out one sentence. (If you read my old post please note that I'm not trying to say that the one plural form is incorrect -- rather that the probable reasoning for the form is based on incorrect historical information. And a shining instance of snobbery ends up backfiring on the shaky brandisher.)