I am desperately searching for information about the epenthetic

"r" found in words like "warsh" (wash) and Warshington (Washington).

I know that this phenomena is only found in certain American dialects

(perhaps in Northern California, Oregon, and Idaho), however that is

as much as I have been able to find. I would really appreciate any

anecdotal, as well as quantitative information concerning this

linguistic variation.

Thank you in advance,

Jennifer van Vorst

Portland State


I'm replying to the List with a copy to Zman, because I think ADS people

will be interested in my response. -- DML


This item has come up in lots of discussions, even on ADS-L if I'm not

mistaken, but I'm not in the mood at present to do any digging for sources.

But I did think of Tom Murray's data in THE LANGUAGE OF ST LOUIS,

MISSOURI: VARIATION IN THE GATEWAY CITY (Peter Lang, 1968). Murray cites

interesting local evidence: signs in laundromats (p. 17). In one case, a

"would-be grammarian" corrected a sign saying "warsher don't work" by

changing the verb to "doesn't" with no indication that the correctionist

found anything amiss in the identity of the machine that didn't work.

Murray has a table (p. 18) summarizing data he collected in St Louis:

Table 3


(as in 'wash')

context in which heard

socio-econ class informal midformal formal

% no. % no. % no.

upper 83 1446 68 1185 39 669

middle 92 1657 76 1369 52 937

lower 99 1766 82 1463 68 1213

It is interesting that this phonological bit of Americana, known by

dialectologists as an Ohio Valley phenomenon (but not limited to this

area), is also thought by someone to exist to some extent (primarily?) in

CA, OR, ID. My students from St. Louis would always mention this "awful"

pronunciation when I would ask about features of St Louis speech; they hung

their heads in shame for not being to erase it from their phonological

systems -- only they thought it was a feature of a single lexical item

rather than a phonological rule in their heads/behavior. I've always

suspected it was more widespread than limited dialectological reports have

suggested. When the last volume of DARE comes out we'll probably get a

nice map and a better answer than we had before.