Three items concerning major North American Dictionary Projects:

The online Middle English Dictionary remains an active project under the custodianship of the University of Michigan Library, where, however, it has to compete for time and attention with other text and software-development projects, including the enlargement of EEBO-TCP (Early English Books Online). Its helpline ( is monitored, its platform remains in development (with date-sorting next on the to-do list), a small team of editors continues to add and correct the text almost daily, and re-loads of the online data should be expected on a roughly annual basis.  Compared with the original e-MED, the present revival effort, now entering its fourth year, has added about 1,600 bibliographic ‘stencils,’ 2,500 dictionary entries, and 16,000 quotations.

Though the project has a dauntingly long list of improvements to work on, it is concentrating this year (2020) on six areas: (1) examining and entering information from the last 1,000 or so ‘Supplement slips’ (produced by editors of the print MED over decades); (2) correcting all mistakes reported by users; (3) adding newly appearing and some previously overlooked texts to the Bibliography (reading them at least minimally for new dates, senses, and words); (4) maintaining and improving a map of MED headwords to OED headwords; (5) making MED data available for projects that could benefit from it; and (6) collaborating with active editorial and scholarly efforts: soliciting pre-publication editions and transcripts, and if possible adding entries and quotations to the MED even before the edition itself appears. Under this last cooperative category thanks are due to Anthony Esposito and the rest of the OED staff, who have acted towards MED in a truly collegial spirit; and particularly to Tess Tavormina and Ruth Harvey of the Henry Daniel Project, who not only supplied a pre-publication copy of Daniel’s seminal treatise on uroscopy, but have been willing to share lexical cruces with MED for joint consideration, even as they are encountered in the course of editing Daniel’s equally massive Herbal.

It is clear to all concerned that in order for MED to thrive, it needs to become a community endeavor, with distributed contributions and contributors properly credited. Volunteer contributions and ideas for collaboration will therefore be welcomed.

Paul Schaffner

For information on the Dictionary of Old English, use this link.

Jost, David. 2020. “Interview with Joan Houston Hall.” Journal of English Linguistics 48(4) with permission from Sage Publishing

DJ: Now, we talked a lot about the procedures of DARE and how it had to struggle through constant changes in technology and also feelings about technology. If DARE was starting today, how would its methods be different?

JHH: I think unless we had a hugely generous sponsor, it would probably not be done with in-person interviews, and I think that’s too bad, because the bonds that developed between field workers and informants were really important to the success of the interviews and the amount of data collected. It’s incredibly time-consuming and therefore incredibly expensive. A new survey would doubtless have to be done digitally. A few years ago, we experimented with some digital fieldwork in the Online Survey of Wisconsin English, and we found some important problems.

We had some funding from NEH to conduct this new fieldwork in Wisconsin, going back to our original communities and also adding some new communities. The idea was to see how the language had changed in one state over the previous fifty years. We tweaked the questionnaire, we deleted all the questions about old-fashioned farming techniques, and we added new questions for things that didn’t even exist fifty years ago. We added questions such as, “What do you call the cardboard device you put around a hot cup of coffee so you don’t burn your hand?” Well, it can be a sleeve, a jacket, a wrap, a collar, and in New York City it can be a zarf (from the Arabic word for the metal framework for a glass cup). And we also asked, “What do you call the raised horizontal barrier on the pavement of a street that’s intended to slow cars down?” It can be a speed bump, speed hump, speed table (some of these vary by how wide they are); it can be a sleeping policeman, a silent policeman; or, one that I found in a Sacramento suburban parking lot: the sign said undulation. I thought that was very funny.

We worked with the UW Survey Center to set up the questionnaire on a website, and we gathered the same information about our respondents: age, gender, ethnicity, education, community type. And we allowed them to answer only the sections of the questionnaire that interested them.

We discovered a few things about online fieldwork that would need to be taken into account if another nationwide survey were done. One is that initial enthusiasm about taking part does not necessarily guarantee follow-through. We had a lot of publicity about the project, we gave talks about it on the radio, and we enlisted librarians in all of the communities we were targeting. We sent them posters and explanations and instructions, and asked them to recruit people in their communities to take part, because they knew who would be good informants.

Well, after a few months when there was very little follow-through, a colleague and I went on a road trip to meet with the librarians. Almost invariably, they would say, “Oh yes, I remember that. That sounded so interesting.” Then they would dig through piles on their desks and find the materials we had sent that had spent the previous months just sitting there. Personal contact is important, not just with the people who answer the questionnaire but with local facilitators as well.

We also discovered how limited the average person’s tolerance is for online questionnaires. The food section is one that people seemed to especially like, but when they got into it and they noticed in the top corner that this was only question 14 of 74, many decided they had had enough. Obviously, we should have shortened that section. Only one person did the entire questionnaire, which included 1764 questions.

The people who did answer parts of the questionnaire were given the opportunity for a phone interview as well as answering the questions, and this was a good way to get samples of conversational speech for analysis and to get recordings of “The Story of Arthur the Rat.”

I neglected to say earlier, when we were talking about fieldwork, that one of the most important things that the field workers were asked to do was to make a tape recording of the informant, including both “The Story of Arthur the Rat” (which was a silly little story that was designed to have all the important contrasts among phonemes in American English) and also about fifteen minutes or more of free conversation about any topic of their choice. And those recordings allowed us to compare the same text (“Arthur the Rat”) for speakers across the whole country and make generalizations about regional pronunciation patterns. But the free conversations provided a treasure trove of oral history for the period 1965-1970.

Well, those recordings have been used in many ways over the decades: for linguistic analysis (by, for example, David Durian, Keelan Evanini, Tim Frazer, Terry Irons, Tom Purnell, Joe Salmons, Cara Shousterman, Erik Thomas, and Walt Wolfram); for introduction of American speech varieties to people around the world (by professors in Belgium, France, Hungary, and Japan, among other places); and for dialect coaching, for actors who want to know how to reproduce the speech of a person in a particular region (see, for example,

I’m very happy to say that recently we were able to make the whole collection of original recordings (I believe there were 1843) freely available to anyone who wants to listen. That wasn’t an easy task because we had promised the informants originally that their names would not be released and so for several years, between about 2013 and 2017, we had a massive project using students and volunteers (and some staff time)—people who listened carefully to each recording and then “bleeped” out any use of the informant’s name or any other names that might betray who the speaker was or somehow be confidential. Those recordings can be found at Now with digital tools to help analyze the recordings, we can compare the speech of people in the same Wisconsin communities fifty years apart and begin to understand how language has changed.