I’ve written about off of on this blog before, in
reaction to British complaints about it as a horrid Americanism. In my day job, I’m writing about
it again from different angles, so I was
thrilled to see that some researchers in Helsinki and Stockholm have undertaken much more wide-ranging
and in-depth research about it than has ever been attempted:
Vartiainen, Turo, and Mikko Höglund.
2020. How to make new use of existing resources: tracing the history of
geographical variation of off of. American Speech 95: 408–40.
Their paper, as the title hints at, is very much about
getting around the problems of studying the history of and
variation within the English language, given the impoverished nature of the data we have. There’s lots of English out there, but it’s not always easy to get a
balanced view of it. For example, it’s not enough to know where a work was published, you
need to know where its author was from. For
another example, if all the evidence you have from Sussex is from farmers and
all that you have from Yorkshire is from school teachers, then your regionality
conclusions are going to be tarnished by other contrasts. Sometimes data sets give this info. More often you either have to go hunting for it and/or the information doesn't exist.
Vartiainen and Höglund have come to their conclusions by
triangulating evidence from a number of corpora, each with their own limitations,
but together rather convincing. At one chronological end, they’re using the
Early English Books Online (EEBO, 1470s–1690s) corpus and, at the other end, a corpus
that is updated daily in the present, News on the Web (from which they only use
regional UK news sources). They’ve also included a range of sources for American English.
Off of only really takes off in the 17th century.
(I won’t go into why that’s so interesting because I have to save things for my book!) In the 19th century, prescriptivists start saying how
horrible it is. British prescriptivists have been more damning of it (“vulgarly
superfluous”, “a Cockneyism and incorrect”), but American style guides advise
against it too (“much inferior to off without the preposition”). The authors
suggest that prescriptive attitudes have colo(u)red linguistic description of the
term, and there’s pretty clear evidence of this, I’d say, in a lot of the
British writing about it, where off of is presented as something from America.
Huddleston and Pullum’s (generally excellent) Cambridge Grammar of the
English Language claims off of is only used in AmE.
Vartiainen and Höglund show that this just isn’t true, and
moreover it never was.
Off of originates in England and has consistently
been used there. What’s striking is how regional it’s stayed.
Here are their maps of where it was most used before 1700 and in the 21st
century. It is very much a southern thing.
This gives a big clue about the
presence of off of in AmE:
Importantly, much of the EEBO data predates the Great Puritan Migration to America that took place between 1620 and 1640 (…). Considering that many of the early colonies were founded by people from East Anglia (…), it is likely that they took this form with them. (p. 428)
They go on to cite examples of off of in the Salem
Witchcraft Trials:
…the older generations may have noticed the increased frequency of off of in public texts (a recency effect), while the younger generations may be sensitive to the form’s high frequency in American English when compared to the other varieties of English. (p. 428)
While it’s certainly possible that the off of surge in AmE could affect current BrE, the evidence from the British data is that it has always been used there. If AmE is having an effect,
perhaps it’s just providing a kind of linguistic mirror that makes the form
feel less non-standard to those who are already hearing and/or using it in
their regional Englishes. The authors conclude that:
…when it comes to regional variation, we have seen that off of is frequently attested in so many parts of England that the whole idea of its being a “regional form” should be questioned. Indeed, based on the results of this study it would seem that in many cases the perceptions that British speakers have of their avoidance of off of [as a regional and/or American form] are due to highly entrenched prescriptive attitudes instead of their actual usage patterns, although we have no doubt that the form is rare enough in some regions, particularly in the West and Northwest of England, to genuinely affect acceptability judgments. (pp. 434–5)
There remain problems in making
direct comparisons of English from different times and places. For example, the AmE corpora include no casual
conversation, but the BrE data do. The authors therefore have to be cautious in
comparing rates of usage in the two countries, There is some indication that off
of is far more widespread in AmE than in other Englishes. In the GloWBE
corpus of web-based English (written, but often not as formal as published
English), AmE has 26.2 off of per million words versus 21.5 in Canadian
English and 8.7 in British. (That data set has not seen the same care as their main data sets, though. It may contain false hits, probably
contains duplications and can’t give a regional picture.)
The paper includes research on the variants offen and offa.
I won’t cover them here, but just mention them to say: oh it’s all so complex
and transatlantic.
In all, a fascinating read for someone who’s always thinking about function words
and transatlantic linguistic comparisons. (That’s me!) I thank the authors for it and American
Speech for publishing it.
Related reading
If you're interested in out of, it's covered at the original off of post. You're welcome to leave comments there and keep that conversation going.
Posts
