The past tense of “dive” exemplifies the quirky differences between American and British English. In Britain, this word is ostensibly “dived,” while Americans increasingly prefer “dove,” the latter likely formed via analogy from drive/drove (Insert joke about America’s auto-centric culture here).
However, it seems that the dived-dove distinction used to be something of a dialect marker in America. E. Bagby Atwood’s 1953 study A Survey of Verb Forms in the Eastern United States found that dove was chiefly Northeastern, used as the dominant form in New England, Upstate New York, and the Northeast corridor as far south as Wilmington, Delaware. In Central Pennsylvania, northeast Maryland, lowland areas of South Carolina, and a few other pockets, dived and dove were relatively interchangeable. Elsewhere in the Eastern US, presumably, dived was typical, along with more unusual forms like div and duv.*
So what’s happened since then? Well, I would hazard to guess that weird, wonderful old dialect forms like “div” and “duv” have waned significantly. But speaking purely from impression, it seems that “dove” has increased exponentially in the sixty years since Atwood’s study. It’s very difficult for me to say this with certainty, however, as we’re discussing an irregular past tense of a rarely used word for which I’ve found few straightforward studies.
So what kind of evidence can be found? Although I’ve often used Google’s NGram Viewer to examine unusual words, I have very little faith in their “American English Corpus.” So I attempted a wholly unscientific test which I call the “first 100 tweets” experiment, whereby one searches for a word on Twitter, looks up the first 100 tweets that use said word, then breaks down the tweets by nationality. After looking up 36 tweets, the ratio was a whopping 17 tweets in the UK using “dived” vs. only 2 tweets from America.
I unfortunately gave up at that point, when it became clear that data (so to speak) was going to be skewed by two important factors. First, the majority of tweets using “dived” involve Association Football (i.e. Soccer), and given the sport’s greater popularity in Britain than North America, it would likely distort things; it’s reasonable to think that, given that “dive” is a word with a particular meaning in the sport, American soccer fans might use “dived” as well.
Second, “dived” is almost certainly used by Americans in compounds such as “scuba dived,” “sky dived,” and “nose dived,” (as opposed to the bizarre-sounding “scuba dove,” “sky dove,” and “nose dove”), which again, would almost certainly make American “dived” usage difficult to assess.
So although the repeated assumption is that “dove” is the unmarked form in America, I can’t be 100% certain there aren’t still Americans who use “dived.” Yet if “dove” is now widespread in the US, why did “dived” so quickly recede? And why did this not occur in Britain?
*An important caveat: I have not read the study itself (it’s old and unavailable online except for the odd used copy sold on Amazon), only its description in R.L. Trask’s Historical Linguistics.
…just to throw a spanner in the works, I’m fairly sure I – English-English speaker – use ‘dove’. (It’s hard to tell when you’re thinking about it!) I suspect the anomalous strong-verb form is Germanic influence in my case, either a northern-England thing or more directly as I have German-speaking family.
Agreed that it isn’t unusual in the UK, though definitely one of those areas some priggish grammar nazis enjoy picking up on, so not quite standard either.
Have there been any studies about irregular vs. regular verbs and perceived “properness?” Irregular verbs as a whole are declining though…
I entered the phrase “he dove” into the NGram Viewer and it seems to have first appeared in about the 1830s. While it doesn’t say exactly where it arose, that’s certainly long before American car culture. Although the “drive/drove” derivation in America could have come from driving cattle, perhaps?
By the way, have you noticed a considerable flaw in this NGram Viewer? From the late 1600s through about 1800 it was common for the sibilant letter s to be printed very similar to the letter f, and since Google is using optical character recognition to produce these graphs, the results can be quite skewed. To take a mild example, enter the word “muft” into this tool and there’s a big spike throughout that period, when obviously the actual word is “must”.
And there’s also a big trough for “Shakespeare” along with a big rise for “Shakefpeare”.
@Tom,
There are actually numerous considerable flaws in the NGram Viewer! Its attempt to extract American and British corpora is largely a failure, it has exhibited numerous issues with publication dates (although these have been rectified somewhat over the past year or so), and as you mention, it has trouble with non-standard orthography. I still find it useful for looking at broad trends, but I’m wary of attempts to use it for any kind of fine-tuned analysis.
Yes, still fascinating for larger trends!
The tarot cards are one of the oldest forms of divination known.
You might detect a trace of deceit or dishonesty this night.
Fitocracy is totally free and available on Apple iphone and Android.
Feel free to visit my web-site AskNow Reviews (Denny)