Language Log

« previous post | next post »

In "Anti-fascist impact" (6/25/2012), I briefly took up the question of where William Strunk might have gotten his 1918 prescription about the use of however:

In the meaning nevertheless, not to come first in its sentence or clause.

The roads were almost impassable. However, we at last succeeded in reaching camp.

The roads were almost impassable. At last, however, we succeeded in reaching camp.

When however comes first, it means in whatever way or to whatever extent.

However you advise him, he will probably do as he thinks best.

However discouraging the prospect, he never lost heart.

As Geoff Pullum and I have pointed out, this principle was not a general feature of the elite usage of Strunk's time, any more than it is of our own: works by Bram Stoker, Joseph Conrad, Jack London, Mark Twain, and Henry James all exhibit violations of Strunk's Rule "In the meaning nevertheless, not to come first in its sentence or clause" ("The Evolution of Disornamentation", 2/21/2005; "However: retraction of a defense of Strunk", 3/26/2009). While recently reading Winston Churchill's 6-volume history of WWII (published between 1948 and 1952), I got the impression that Sir Winston violated Strunk's Rule rather frequently; and since Churchill's first writings were published in the late 1890s, I wondered whether his style might have changed over the half-century in question. So I did a quick count before breakfast, and here's what I came up with. In three early books (The Story of the Malakand Field Force, 1897; The River War, 1899; From London to Ladysmith via Pretoria, 1900), I counted

non-initial		initial
"in whatever way or extent"	"nevertheless"	"in whatever way or extent"	"nevertheless"
17 (6%)	256 (91%)	3 (1%)	6 (2%)

In his writings around 1900, Churchill used however a lot – 282 times in 300,139 words, or about 940 times per million. 97% of these were non-sentence-initial uses, so Strunk's Rule doesn't get very many chances to apply. Still, out of Churchill's nine sentence-initial uses of however, six were instances of the forbidden "nevertheless" meaning. So Strunk's dictum is supported to the degree that sentence-initial however is rare — but violated in that when sentence-initial however does occur, it usually has the expressly forbidden meaning. How about Churchill's works published around 1950? I checked the first (The Gathering Storm, 1948) and last (Triumph and Tragedy, 1952) in the six-volume history, with this result:

non-initial		initial
"in whatever way or extent"	"nevertheless"	"in whatever way or extent"	"nevertheless"
20 (6%)	231 (68%)	0 (0%)	89 (26%)

Here Churchill's style has moved from a mild divergence with Strunk's prescription to an unequivocal statistical contradiction. In his later writing, the proportion of sentence-initial however has increased to about a quarter — and all of these (in the sample I checked) had the meaning that Strunk proscribed in that position. Was this a general trend in formal written English over that period? More research is needed; but it's clear that the pattern is complicated, since (for instance) Mark Twain was already in unequivocal violation, a generation or two before Churchill and Russell, with about two thirds of his uses of however being of the forbidden sentence-initial "nevertheless" type. A similarly transgressive pattern can be found in Henry David Thoreau — a quick scan of Walden (1854), Canoeing in the Wilderness (1857), and Cape Cod (1865) yields:

non-initial		initial
"in whatever way or extent"	"nevertheless"	"in whatever way or extent"	"nevertheless"
21 (23%)	34 (38%)	4 (4%)	31 (34%)

And Thomas Hardy, though later, goes even more strongly anti-Strunkwards — the counts for Jude the Obscure (1895) and Far from the Madding Crowd (1874) are

non-initial		initial
"in whatever way or extent"	"nevertheless"	"in whatever way or extent"	"nevertheless"
10 (8%)	55 (44%%)	5 (4%)	56 (44%)

We can certainly find other writers of the late 19th and early 20th century who rarely used sentence-initial however at all. Thus Bertrand Russell in Problems of Philosophy (1912) and Analysis of Mind (1921) shows this pattern:

non-initial		initial
"in whatever way or extent"	"nevertheless"	"in whatever way or extent"	"nevertheless"
28 (23%)	94 (76%)	2 (2%)	0 (0%)

This is consistent with Strunk — though the most striking fact is perhaps the low rate of sentence-initial however of any type (2%). Russell's History of Western Philosophy (published in 1945) is a bit more like 1900-era Churchill, in that 94% of the instances of however remain non-inititial, but two thirds of the few initial howevers violate Strunk's Rule:

non-initial		initial
"in whatever way or extent"	"nevertheless"	"in whatever way or extent"	"nevertheless"
31 (10%)	317 (85%)	8 (2%)	15 (4%)

And we can certainly find writers today who mostly follow Strunk's Rule. This looks to me like a case of sociolinguistic variation, though in the formal written language rather than in the vernacular, with complex patterns in space, time, and individual identity.  It might repay more extensive study. [Further discussion of the phenomena and the relevant prescriptivist literature can be found in a LL post by Arnold Zwicky, "However, …", 11/1/2006; and in a handout by Zwicky and Kenter, "Avoid vagueness?  The case of sentence-initial linking however", 7/11/2007.] [Note: Strunk's glosses for the various uses of however are not very satisfactory, in my opinion. They're good enough to serve as names for the uses in question; but a serious stylistic analysis needs to unpack the semantics and pragmatics more carefully, a process that Zwicky and Kenter make a good start on.]

July 1, 2012 @ 10:26 am · Filed by Mark Liberman under Peeving

Permalink