Fowler’s Dictionary of Modern English Usage (4th edition) – by Jeremy Butterfield

“Is that good English?”

Century after century, people ask this and publishers sell books to answer them. A lot of these books are terrible but some are linguistic treasure troves. A lot of them sink without trace but some endure for decades.

This is one of the good ones.

A brief history of Fowler

One of the biggest names in usage guidance over the last 100 years, especially in Britain, has been Fowler.

Henry Fowler published his Dictionary of Modern English Usage in 1926. It contained a lot more learning, insight and modesty than many of today’s usage guides, but it was fundamentally a book of Fowler’s opinions. These opinions, shaped by his lexicographical work, were sometimes wise, sometimes idiosyncratic. His writing style was sometimes witty, sometimes obscure. It was a hugely influential book, which is why its successors still bear the Fowler name.

In 1965, Ernest Gowers edited the second edition. There were cuts and additions and polishes, but it was essentially an update, and the bulk of the book stayed the same.

The third edition came in 1996, edited by Robert Burchfield. This was a near-total rewrite, although there were still plenty of traces of Fowler. It marked a shift in approach towards greater objectivity and more systematic analysis. Burchfield had built up a database of language use far exceeding Fowler’s more hotchpotch collection, and he used it to inform his judgements. That said, he wasn’t shy with his opinions.

Now, in 2015, we have the fourth edition, edited by Jeremy Butterfield. It’s an update of Burchfield’s version, but a pretty substantial one. There are over 250 new entries and many of the others have been partly or mostly rewritten. Butterfield again improves the standard of evidence by using the immense resources of the Oxford English Corpus – a database holding 2.5 billion words of 21st-century usage from all sorts of sources. And while his own opinions still figure, they play a smaller role than Burchfield’s did.

What’s new?

Some of the changes are to make the entries more reader-friendly: Butterfield makes better use of paragraph breaks and subheadings. The language, although mostly still formal, is less stiff and stuffy: where Burchfield had sections on “diachronic” and “synchronic” usage, Butterfield sensibly calls them “historical” and “current”.

Many of the explanations are better. Butterfield is more likely than Burchfield to begin a section with a short statement of the issue and not just launch into the discussion of it, and he’s more likely to state what a list of examples illustrates and not leave it to the reader to figure out.

He also shows more interest in analysing contentious usages, and not just rating their acceptability. For instance, this gem on the logical evolution of literally:

From emphasizing non-metaphors [as in “What punishment has he suffered? Literally none”] literally has bit by bit passed to emphasizing metaphors and figures of speech [as in “with his eyes he literally scoured the corners of the cell”]… This historical development explains how the word has apparently reversed its meaning. In fact it has done no such thing but has been absorbed into the metaphor. Once it is analysed as part of the verbal image and not external to it, the use makes good linguistic sense…

Butterfield often takes a more liberal line than Burchfield on disputed usages, such as anticipate to mean “expect” and disinterested to mean “uninterested”.

In some cases this simply reflects language change between the two editions. For instance, on the prepositional use of due to (“the train was delayed due to bad weather”), Burchfield said: “Opinion remains sharply divided but it begins to look as if this use of due to will form part of the natural language of the 21c, as one more example of a forgotten battle.” Butterfield fulfils the prophecy: “it looks as if this use of due to is now part of the natural language of the 21c”.

The greater liberalism also reflects differences in opinion between the two editors – or differences in willingness to put opinion aside and be led by the evidence.

Pet peeves and linguistic zoology

Still, Butterfield often gives his opinions too. I’m not too concerned about that – the mammoth task of editing a book like this entitles you to mouth off a bit – as long as the opinions are presented as opinions and don’t unduly sway the advice.

I think he stays on the right side of the line. For instance, he notes a “personal distaste” for the use of cohort to mean a single colleague, but he accepts that it’s not really objectionable and is “now here to stay in all varieties of English”. Conversely, while he makes a strong case for the syntactical usefulness of different than, he warns that British audiences might be irked by it.

And he indulges himself in the occasional rant. For instance, he says that overuse of like as a filler word may cause listeners “to ignore the content of the message completely, to assume that the speaker is little short of brain-dead, or, in extreme cases, to wish they had a discreet firearm to hand”. You might find this sort of thing amusing or tiresome, but either way it’s so theatrical that it must be (mostly) tongue-in-cheek.

(Compared with two other hefty usage guides of recent years, my impression is that Butterfield is less opinionated than Bryan Garner but more than Pam Peters.)

He unconditionally rejects some of the more notorious grammar folklore – such as the wrongness of beginning sentences with and or ending them with prepositions – but still urges us to pay some respect to the most famous superstition of them all: don’t split infinitives. While he rejects the idea that split infinitives are ungrammatical, he thinks this idea is commonly enough held to recommend “avoidance where possible”.

This policy, alas, can lead to awkwardness, such as: “on the other hand, always to put the preposition in front of its object will result in a starchy, stilted, formal style”. I suspect more readers would find “always to put” a strained turn of phrase than would object to the grammar of “to always put”.

Looking over the entries for the commonest bugbears, I wonder how he decided whether to recommend shooting them or cowering at their feet. I wonder this whenever I read a usage guide that favours evidence over prejudice. Corpus data tells us about usage, but not about reactions to usage.

In other places, he doesn’t give a clear steer either way. That might dismay some people, but for me it’s a strength.

It’s easy to find usage guides that decree right and wrong answers on every issue. I suspect it’s easy to write them, too. But a serious guide will acknowledge grey areas, and map them so the reader can decide how to navigate. In this spirit, many of Butterfield’s entries on problem words – such as comprise, data, masterful/masterly and try and – sketch the disputed territory without adjudicating. On who/whom, he sets out the debate and explains the traditional distinction, but also makes clear that your use of whom isn’t a matter of objective right and wrong: it depends on how formal you want to be.

This fulfils his promise, in the introduction, to offer “detailed, reasoned guidance” and to present “the facts of disputed or controversial usages… thereby enabling readers to make informed choices for themselves”.

21st-century usage

Among the new entries, there is thoughtful advice on bemused, brainstorm, deceptively and epicentre, on how to use bullet-points, and on language conventions when writing emails.

Other additions I’m not so sure about. For instance, he says that Web and Internet need initial capitals. This agrees with AP style but it’s a practice many of us abandoned years ago. A quick glance at the Glowbe corpus of online usage suggests that in 2012 Web took a cap about a third of the time and Internet about two-thirds (no, I’m not counting spider webs or starts of sentences). The Chicago Manual of Style caps Internet but not web, and the Guardian, Economist and Telegraph use lower case for both.

The oddest of the new entries is on wilfing, a word that seems to have been invented by a market research company in 2007, to the brief amusement of a few journalists, and then abandoned.

And some more of the entries left over from the earlier editions could have been retired: euphuism, Pindaric and Wardour Street English are not terms you come across every decade.

This brings me to the book’s biggest weakness. While it contains so much valuable advice and information, it’s haunted by the ghost of Fowler. The F-word doubtless has brand value, but a book of this calibre can stand on its own merits. The content and even the approach have moved so far from Henry Fowler’s that it seems odd for so many of the entries to still examine usage through the prism of his views. While Butterfield cites plenty of other, more contemporary usage commentators for support and contrast, by far his biggest reference point is a man who died in 1933.

I love the history of usage and usage disputes; I’m less interested in the history of one particular series of usage guides.

This ought to be the last edition of Fowler. The OUP’s successor volume (around 2030?) should drop the name and most of the references to him, and simply be its own book.

On a more trivial note, I wasn’t looking for typos but I can never completely stop proofreading: there are a few. A rogue comma in a sentence illustrating the Oxford comma is unfortunate, and I winced at the mention, in the entry on the passive voice, of “Barrack Obama”.

But, all in all, this is a really good usage guide. While smaller than Merriam-Webster’s Dictionary of English Usage, and so necessarily less encyclopedic, it’s more than 20 years newer, which is a big plus. Butterfield has decisively earned his place on my desk.

Both comments and trackbacks are currently closed.


  • Andriy Karamazov  On July 16, 2015 at 12:34 pm

    I must shamefully admit I have never heard of this book, although I study linguistics.

  • danseovietnam  On July 30, 2015 at 12:29 am

    This policy, alas, can lead to awkwardness, such as: “on the other hand, always to put the preposition in front of its object will result in a starchy, stilted, formal style”. I suspect more readers would find “always to put” a strained turn of phrase than would object to the grammar of “to always put”.

%d bloggers like this: