Eats, Shoots & Leaves: The Zero Tolerance Approach to Punctuation

That’ll Do, Comma

When the humorist James Thurber was writing for New Yorker editor Harold Ross in the 1930s and1940s, the two men often had very strong words about commas. It is pleasant to picture the scene: two hard-drinking alpha males in serious trilbies smacking a big desk and barking at each other over the niceties of punctuation. According to Thurber’s account of the matter (in The Years with Ross [1959]), Ross’s “clarification complex” tended to run some-what to the extreme: he seemed to believe there was no limit to the amount of clarification you could achieve if you just kept adding commas. Thurber, by self-appointed virtuous contrast, saw commas as so many upturned office chairs unhelpfully hurled down the wide-open corridor of readability. And so they endlessly disagreed. If Ross were to write “red, white, and blue” with the maximum number of commas, Thurber would defiantly state a preference for “red white and blue” with none at all, on the provocative grounds that “all those commas make the flag seem rained on. They give it a furled look.”

If you want to know about editorial “commaphilia” as a source of chronic antagonism, read The Years with Ross. Thurber once went so far as to send Ross a few typed lines of one of Wordsworth’s Lucy poems, repunctuated in New Yorker style:

But Ross, it seems, was unmoved by sarcasm, and in the end Thurber simply had to resign himself to Ross’s way of thinking. After all, he was the boss; he signed the cheques; and of course he was a brilliant editor, who endearingly admitted once in a letter to H. L. Mencken, “We have carried editing to a very high degree of fussiness here, probably to a point to get it under control.” And so the comma proliferated. Thurber was once asked by a correspondent: “Why did you have a comma in the sentence, ‘After dinner, the men went into the living-room’?” And his answer was probably one of the loveliest things ever said about punctuation. “This particular comma,” Thurber explained, “was Ross’s way of giving the men time to push back their chairs and stand up.”

Why the problem? Why the scope for such differences of opinion? Aren’t there rules for the comma, just as there are rules for the apostrophe? Well, yes; but you will be entertained to discover that there is a significant complication in the case of the comma. More than any other mark, the comma draws our attention to the mixed origins of modern punctuation, and its consequent mingling of two quite distinct functions:

1 To illuminate the grammar of a sentence

2 To point up – rather in the manner of musical notation – such literary qualities as rhythm, direction, pitch, tone and flow

This is why grown men have knock-down fights over the comma in editorial offices: because these two roles of punctuation sometimes collide head-on – indeed, where the comma is concerned, they do it all the time. In 1582, Richard Mulcaster’s The First Part of the Elementarie (an early English grammar) describedthe comma as “a small crooked point, which in writing followeth some small branch of the sentence, & in reading warneth vs to rest there, & to help our breth a little”. Many subsequent grammars of the 17th, 18th and 19th centuries make the same distinction. When Ross and Thurber were threatening each other with ashtrays over the correct way to render the star-spangled banner, they were reflecting a deep dichotomy in punctuation that had been around and niggling people for over four hundred years. On the page, punctuation performs its grammatical function, but in the mind of the reader it does more than that. It tells the reader how to hum the tune.

If only we hadn’t started reading quietly to our-selves. Things were so simple at the start, before grammar came along and ruined things. The earliest known punctuation – credited to Aristophanes of Byzantium (librarian at Alexandria) around 200 bc – was a three-part system of dramatic notation (involving single points at different heights on the line) advising actors when to breathe in preparation for a long bit, or a not-so-long bit, or a relatively short bit. And that’s all there was to it. A comma, at that time, was the name of the relatively short bit (the word means in Greek “a piece cut off”); and in fact when the word “comma” was adopted into English in the 16th century, it still referred to a discrete, separable group of words rather than the friendly little tadpoley number-nine dot-with-a-tail that today we know and love. For a millennium and a half, punctuation’s purpose was to guide actors, chanters and readers-aloud through stretches of manuscript, indicating the pauses, accentuating matters of sense and sound, and leaving syntax mostly to look after itself. St Jerome, who translated the Bible in the 4th century, introduced a system of punctuation of religious texts per cola et commata (“by phrases”), to aid accurate pausing when reading aloud. Cassiodorus, writing in the 6th century in southern Italy for the guidance of trainee scribes, included punctuation in his Institutiones Divinarum et Saecularium Litterarum, recommending “clearpausing in well-regulated delivery”. I do hope Harold Pinter knows about all this, by the way; who would have thought the pause had such a long and significant history?

Most of the marks used by those earnest scribes look bizarre to us now, of course: the positura, a mark like a number 7, which indicated the end of a piece of text; the sinister mark like the little gallows in a game of hangman that indicated the start of a para-graph (paragraphs weren’t indented until much later); and, significantly here, the virgula suspensiva, which looked like our present-day solidus or forward slash (/), and was used to mark the briefest pause or hesitation. Perhaps the key thing one needs to realise about the early history of punctuation is that, in a literary culture based entirely on the slavish copying of venerated texts, it would be highly presumptuous of a mere scribe to insert helpful marks where he thought they ought to go. Punctuation developed slowly and cautiously not because it wasn’t considered important, but, on the contrary, because it was such intensely powerful ju-ju. Pause in the wrong place and the sense of a religious text can alter in significant ways. For example, as Cecil Hartley pointed out in his 1818 Principles of Punctuation: or, The Art of Pointing, consider the difference between the following:

and:

Now, huge doctrinal differences hang on the placing of this comma. The first version, which is how Protestants interpret the passage (Luke, xxiii, 43), lightly skips over the whole unpleasant business of Purgatory and takes the crucified thief straight to heaven with Our Lord. The second promises Paradise at some later date (to be con-firmed, as it were) and leaves Purgatory nicely in the picture for the Catholics, who believe in it. Similarly, it is argued that the Authorised Version of the Bible (and by extension Handel’s Messiah) misleads on the true interpretation of Isaiah xl, 3. Again, consider the difference:

and:

Also:

and

Of course, if Hebrew or any of the other ancient languages had included punctuation (in the case of Hebrew, a few vowels might have been nice as well), two thousand years of scriptural exegesis need never have occurred, and a lot of clever, dandruffy people could definitely have spent more time in the fresh air. But there was no punctuation in those ancient texts and that’s all there is to it. For a considerable period in Latin transcriptions there were no gaps between words either, if you can credit such madness. Texts from that benighted classical period – just capital letters in big square blocks – look to modern eyes like those word-search puzzles that you stare at for twenty minutes or so, and then (with a delighted cry) suddenly spot the word “PAPERNAPKIN” spelled diagonally and back-wards. However, the scriptio continua system (as it was called) had its defenders at the time. One fifth-century recluse called Cassian argued that if a text was slow to offer up its meaning, this encouraged not only healthy meditation but the glorification of God – the heart lifting in praise, obviously, at the moment when the word “PAPERNAPKIN” suddenly floated to the surface, like a synaptic miracle.

Isn’t this history interesting? Well, I think so – even though, for a considerable time, admittedly, not much happened. That imaginative chap Charlemagne (forward-looking Holy Roman Emperor) stirred things up in the 9th century when Alcuin of York came up with a system of positurae at the ends of sentences (including one of the earliest question marks), but to be honest western systems of punctuation were damned unsatisfactory for the next five hundred years until one man – one fabulous Venetian printer – finally wrestled with the issue and pinned it to the mat. That man was Aldus Manutius the Elder (1450–1515) and I will happily admit I hadn’t heard of him until about a year ago, but am now absolutely kicking myself that I never volunteered to have his babies.

The heroic status of Aldus Manutius the Elder among historians of the printed word cannot be overstated. Who invented the italic typeface? Aldus Manutius! Who printed the first semicolon? Aldus Manutius! The rise of printing in the 14th and 15th centuries meant that a standard system of punctuation was urgently required, and Aldus Manutius was the man to do it. In Pause and Effect (1992), Malcolm Parkes’s magisterial account of the history of punctuation in the West, facsimile examples of Aldus’s groundbreaking work include a page from Pietro Bembo’s De Aetna (1494) which features not only a very elegant roman typeface but the actual first semicolon (and believe me, this is exciting). Of course we did not get our modern system overnight, but Aldus Manutius and his grandson (conveniently of the same name) are generally credited with developing several of our modern conventional signs. They lowered the virgule and curved it, for a start, so that it began to look like the modern comma. They put colons and full stops at the ends of sentences. Like this. And also – less comfortably to the modern eye – like this:

Most significantly of all, however, they ignored the old marks that had aided the reader-aloud. Books were now for reading and understanding, not intoning. Moving your lips was becoming a no-no. Within the seventy years it took for Aldus Manutius the Elder to be replaced by Aldus Manutius the Younger, things changed so drastically that in 1566 Aldus Manutius the Younger was able to state that the main object of punctuation was the clarification of syntax. Forget all that stuff about the spiritual value to the reader of working out the meaning for himself; forget as well the humility of those copyists of old. I’m sure people did question whether Italian printers were quite the right people to legislate on the meaning of every-thing; but on the other hand, resistance was obviously useless against a family that could invent italics.

So what happened to the comma in this process? Well, between the 16th century and the present day, it became a kind of scary grammatical sheepdog. As we shall shortly see, the comma has so many jobs as a “separator” (punctuation marks are traditionally either “separators” or “terminators”) that it tears about on the hillside of language, endlessly organising words into sensible groups and making them stay put: sorting and dividing; circling and herding; and of course darting off with a peremptory “woof” to round up any wayward subordinate clause that makes a futile bolt for semantic freedom. Commas, if you don’t whistle at them to calm down, are unstoppably enthusiastic at this job. Luckily the trend in the 20th century (starting with H. W. Fowler’s The King’s English in 1906) has been towards ever-simpler punctuation, with fewer and fewer commas; but take any passage from a non-contemporary writer and you can’t help seeing the constituent words as so many defeated sheep that have been successfully corralled with the gate slammed shut by good old Comma the Sheepdog.

No wonder feelings run high about the comma. When it comes to improving the clarity of a sentence, you can nearly always argue that one should go in; you can nearly always argue that one should come out. Stylists have meanwhile always dickered with the rules: Oscar Wilde famously spent all day on a completed poem, dangling a questionable comma over it; Gertrude Stein called the comma “servile” and refused to have anything to do with it; Peter Carey cleverly won the Booker Prize in 2001 for a book that contained no commas at all (True History of the Kelly Gang); and I have seen an essay on the internet seriously accusing John Updike, that wicked man, of bending the rules of the comma to his own ends “with fragments, comma splices, coordinate clauses without commas, ellipted co-ordinate clauses with commas, and more” – charges to which, of course, those of us with no idea what an ellipted-coordinate-clause-with-a-comma might look like can only comment, “Tsk”.

Meanwhile, lawyers eschew the comma as far as possible, regarding it as a troublemaker; and readers grow so accustomed to the dwindling incidence of commas in public places that when signs go up saying “No dogs please”, only one person in a thousand bothers to point out that actually, as a statement, “no dogs please” is an indefensible generalisation, since many dogs do please, as a matter of fact; they rather make a point of it.

“The use of commas cannot be learned by rule.” Such was the opinion of the great Sir Ernest Gowers; and I have to say I find that a comfort, coming from the grand old boy himself. However, rules certainly exist for the comma and we may as well examine some of them. The fun of commas is of course the semantic havoc they can create when either wrongly inserted (“What is this thing called, love?”) or carelessly omitted (“He shot himself as a child”).[1] A friend of mine who runs a Shakespeare reading group in New England tells a delightful story of a chap playing Duncan in Macbeth who listened with appropriate pity and concern while the wounded soldier in Act I gave his account of the battle, and then cheerfully called out, “Go get him, surgeons!” (It’s supposed to be “Go, get him surgeons.”)

But we’ll come to such lovely enjoyable things by and by. In the meantime, however, this is serious. Sharpen a pencil, line up your favourite stimulants, furrow the brow, and attempt to concentrate on the following.

This is probably the first thing you ever learn about commas, that they divide items in lists, but are not required before the and on the end:

The rule here is that the comma is correct if it can be replaced by the word and or or. For example: “I had a marvellous time eating in tavernas and swimming in the turquoise water and getting sloshed on retsina and not sending postcards.” This would be the grammatical consequence of omitting the comma: a sentence that is clumsy (and sounds a lot more sloshed), but still counts as grammatical. What a loss to the language it was, incidentally, when they changed the name of Opal Fruits to Starburst.

However, if you feel you are safe paddling in these sparklingly clear shallows of comma usage, think again. See that comma-shaped shark fin ominously slicing through the waves in this direction? Hear that staccato cello? Well, start waving and yelling, because it is the so-called Oxford comma (also known as the serial comma) and it is a lot more dangerous than its exclusive, ivory-tower moniker might suggest. There are people who embrace the Oxford comma and people who don’t, and I’ll just say this: never get between these people when drink has been taken. Oh, the Oxford comma. Here, in case you don’t know what it is yet, is the perennial example, as espoused by Harold Ross: “The flag is red, white, and blue.”

So what do you think of it? (It’s the comma after “white”.) Are you for it or against it? Do you hover in between? In Britain, where standard usage is to leave it out, there are those who put it in – including, interestingly, Fowler’s Modern English Usage. In America, conversely, where standard usage is to leave it in, there are those who make a point of removing it (especially journalists). British grammarians will concede that sometimes the extra comma prevents confusion, as when there are other ands in the vicinity:

But this isn’t much of a concession, when you think about it. My own feeling is that one shouldn’t be too rigid about the Oxford comma. Sometimes the sentence is improved by including it; sometimes it isn’t. For example, in the introduction to this book (page 7) I allude to punctuation marks as the traffic signals of language: “they tell us to slow down, notice this, take a detour, and stop”. And, well, I argued for that Oxford comma. It seemed to me that without the comma after “detour”, this was a list of three instructions (the last a double one), not four. And here was a case where the stylistic reasons for its inclusion clearly outweighed the grammatical ones for taking it out. This was a decelerating sentence. The commas were incrementally applying the brakes. To omit the comma after “detour” would have the sentence suddenly coasting at speed again instead of slowing to the final halt.

Anyway, there are some more points about commas in lists before we move on. In a list of adjectives, again the rule is that you use a comma where an and would be appropriate – where the modifying words are all modifying the same thing to the same degree:

But you do NOT use a comma for:

This is because, in each of these cases, the adjectives do their jobs in joyful combination; they are not intended as a list. The rhino isn’t endangered and white. The wines aren’t Australian and red. The Duke of York wasn’t grand and old. The wedding wasn’t big and fat and Greek.

Commas are used when two complete sentences are joined together, using such conjunctions as and, or, but, while and yet:

If this seems a bit obvious to you, I apologise. But trouble arises with this joining-comma rule from two directions: when stylists deliberately omit the conjunction and just keep the comma where a semi-colon is called for (this is the “splice comma” John Updike is accused of ), and when the wrong joining words are used. The splice comma first.

Now, so many highly respected writers adopt the splice comma that a rather unfair rule emerges on this one: only do it if you’re famous. Samuel Beckett spliced his way merrily through such novels as Molloy and Malone Dies, thumbing his nose at the semicolon all the way: “There I am then, he leaves me, he’s in a hurry.” But then Beckett was not only a genius, he was a man who wrote in French when he didn’t have to; we can surely agree he earned the right to be ungrammatical if he felt like it. Besides, he is not alone. E. M. Forster did it; Somerset Maugham did it; the list is endless. Done knowingly by an established writer, the comma splice is effective, poetic, dashing. Done equally knowingly by people who are not published writers, it can look weak or presumptuous. Done ignorantly by ignorant people, it is awful.

Meanwhile, words that must not be used to join two sentences together with a comma are however and nevertheless, as in, “It was the Queen’s birthday on Saturday, nevertheless, she had no post what-ever”; “Jim woke up in his own bed, however, he felt great.” Again, the requirement is for either a new sentence or one of those unpopular semicolons.

Are we halfway yet? I hope so, but I doubt it. Anyway, this one is quite simple, involving missing words cunningly implied by a comma:

This doesn’t arise very much these days, though, does it? I wonder why?

This usage is likely to lapse. Many writers prefer to use colons; others just open the inverted commas – a pretty unambiguous sign that direct speech is coming. Personally, I seem to ring the changes. Since this is a genuine old pause-for-breath use of the comma, however, it would be a shame to see it go.

This is where comma usage all starts getting tricky. The first rule of bracketing commas is that you use them to mark both ends of a “weak interruption” to a sentence – or a piece of “additional information”. The commas mark the places where the reader can – as it were – place an elegant two-pronged fork and cleanly lift out a section of the sentence,

In all these cases, the bits between the commas can be removed, leaving the sentence arguably less interesting, but grammatically entire.

As with other paired bracketing devices (such as parentheses, dashes and quotation marks), there is actual mental cruelty involved, incidentally, in opening up a pair of commas and then neglecting to deliver the closing one. The reader hears the first shoe drop and then strains in agony to hear the second. In dramatic terms, it’s like putting a gun on the mantelpiece in Act I and then having the heroine drown herself quietly offstage in the bath during the interval. It’s just not cricket. Take the example, “The Highland Terrier is the cutest, and perhaps the best of all dog species.” Sensitive people trained to listen for the second comma (after “best”) find themselves quite stranded by that kind of thing. They feel cheated and giddy. In very bad cases, they fall over.

However, why is it that sometimes these pairs of commas are incorrect? One Telegraph correspondent wrote to complain about a frequent newspaper solecism, and the example he gave was, “The leading stage director, Nicholas Hytner, has been appointed to the Royal National Theatre.” Shouldn’t the commas be removed in cases such as this, he asked? Well, yes. Absolutely. For a start, if you removed the name “Nicholas Hytner” from this particular sentence, it would make no sense at all. But there is a larger grammatical point here, too. Consider the difference between:

and:

In the first case, the reader infers from the absence of commas that not everyone in the queue was fortunate. Some people did not get tickets. (The ones who did were, naturally, cock-a-hoop.) In the second version everyone in the queue gets tickets, hurrah, and I just hope it turned out to be for some-thing nice. The issue here is whether the bit between the commas is “defining” or not. If the clause is “defining”, you don’t need to present it with a pair of commas. Thus:

If the information in the clause is “non-defining”, however, then you do:

Now, here’s a funny thing. When the interruption to the sentence comes at the beginning or at the end, the grammatical rule of commas-in-pairs still applies, even if you can only see one of them. Thus:

is, from the grammatical point of view, the same as:

as well as:

In many cases nowadays, the commas bracketing so-called weak interruptions are becoming optional. And I say three cheers for that, quite frankly. Where I get into a tangle with copy-editors is with sentences such as:

This is, actually, all right. True, it isn’t elegant, but it uses the comma grammatically as a “joining” comma, before the “and”. Most editors, however, turn purple at the sight of such a sentence. It becomes, suddenly:

It seems to me that there are two proper uses of the comma in conflict here, and that the problem arises simply from the laudable instinct in both the writer and the editor to choose just one use at a time. In previous centuries – as we can see in those examples from Fielding and Dickens – every single use of the comma would be observed:

Nowadays the fashion is against grammatical fussiness. A passage peppered with commas – which in the past would have indicated painstaking and authoritative editorial attention – smacks simply of no backbone. People who put in all the commas betray themselves as moral weaklings with empty lives and out-of-date reference books. Back at The New Yorker, Thurber tells the story of “the grisonanecdote” – a story about a soap salesman who belatedly spots a grison (a South American weasel-like carnivore) on a porch in New Jersey. Now, Thurber says he commanded Ross not to change a word of this piece, but he was obviously asking for trouble. “It preserves the fine texture of the most delicate skin and lends a lasting and radiant rosiness to the complexion my God what is that thing?” says the salesman. Ross, of course, inserted a comma after “my God”. He just couldn’t help himself.

The big final rule for the comma is one that you won’t find in any books by grammarians. It is quite easy to remember, however. The rule is: don’t use commas like a stupid person. I mean it. More than any other mark, the comma requires the writer to use intelligent discretion and to be simply alert to potential ambiguity. For example:

1 Leonora walked on her head, a little higher than usual.

2 The driver managed to escape from the vehicle before it sank and swam to the river-bank.

3 Don’t guess, use a timer or watch.

4 The convict said the judge is mad.

In the first example, of course, the comma has been misplaced and belongs after “on”. The second example suggests that the vehicle swam to the river-bank, rather than the passenger. It requires a comma after “sank”. The third is pretty interesting, since it actually conveys the opposite of its intended meaning. What it appears to say is, “Don’t guess, or use a timer or a watch”, when in fact it only wants to tell you not to guess. It therefore requires a semi-colon or even a full stop after “guess”, rather than a comma. The fourth makes perfect sense, of course – unless what’s intended is: “The convict, said the judge, is mad.”

Two particular stupid uses of the comma are pro-liferating and need to be noted. One is the comma memorably described in the “This English” column of the New Statesman in the late 1970s as “the yob’s comma”: “The yob’s comma, of course, has no syntactical value: it is the equivalent of a fuddled gasp for breath, as the poor writer marshals his battered thoughts.” Examples cited in the New Statesman included this, from The Guardian:

The comma after “animals” is not only ungrammatical and intrusive, but throws the end of the sentence (“on legal advice”) into complete semantic chaos. Meanwhile, moronic sentences such as “Parents, are being urged to take advantage of a scheme designed to prevent children getting lost in super-markets” and “What was different back then, was if you disagreed with the wrong group, you could end up with no head!” are observably on the increase.

Less stoppable is the drift towards American telegraphese in news headlines, where the comma is increasingly given the job of replacing the word “and”. Thus:

So that’s nearly it for the comma. Although it is not true that the legal profession has historically eschewed commas altogether, one begins to realise there is a sensible reason for its traditional wariness. It is sometimes said, for instance, that Sir Roger Casement (1864–1916), the Irish would-be insurrectionist, was actually “hanged on a comma”, which you have to admit sounds like a bit of very rough justice, though jolly intriguing. How do you get hanged on a comma, exactly? Doesn’t the rope keep slipping off ? Well, having landed in Ireland in 1916 from a German submarine, Casement was arrested and charged under the Treason Act of 1351, where-upon his defence counsel opted to argue a point of punctuation – which is the last refuge of the scoundrel, of course; but never mind, you can’t blame the chap, it must have seemed worth a go. His point was that the Treason Act was not only written in Norman French but was unpunctuated, and was thus open to interpretation. The contested words in question, translated literally, were:

Casement’s defence argued that, since Casement had not been adherent to the king’s enemies “in the realm” (indeed, on the contrary, had scrupulously conducted all his treasonous plotting abroad), he was not guilty. Now, I guarantee you can look at this set of words for hours at a stretch without seeing any virtue in this pathetic contention. Casement was clearly condemned by the phrase “or elsewhere”, regardless of how you punctuate it. However, two judges duly traipsed off to the Public Record Office to examine the original statute and discovered under a microscope a faint but helpful virgule after the second “realm” which apparently (don’t ask) cleared up the whole thing. Mr Justice Darling ruled that “giving aid and comfort to the king’s enemies” were words of apposition:

How this story ever got the sensational name “hanged on a comma”, however, is an interesting matter. “Tried to get off on a comma” is a more accurate representation of the truth.

A similar comma dispute still rages today, in a case with less explosive overtones. On his deathbed in April 1991, Graham Greene corrected and signed a typed document which restricts access to his papers at Georgetown University. Or does it? The document, before correction, stated:

Being a chap who had corrected proofs all his life, Greene automatically added a comma after “excluding any other” and died the next day without explaining what he intended by it. And a great ambiguity was thereby created. Are all other researchers excluded from quoting the material? Or only other biographers? The librarian at George-town interprets the document to mean that nobody besides Norman Sherry can consult the material at all. Meanwhile others, including Greene’s son, argue that the comma was carefully inserted by Greene only to indicate that Sherry was the sole authorised biographer. It is worth pointing out here, by the way, that legal English, with its hifalutin efforts to cover everything, nearly always ends up leaving itself semantically wide open like this, and that if Greene had been allowed to write either “Let Norman Sherry see the stuff and no one else” or, “Don’t let other biographers quote from it, but otherwise all are welcome”, none of this ridiculous palaver would have transpired.