Arrant Pedantry


Is the Oxford Comma Ungrammatical?

Few language issues inspire as much fervent debate as the question of whether you need a comma before the last item in a series, also known as the Oxford, Harvard, or serial comma. This is the comma that you sometimes see before and in lists, such as “I need you to go to the store and get bread, milk, and butter.” The Chicago Manual of Style, which is used by many book publishers and some academic journals, requires the comma. The AP Stylebook, which is used by newspapers and many magazines, omits the comma unless it’s necessary to avoid ambiguity.

It seems like such a trifling thing, yet it inspires impassioned debate among editors and writers on both sides of the issue. Last year the satirical news site the Onion joked about violence between the AP and Chicago gangs and wrote that “an innocent 35-year-old passerby who found himself caught up in a long-winded dispute over use of the serial, or Oxford, comma had died of a self-inflicted gunshot wound.”

But more recently, Walt Hickey at the FiveThirtyEight blog decided to approach the argument in a more scientific fashion, by polling readers on their preference. What he found was that readers are fairly split: 57 percent prefer the serial comma, while 43 percent dislike it. So while there’s a preference for the serial comma, at least among FiveThirtyEight readers, it’s not an overwhelming one.

Roy Peter Clark at the Poynter Institute followed up by arguing that journalists should adopt the serial comma, after which Sam Kirkland posted a poll on Poynter asking if the Associated Press should make the switch. Surprisingly, a whopping 71 percent of respondents said yes.

But none of this was good enough for Poynter blogger Andrew Beaujon. He took to Twitter to lay out his arguments against the serial comma. The first argument—that the serial comma is simply ungrammatical—is actually the easiest to refute. He says that since you can’t write “My wife, and I drove to work” (note the comma before the “and”), you can’t write “Bob, my wife, and I drove to work.” But he’s simply presupposing that there’s a rule that you can never have a comma before a conjunction in a list, which is obviously not true.

He’s also presupposing that lists of two items behave exactly like lists of three or more items, but this is also untrue. I can’t write “Bob, my wife drove to work” (at least not with the intended meaning), but I can certainly write “Bob, my wife and I drove to work.” By Beaujon’s logic, the fact that you add a third item doesn’t magically obviate the rule that you can’t use a comma to join two things together. (Of course, this is done all the time in headline style, but it’s not allowed in normal prose.) It’s clear that the structure of lists changes a bit when you have three items or more, but it is not clear that there is any grammatical rule forbidding commas.

His next point, that “the sentences people employ to show the need for a serial comma are usually ridiculous”, is weak at best. Yes, it’s true that most contrived example sentences are a little ridiculous, but that’s just a problem with contrived example sentences, not with the argument for the serial comma. And consider this example: “The highlights of his global tour include encounters with Nelson Mandela, an 800-year-old demigod and a dildo collector.” It may be a ridiculous sentence, but it’s real.

Beaujon’s final argument is that the serial comma arises from an urge to overpunctuate because we believe readers are too stupid to figure things out on their own. He says that “prescriptivism is not for [the readers’] benefit; its purpose is to make those of us in the publishing game to feel important and necessary”, but how is his own prescriptivism any different? He’s instructing writers and editors in comma rules and telling them that following his rule means they’re good writers. In other words, my writing is so clear that I don’t need a crutch like the serial comma; if you disagree, it’s only because you don’t trust readers and want to make yourself feel important.

But consider this: most people in both the FiveThirtyEight and Poynter polls prefer the serial comma. That means most readers prefer it. If they find it helpful, who are we to argue that it’s some sort of crutch of bad writers or source of job security for copy editors? And AP style does in fact use the serial comma to prevent ambiguity (though apparently not in the above example regarding the late Mr. Mandela), so what’s the harm in using it all the time?

Because the fact is that I often stumble over sentences that lack the serial comma. Even though I’m well aware that AP and other styles omit the comma before the “and”, I still tend to read “bacon and eggs” in “He made muffins, bacon and eggs” as a single item, not the final two items in a list. The sudden end of the list after “eggs” throws me because I was expecting something to follow it.

I suppose you could conclude that I’m an idiot that can’t work out writing on his own, but you could just as easily (and much more charitably) conclude that the serial comma really is helpful because it signals something about the structure of the sentence. In speech, we can rely on a speaker’s prosody—the rise and fall of pitch—to tell us where the syntactic units begin and end. In writing, we have to rely on punctuation marks to serve as signposts.

You can claim all you want that your writing is so clear that it can do without these signposts, but if you leave out too many, your readers may feel lost, wandering through meandering sentences without knowing where they’re going. Did your reader immediately understand what you wrote, or did they stumble, backtrack, and read it again before they got your message?

This isn’t to say there’s one right way to punctuate, and it’s to use the serial comma. As we saw from the polls above, opinion on its use is still fairly divided. But rather than accusing your opponents of distrusting readers or being self-aggrandizing, you could take them at their word. Maybe there really are legitimate reasons to prefer the serial comma, just as there are legitimate reasons to prefer omitting it.

I find the arguments in favor of including the serial comma stronger than the arguments in favoring of leaving it out, but I don’t pretend that my preference is an ironclad grammatical law or proof of my superiority. It’s just that—a preference. You are free to choose for yourself.

11 Responses to Is the Oxford Comma Ungrammatical?

  1. Karen says:

    It’s easier to just use it every time—that way I don’t have to stop and reread every list and imagine whether it could be misinterpreted (which, presumably, is what the AP folks are doing—”unless it’s necessary to avoid ambiguity”). Granted, all of my clients are book publishers, and all follow Chicago rather than AP, so it’s not my choice anyhow. But it’s a simple way to ensure clarity, and that’s my goal.

  2. Jeromos Ferenc says:

    I agree entirely that people should be free to use the comma according to their preference but I am not entirely convinced by some of the things you say. For example, “if you leave out too many, your readers may feel lost” seems to me an argument in favour of “use it when it will prevent ambiguity” – otherwise “too many” is meaningless, surely?

    The sentences you give aren’t particularly helpful, either. The Mandela example would be ambiguous whether it included a serial coma or not. The “bacon and eggs” example fits the “if it will prevent ambiguity” model because bacon and eggs, when being referred to in a cooking context, are thought of as a discrete item so it is not a missing comma that is the problem but a missing “and”, in my view. “He made muffins and bacon and eggs” would fix that. Adding a comma (“He made muffins, bacon, and eggs”) wouldn’t be wrong but it doesn’t address the problem of “bacon and eggs”. A true three-part list might be ““He made muffins, bacon and eggs, and toast and marmalade” where the comma before the last item is necessary because of the two “and”s in the compound objects – at least, that is how I read it.

  3. Chris says:

    “Hillary, my wife, and I went to the store.” is ambiguous with the comma. If the reader had reason to believe that I had a wife named Hillary, they might assume that only two people went to the store. In speech, we’d be careful to set off the “my wife” appropriately, depending on whether Hillary was my wife or a different person. In text, we get commas, a much blunter tool.

    In reality, of course, we’d say, “All three of us went to the store: Hillary, my wife, and I.” so as to include plenty of information and be clear.

  4. The civil and uncivil wars over the serial comma extend deep into the Poynter Institute’s rank and file writers and teachers. Want to make sure you and your readers caught my defense of the Oxford comma, which puzzled and frustrated my colleague Andrew Beaujon. Happy to see so many of the Poynter readers, many of whom are journalists and follow AP Style, support that final comma.

  5. Pingback: Oxford commas, Nelson Mandela, and Stephen King | Sentence first

  6. Karen: Agreed. Though as some point out, there still is potential ambiguity with some lists, like the Nelson Mandela one above, because the middle item could be read as an appositive describing the first. I think context usually disambiguates these, but context doesn’t usually help me not jerk to a stop when I read a list that lacks the serial comma.

    Jeromos: In that line you quoted about leaving things out, I was talking about punctuation generally, not serial commas specifically. I meant that omitting punctuation whenever you can may make your audience feel lost, which is an argument in favor of not reflexively omitting things.

    And yes, you’re right that the Mandela line would still be ambiguous. And to be fair, it’s only technically ambiguous without the comma. I doubt any educated reader would seriously take the last two items as an appositive, but I think including the comma would reduce the numbers of double-takes by misled readers.

    If you dislike my “bacon and eggs” example, you can substitute something else. “He made muffins, bacon and toast” would have the same problem from my point of view—I assume that the reader conceives of “bacon and toast” as a single item, and I expect another item to follow.

    Chris: Yes, there’s still a potential ambiguity, but I think it’s a very small one. I think people are more likely to say “my wife, Hilary,” than “Hilary, my wife.” But you’re right that we are usually avoid these problems in speech. We have the option there of not just wording things differently but also inflecting them differently. And if all else fails, we can just explain to our confused interlocutors. In writing, you don’t get that chance, and so it helps to try to be clear the first time.

    Roy: Thanks for the link. I should have included it in my post.

  7. Susan says:

    Thank you Professor Jonathan for posting a blog on this comma, which has been splicing my life for the past two months.
    I was taught from childhood, that there is no need for a comma before a conjunction such as and as this was redundant there.
    It was recently when I started writing that people were sending me back my work asking me to edit and add the comma-I am a writer at a client’s mercy, so I had to concur. But a doubt nagged me all the while, is it necessary or not.
    And I am glad to know that there are two viewpoints.
    So it depends on us readers to advocate the use of the comma before an “and”.

  8. Jake says:

    I think your last point (“It’s just that—a preference”) really sums up the situation. I’m convinced that any attempt to appeal to grammar in an argument for or against the Oxford comma is bound to fail, simply because this is not at its heart a grammatical issue. It is simply a style choice. Granted, the Oxford comma may sometimes prevent ambiguity, but that’s not really grammar, anyway.

  9. Chris Elliott says:

    I prefer omission. I gathered my thoughts on the topic and prepared this essay, which lays it out in terms of The Three Stooges. The main argument is that a good writer should be able to exercise his or her option on this, relative to the intended rhythm of the sentence.

  10. Sylvius the Mad says:

    If we truly wanted to reduce ambiguity, we would eliminate any uncertainty about the use of the serial comma. Either we all use it, or none of us do. Alternatively, every piece of text containing a list should explicitly declare whether it uses the serial comma.

    The absence of the serial comma is especially confusing when we were expecting it, and the presence of the serial comma is confusing when we are not. We need to know before we start reading what the rules would be. We either need one set of rules everyone follows, or we need each author to declare which rules are being used in advance.

  11. We need to know before we start reading what the rules would be. We either need one set of rules everyone follows, or we need each author to declare which rules are being used in advance.

    But wouldn’t this amount to tacking an entire style guide onto every article or book written? But then you might as well shorten it to a statement like “This text uses The Chicago Manual of Style.” I suspect, though, that vanishingly few readers would even care, and those who do are probably astute enough to recognize which style is being followed anyway.

Leave a Reply

Your email address will not be published. Required fields are marked *

%d bloggers like this: