The hidden mental models behind the fight over the Oxford comma

A nasty, permanent spat burns in the heart of analysts of the English language. It’s the fight over the need for the Oxford or serial comma — for example, do you really need the final comma in the phrase “passive voice, weasel words, and jargon”? Now that a court in Maine has decided a case based on the absence of an Oxford comma, the dogmatic dispute between its proponents and opponents will only get worse. Today I’ll reveal why each side thinks its right — and give you the ammunition to make the proper choice.

How the Oxford comma decided a case in Maine

Drivers for the Oakhurst dairy in Maine think they deserve overtime. Whether they do or not depends on how you interpret this clause in the Maine state law that lists the activities that do not qualify for overtime pay:

The canning, processing, preserving, freezing, drying, marketing, storing, packing for shipment or distribution of:
(1) Agricultural produce;
(2) Meat and fish products; and
(3) Perishable foods.

Clearly, you can’t get paid overtime for canning, processing, preserving, drying, marketing, or storing food. But what about distribution (like, say, driving a milk truck)? Well it depends on whether the final item in the series is “packing for shipment or distribution” or just “distribution.” If there were an Oxford comma before “or distribution,” then distribution would be another activity that doesn’t qualify for overtime, and the dairy managers would win. But without the comma, it’s ambiguous, because the overtime-prohibited activity could be “packing for shipment or distribution.” The court ruled that because of that ambiguity, the state law does not specifically exempt distribution activities from overtime, so the drivers won their case.

Note that the semicolon on line starting with “(2)” is a serial semicolon — if you include one there before “and,” why not include a comma before the “or”?

The style guide for the state of Maine recommends against the Oxford comma. But those who wrote the guide didn’t foresee that omitting it (or, as in this case, using it inconsistently) would result in ambiguities like this one.

Oxford comma proponents like me can now hit their opponents over the head with this case. But why are the people on both sides so passionately sure that they’re right?

Why you think the way you do about the Oxford comma

Oxford comma proponents insist that it promotes consistency. Its opponents say the same. To explain why they’re both so sure, I’ll use the logic of computer code. If you’re not a coder, just skip the code samples here; I’ll explain in words as well.

Suppose you were writing about a series of n parallel terms, call them A(1), A(2), . . . A(n). An Oxford comma proponent needs to make an exception if there are only two terms, in which case there is no comma. So if they were to write a computer program to create text from that list, they’d write this algorithm. (This is “pseudocode,” not an actual programming language; I assume that there are at least two terms; and the stuff after the semicolons represents comments, not code.)

SUBROUTINE TypeSerialListOxford(A)

LET n=count(A) ; n is the number of items in the series

IF n = 2 THEN TYPE A(1)+” and “+A(2) ; type the only two terms with ” and ” between them

ELSE FOR i = 1 to n-1

TYPE A(n)+”, ” ; use Oxford comma after all elements of the list except the last

NEXT i

TYPE ” and “+A(n) ; type the ” and ” and the last element

END SUBROUTINE

In words, this algorithm says “If there are two terms in the list, just write them with ‘ and ‘ between them. If there are more than two, type each term followed by a comma until you get to the last one, they type ” and ” and the last term. Take note that the special case here is the two-term list, which we handle differently from longer lists.

Those who eschew the Oxford comma can write their own program, which looks like this (it’s recursive):

SUBROUTINE TypeSerialListNonOxford(A)

LET n=count(A) ; n is the number of items in the series

IF n = 2 THEN LET TypeString = A(1)+” and “+A(2) ; if the list is reduced to two terms, type them with ” and ” between them

ELSE TYPE A(1)+”, ” ; if longer than two, type the first element and then apply the algorithm to the remainder of the list

LET B = last n-1 terms of A ; remainder of the list

RUN TypeSerialListNonOxford(B) ; recursively run algorithm on the remainder of the list

END SUBROUTINE

Here’s how this recursive program works. If it’s a two-term list, it just writes the two terms with “and” between them. But if there are three or more terms, you type the first term with a comma, then apply the algorithm to the rest of the list.

Which one is more consistent?

• The Oxford comma algorithm works through the list from the first element to the last, which is the natural order for writers. It is inconsistent, though, in that it treats two-element lists differently from other lists.
• The non-Oxford comma starts with the last two elements of the list, which it always types without a comma. For longer lists, it disposes of the first element and its comma, then works on the smaller remaining list, until it gets down to the last two elements, which it of course always types without a comma. This makes it more consistent, but at the cost of thinking recursively rather than iteratively.

I doubt that proponents of either side would put their arguments in these terms, but this is what is happening.

• The Oxford proponents think they are absolutely consistent, because they’re always putting commas after list terms. They are — but their long lists are inconsistent with their two-element lists.
• The Oxford opponents think they are absolutely consistent, because they treat all lists the same, with the final two elements comma-free. While they treat all lists the same, there is an inconsistency in every list they type: they have to handle the last two elements differently from the other elements.

So who’s right?

If we can’t settle the argument on logic and consistency, can we settle it empirically?

The Oxford comma proponents cite phrases that are ambiguous if you leave the Oxford comma out. That’s what happened in the Maine court case, and it’s also what’s going on in this famous example of a BBC description of a program Peter Ustinov:

By train, plane and sedan chair, Peter Ustinov retraces a journey made by Mark Twain a century ago. The highlights of his global tour include encounters with Nelson Mandela, an 800-year-old demigod and a dildo collector.

Without the Oxford comma, the great South African leader becomes not just divine, but a fetishist.

The case for the Oxford comma opponents centers on the concept that additional punctuation slows people down. If you subscribe to the idea of omitting extra words (which I do), then shouldn’t you also omit needless punctuation? We’ve got the “and” to tell us we’re at the end of the list; who needs a comma, too?

This is a reasonable argument. But you have to make a decision, and I’ve made mine. To save the reputation of Nelson Mandela, avoid ambiguity in labor laws in Maine, and preserve the clarity of everything you write, hold onto your Oxford comma.