Punctuation as a system

When readers see a comma, they get a signal like "These two parts of the sentence are being separated for a reason." -- Ann Raimes

Ann Raimes' statement cuts through so much confusion about punctuation, it makes an ideal opener for this discussion. It contains two crucial ideas. One, punctuation is a set of signals for the reader on how to read the words, not a set of rules in a writing manual. It is, in other words, part of the living dynamic of rhetoric. Two, most punctuation marks are there for a reason, not just because some usage manual says to put them there.

Style manuals and handbooks "have it all wrong," argues John Dawkins in a recent article in College Composition and Communication. A startling claim, but Dawkins is serious. All the manuals, he points out, tend to fragment punctuation into an incoherent body of instruction by dealing with each mark separately: a section for commas, a section for semicolons, a section for colons, and so forth, each with its own set of rules for proper use. They fail to present punctuation as a hierarchical system, one that Dawkins calls "surprisingly simple. . . [one] that enables writers to achieve important -- even subtle -- effects."

Dawkins illustrates the system with a series of tables, which we have combined into the single table on page 3. Because the table at first glance might look a little daunting, we will try to provide some explanation and examples to guide you through it.

A hierarchy of punctuation

The main feature to notice about the table is that the punctuation marks are shown in relation to one another, in hierarchy. As you move up the scale you create more separation between statements, and more emphasis on the word or words just before the punctuation mark. (The dash and the colon are about equal, but they serve somewhat different purposes.)

Consider the following famous aphorism by Mark Twain:

[1] Man is the only animal that blushes. Or needs to.

Twain might have punctuated it several other ways.

[2] Man is the only animal that blushes or needs to.
[3] Man is the only animal that blushes, or needs to.
[4] Man is the only animal that blushes -- or needs to.
[5] Man is the only animal that blushes: or needs to.
[6] Man is the only animal that blushes; or needs to.

All of these are possible, and each of them creates a subtly different effect. Sentence [1], the one Twain actually wrote, has maximum separation between the parts and puts maximum emphasis on blushes. Sentence [2], at the opposite extreme, has no separation. Notice how it seems to flow by all in one breath, so to speak, with no emphasis on the key word, blushes. It's pretty clear why Twain didn't choose zero punctuation.

Sentences [3] through [6] have increasing separation and emphasis: try saying them aloud. Of these options, the most likely candidate might be [4], with the dash. But there is an important difference between [1] and [4]. In [4], the dash says, "We are pausing here, but watch out, because something even more important is coming next." In [1], the period says, "That's it; stop." So when "Or needs to" comes, it comes not as an expected punch but almost as an afterthought. Which sentence, for you, makes the better joke?

But what about correctness?

"Now hold on a minute," you might be saying, "the period in Twain's sentence creates a fragment, and I'm trying to get my students not to write fragments." Yes, with good reason. Twain was writing literary humor, not a research report or a business letter, where readers might be confused or irritated by fragments. We've tried to cover that distinction in the table by labeling such use of the period "for special effects only."

The same is true of comma splices (joining of complete sentences with commas but no conjunction). In certain contexts they are not allowed, because they can confuse readers as to where one statement ends and the next one begins. In technical and legal documents it's particularly important to avoid any uncertainty. But in other contexts they can be made to work. Novelists and essayists have used them for years. E. M. Forster was what Dawkins calls "a fearless comma splicer" and wrote sentences like [7], quoted by Dawkins.

[7] He could not stand the insecurities that are customary between officials, he refused to make use of the face- saving apparatus that they so liberally employ.

Forster apparently wanted only slight separation between statements, and he did not want to unduly emphasize officials. His comma splice is not confusing or irritating, because the two parts are parallel, each beginning with the same subject and an equivalent verb: "He could not stand. . . , he refused. . . ." The connection is reinforced by the tie between officials and its pronoun, they.

Fragments and comma splices, in some genres of writing, are not errors but tools to create special effects. So are uses of the semicolon to create separate parts that aren't sentences in themselves. As with other tools they take practice to use well, and a good ear for rhetorical subtlety.

The versatile middle group

The middle or medium marks, colons and dashes, are interesting because they are so versatile. They can separate almost anything, from sentences to single words.

[8] Midas was interested in one thing: gold.
or [9] Gold: Midas was interested in that alone.
[10] Jack is a true vintage-car fanatic: he'll hear about some restored Maxwell or Reo and drive hundreds of miles just to look at it.

A dash could be used in place of the colon in either sentence. But put dashes in place of the colons, and you hear a slightly different sentence. Dawkins calls the colon anticipatory and the dash emphatic. We might quibble that the difference is more subtle than that, but in most cases his descriptions will also work. Of course, a period or semicolon could be used in [10]. Try substituting those for the colon and see how the separation and emphasis change.

The semicolon

The semicolon is an odd duck; it was borrowed from the Greek question mark. Lewis Thomas, in his delightful essay "Notes on Punctuation," observes that one gets "a strange sensation to read a Greek sentence which is a straightforward question: Why weepest thou; (instead of Why weepest thou? . . .)."

Thomas continues:

I have grown quite fond of the semicolon in recent years. The semicolon tells you that there is still some question about the preceding full sentence; something needs to be added; it reminds you sometimes of the Greek usage. It is almost always a greater pleasure to come across a semicolon than a period. The period tells you that that is that; if you didn't get all the meaning you wanted or expected, anyway you got all the writer intended to parcel out and now you have to move along. But with a semicolon there you get a pleasant little feeling of expectancy; there is more to come; to read on; it will get clearer.

Thomas captures the spirit of punctuation as a form of play -- which is maybe what this discussion is about. When writers are encouraged to play, they learn more about the rhetorical effects produced by their writing.

In most genres, of course, room to play is fairly limited; all of us have to learn to follow conventions much of the time. But the punctuation chart can still be useful, to students and teachers alike. It replaces (partially, at least) the distasteful task of policing and correcting punctuation at random with a system that can be taught and learned: separation and emphasis, raising and lowering. The three most common punctuation errors: placing commas where no punctuation is needed, placing commas where higher punctuation is needed, and placing periods or semicolons where lower punctuation is needed, can all be dealt with in terms of raising and lowering. Writers tend to commit the same few punctuation errors repeatedly, so that what looks like a paper riddled with bad punctuation really has just two or three different errors, repeated.

And though Dawkins debunks the handbooks and manuals, they still have their place. We need them to look up usages where convention, not rhetorical effect, rules in all genres. (For example, in some respects the American convention for punctuating in and around quotations is the opposite of the British; both being largely arbitrary.) And we can still refer to the separate sections in the handbooks for advice on commas, colons, semicolons, and periods -- once we understand that they are interrelated. And that there's a hierarchical system the handbooks fail to acknowledge.

RL