Nancy's Wordsmithy
Rules You Don't Have to Obey, Part III

by Nancy Allison

Home / Archives / Back Issues 1989-1990 /

No self-respecting wordsmith can avoid this one forever. I may as well get it over with. I don't have to tell you which rule I mean. You know the one... But in case you don't, I will state it here:

Use the relative pronoun that with a restrictive clause and don't set the clause off with commas. Use the relative pronoun which with a non-restrictive clause and set the clause off with commas.

Yes, thatrule, the one over which gallons of writerly blood have been shed throughout the past century or two. I myself have witnessed, and even participated in, a number of fierce duels which, needless to say, failed to resolve the dispute.

The funny thing is, this rule should be running out of steam, because certain standards of written English have changed in ways that make the rule at least partly obsolete. Learning it is kind of like learning to change a cloth ribbon on an old manual typewriter. You get hot and messy and short-tempered doing it, and when you're done you don't have much use for your accomplishment. You use the manual typewriter only when the computer is down, and you use the sentences produced by the that/whichrule only when no better sentence structures are possible.

Still, the passions aroused by this rule are so great that none of us can afford to ignore it. Let's take a good look at the source of our discomfort.

The first thing to note is that this rule isn't exactly intuitive. It requires us to learn a pattern that uses a class of words (relative pronouns) and a punctuation mark (the comma) to achieve its goal. For starters, the difference between that and which is not self-evident. Their meanings overlap: The bus that I tookand the bus which I took-- what's the difference? What does that mean that which couldn't also mean? And vice versa.

To be fair, their meanings don't overlap one hundred percent. Thatsounds all right with either animate or inanimate things:

The program that Tina worked on is in this directory.

The engineers that Paul interviewed work here.

Which,on the other hand, usually sounds right only with inanimate things:

The program, which Tina worked on, is in this directory.

Sounds good.

The engineers, which Paul interviewed, work there.

Doesn't sound right. We use that(or whoor whom) with people.

So far, so good, almost. We can at least say that thatand whichare appropriate for slightly different subjects. If this was all the rule required us to learn, we wouldn't have much trouble with it.

Unfortunately, there's more.

Use that with a restrictive clause not set off with commas. Use which with a non-restrictive clause set off with commas.

OK, so what's a restrictive clause? A restrictive clause is one that contains information essential to its subject.

The keyboards that were manufactured in Korea have been sold.

In this example, the clause that were manufactured in Koreaadds information that limits the subject to a particular group: those keyboards that were manufactured in Korea, as opposed to those manufactured in the rest of the world. If we omit this information, we change the subject of the sentence. To indicate that this information should not be omitted, no commas are used and the clause is visually joined to the rest of the sentence.

A non-restrictive clause, on the other hand, adds information that is not essential to the identity of the subject.

The keyboards, which were manufactured in Korea, have been sold.

In this sentence, the subject is an entire group of keyboards. The information contained in which were manufactured in Koreaadds information about the subject but does not limit it. If the clause is omitted, the subject of the sentence remains the same: the entire group of keyboards. To give a visual cue that this information is optional, it is set off by commas.

Thus, in these two sentences we see the rigorous logic, the splendid clarity, of this rule. Breath-taking, isn't it? If properly employed, it will keep us from ever scratching our heads and muttering "Now, is that supposed to be restrictive or non-restrictive?" In fact, the rule is even fail-safe, since it employs redundant markers: it uses both relative pronouns and commas to make its point. Yes, this rule comes with its own backup system.

Well, is this enough to recommend it? The rule is fairly easy to follow -- fool-proof, even -- and it delivers clear-cut results. Does this mean we should apply it wherever suitable?

No, of course not. As with every other grammatical rule, we need to ask ourselves, Does this rule produce the best writing possible? And the answer is no, not always. Example:

The figures that are labeled show the A/D card.

Wouldn't most of us tighten up this sentence, producing this result:

The labeled figures show the A/D card.

Or this:

The car that Jerry bought has been recalled.

Wouldn't most editors omit the that:

The car Jerry bought has been recalled.

What's at issue here isn't whether each of us agrees with this rule. Probably most of us would say that it is important to show a distinction between restrictive and non-restrictive clauses. We would probably disagree over the choice of pronouns but agree that the presence of absence of commas is required to show whether the clause is restrictive or not.

The bigger consideration is that the rule itself is less and less relevant to our work because we often don't like the sentences it produces. Some of those sentences were acceptable back in the days of cloth typewriter ribbons but are a bit too leisurely for the era of PCs and Macintoshes. Thus, if we only struggle to apply the rule to our work and fail to criticize its results, we will frequently write sentences that are technically correct, but too wordy to be acceptable.


© 2000 by STC Boston, Boston, Massachusetts, USA
Originally published May/June, 1989 in the Boston Broadside