What word is THAT?

What word is that?

More accurately, what word is “that”? For such a common and apparently simple word, it presents some interesting problems, and can be misused. Since misuse can confuse the reader, and since our first responsibility as communicators is to be clear, it seems like a good idea to clarify the principles underlying its use.

In fact, traditional grammar recognises four words spelt and pronounced the same way:

Demonstrative pronoun, as in “That might happen if you omit the regular dose.”
Demonstrative adjective, as in “That function is not available.”
Relative pronoun, as in “The function that you have requested is not available.”
Complementiser (ugly name!), as in “This reaction indicates that the patient is allergic.”

The two demonstratives are usually contrasted with “this”: “Do you want this, or would you prefer that?” “This book is more interesting than that one.”

The relative pronoun is sometimes interchangeable with “which” or “who[m]”, or can be omitted: “The function which you have requested is not available.” “Users who do not have administrator rights cannot perform this function.” “The function you have requested is not available.”

The complementiser offers no alternatives or distinctions, but can sometimes be omitted: “This reaction indicates the patient is allergic.” “That the patient was allergic seemed obvious.”

There’s little chance of going wrong with the choice between “this” and “that”; in general, “this” is closer, while “that” is further away. But the relative pronoun and the complementiser do require some care. Let’s take them one at a time.

Use of “that” as a relative pronoun presents two problems: whether to use “that” or “who”; and whether to use “that” or “which”. The first is simple: never use “that” to refer to a human. Avoid constructions like “Patients that take this medication …”

The second is a little more difficult, because there are two kinds of relative clause. Traditionally, these are called restrictive and non-restrictive, but I prefer the terms defining and describing. The restrictive or defining type provides information that the reader needs, to identify what you are referring to: “The OK button that is in the Find window should be disabled” suggests that there are other OK buttons in other windows, but they aren’t what you’re interested in. The non-restrictive or describing type gives additional, non-essential information: “The tablets, which are wrapped in foil, are pink and triangular.” Notice that you cannot use “that” instead of “which” in the second example. However, you could use “which” instead of “that” in the first — except in US English, which insists on “that”. Notice too that you can reconstruct both examples to use a phrase or subjectless clause, thereby avoiding both “which” and “that”: “The OK button in the Find window …” and “The tablets, wrapped in foil, …”

Use or omission of “that” as a complementiser also presents an interesting problem. In general, it seems like a good idea to leave out unnecessary words — so, if we can omit the complementiser, is there ever any point in retaining it? The short answer, as I will illustrate, is “Yes” — and, if it is sometimes necessary, then perhaps stylistic consistency requires that we never omit it.

Consider what happens as we read a sentence like this:

You should know the French language is not supported in this version.

Many — perhaps most — readers begin by interpreting the first six words at their face value: You should know the French language. They are probably already forming the question “Why?” when they read on to discover that this is the wrong interpretation. (This is an example of what linguists call garden path ambiguity; the most obvious interpretation of the first few words leads the reader up the garden path.) If we change the sentence to You should know that the French language is not supported in this version, the problem goes away.