On the state of Scientific English and how to improve it – Part 3


How monotremes and planets help me communicate the distinction between “which” and “that”

Figure 1.

Starter for 10 points: Do all monotremes lack imprinting? OK, I'm trying to sell a rather drab topic with a more interesting lead. I've stood in front of audiences and stated that it is important to make the distinction between “which” and “that”, and eyes have glazed over in a combination of incomprehension, indifference and incredulity. If anyone is even vaguely aware of a difference, he or she has probably read something about “which” being used in a non-restrictive clause, and “that” being used in a restrictive clause; but that's probably best left to the linguists. We need some tangible, scientific, examples, because most people I've talked to think that there is no difference between “which” and “that”, and certainly none worth mentioning. After all, don't we use them interchangeably? Often, yes; and in UK English, very little distinction is made between the two; but in American English there's a very useful little rule. It is precisely because of different cultural habits in using these two relative pronouns that a problem arises. And here it is:

“…recent analyses made possible by sequencing of marsupial and monotreme genomes also revealed an accumulation of DNA and LTR repeats in imprinted domains in eutherians and marsupials compared to monotremes that lack imprinting.”

Had I read this in a published manuscript, I would simply have continued – somewhat uncomfortably, mind – with less than certain factual knowledge: given that I am not an expert on monotreme genetics, the authors might be right in using “that”. Why even question it? Because there is a logical distinction in American English between “that” without punctuation, and “which” preceded by a comma, and I've learnt to appreciate it in science; reading such a sentence as an editor now automatically triggers a question, because I want readers to have the most unambiguous understanding of an article as possible: an American reader would be grammatically justified in understanding the above sentence as “compared to that sub-set of monotremes that lacks imprinting”, i.e. “that” indicates a sub-set of the entire set; a UK reader would have two ways of interpreting the sentence. Perhaps as little as 1/4 of the readers would be absolutely sure to have interpreted the sentence correctly, because what the author really meant was:

“…compared to monotremes, which lack imprinting.”

hence – according to the rule – leaving no doubt that it is not a sub-set of all monotremes that is being described here, but all monotremes. “Which” preceded by a comma indicates that the qualification refers to the whole set of objects mentioned. Scientists are more likely to understand the grammatical concept in terms of its consequences for membership of sets, and so if any tutor of science writing courses reads this editorial, I can recommend using that logic, rather than trying to explain restrictive and non-restrictive clauses.

But maybe I've picked the only ambiguous example amongst countless thousands of unambiguous ones from my editorial work. Not so. Here's another:

“Cell extracts from the different cell lines were treated with the denaturing reagent. The supernatant which contained the protein of interest was aspirated and stored on ice.”

What would readers understand here? A UK native English reader could choose to read “which” as “that”, hence defining the meaning as only a sub-set of the supernatants containing the protein of interest. A US native English reader might think that the author mistakenly used “which”, and intended “that”, or that he omitted the comma before “which”. In fact, the author clarified the sentence to read:

“The supernatant, which contained the protein of interest…”

because all of the supernatants contained the protein.

Here's the rule, once again, with celestial examples:

The moon, which orbits at around 384,400 km from the Earth, is covered with craters. (i.e., there is only one moon orbiting Earth; hence by definition this describes the whole set)

The moon of Saturn that I saw last night with my telescope is called Titan. (Saturn has 62 moons!!)


Andrew Moore