Deborah Cameron on thats and whiches

Linguist Arnold Zwicky's delightful Language Log post on the brouhahas over whiches and thats, Five more thoughts on the That Rule, reminded me that I had the following piece from Deborah Cameron lurking on my hard drive.

Deborah Cameron, Verbal Hygiene (Routledge: 1995), 50ff.

A friend of mine once worked as a copy editor for a major publisher in New York. The project she worked on was a large encyclopedia, and her job was to edit a large number of contributions from all over the world in accordance with the rules laid down in the Chicago Manual of Style.

Among these rules is one that concerns the use of that or which in relative clauses. That is prescribed in those cases where the clause is 'restrictive', e.g 'the book that Nigel gave me was no good', while which is used in 'nonrestrictive' clauses, e.g 'the book, which Nigel gave me, was no good'.

The difference between the two sentences above is one of those subtleties beloved of language mavens everywhere. In the first sentence, the relative clause that Nigel gave me 'restricts' the reference of the book, making clear that I am talking specifically about the particular book that Nigel gave me, as distinct from all the other books I possess. In the second sentence the information that Nigel gave me the book in question is still present, but it simply adds incidental information rather than being necessary for the identification of one out of a whole class of possible referents.

English speakers normally put which in non-restrictive clauses, but they quite often fail to observe the part of the rule that prescribes that in restrictive clauses: many native speakers find it equally acceptable to use either 'the book that Nigel gave me' or 'the book which Nigel gave me' (as well as a version with no relative pronoun, 'the book Nigel gave me'). It is not that such people perceive no difference between restrictive and non-restrictive relative clauses; it is rather that, for them, the distinction is carried by the commas that mark off the non-restrictive clause (or the prosody, in the unlikely event of someone uttering this sentence), and not by the choice of pronoun. Nevertheless, the Chicago Manual of Syle insists on that rather than which in restrictive relative clauses. Copy editors therefore spend a good deal of time correcting which to that in writers' copy.

British readers unacquainted with the Chicago Manual of Syle may well have followed this discussion with a degree of bewilderment. The rule about that and which is not insisted on in Britain as it is in the US; though some authorities (such as Fowler and The Times guide) do recommend it, it is not an absolute prescription. Bewilderment mixed with irritation was certainly the reaction of many British contributors to the encyclopedia my friend was editing. She sent them proofs with the whiches changed to thats; they promptly returned them with the thats changed back to whiches. My friend referred the matter to her boss, the 'copy chief', for an authoritative ruling on the entire pronoun question. After due deliberation he handed down his decision. Britons could write which if they wanted, but Americans must go by the book and write that.

This incident is absurd, and it was recounted to me as an absurdity. Nevertheless, it reveals a number of interesting things. First, it reveals that editorial practices need have nothing to do with communicational efficiency. Although the use of which in restrictive clauses may strike educated Americans as inelegant, one can scarcely imagine it interfering with their comprehension of the text. Second, it is notable that the copy chief did not try to resolve the that versus which problem by an appeal to rational principles. His solution implicitly acknowledged that no principled argument could be advanced in support of either alternative. Third, the outcome reveals that the much-vaunted principle of consistency can be ignored in certain circumstances. The copy chief did not say, as one might have expected, that while the rule might be arbitrary, it must be observed by all. On the contrary, he ended up allowing the encyclopedia as a whole to exhibit the very inconsistency the rule was supposed to eliminate. The chief did not however go so far as to give individual authors freedom of choice on the relative pronoun issue. The overall inconsistency of usage had to be structured by national affiliation — one rule for British writers, another for US writers (inevitably disputes arose later about how to categorize Australian and Canadian contributors). The underlying concern, then, was not that the text should be either clear or consistent: it was that people should follow rules. Indeed we might read the whole affair as a kind of demarcation dispute in which members of one 'guild', the American copy editors, agreed to respect the differing craft practices of their colleagues across the Atlantic, while continuing to uphold the authority of such practices in themselves.

Arguably, however, it was not only the need to preserve editorial authority that led to the bizarre outcome of the relative pronoun controversy; it is also relevant to consider the working culture of the copy editors. Editors are not just automata, mindlessly applying the rules. What looks like excessive zeal on their part may in fact be a mixture of self-interest and subversion, as practised by alienated workers everywhere.

My friend and most of her fellow toilers on the encyclopedia fell short of Elsie Myers Stainton's ideal: they were not caring fusspots but graduate students working for doctoral degrees. Casual editorial work suited their need for flexibility at reasonable rates of pay, and they suited the publisher's need for a smart and highly literate workforce which would nevertheless be relatively cheap and disposable. They were hired to work on a specific project, and liable to be laid off when it was finished. While they worked they were paid by the hour. This particular group may well have represented an extreme of casualization and at times disaffection, but freelance arrangements of a roughly comparable kind are common in the editing trade, and these conditions affect the way the job is done.

For example, it was in the encyclopedia editors' interests to prolong work rather than hurrying to finish it. They lost money — and ultimately their jobs — if they worked too quickly. On the other hand, as casual workers they were easy to fire if they were thought to be shirking or 'padding their hours'. The best way to cope with these conditions was to edit copy with extreme thoroughness, both to display conscientiousness and to maximize the hours for which they would be paid. In their attempts to meet these criteria, they adopted the maxim of 'not just passing copy'. Whatever could be queried would be. This group of editors took particular pleasure in generating an obscure query that would need referring up to the copy chief. This was at once a good delaying tactic, proof of keenness and an outlet for underused creativity.

From an editor's perspective, then, hyperstandardization has its advantages: it makes a thorough editing job a relatively long job, a source of financial as well as professional satisfaction. l am not, in fact, the only linguist to have harboured this sort of suspicion. A review of a (British) monograph in the (US) scholarly journal Language ends with the following remark:

[The author] and presumably the [publisher's] copy editor make no attempt to observe the infamous that/which distinction in restrictive relative clauses. For this relief, much thanks, and why can't American publishers give up on this device, whose sole virtue (speaking as the husband of a copy editor) is to give copy editors more billable hours?

(Aronoff 1992: 610)

That billable hours are not the sole reason for editorial practices is strikingly borne out by the fact I noted earlier: that some copy editors continue to enact the role of 'fusspot' even when their employers are begging them not to (the employer who told me this did pay by the hour, but there was a limit beyond which editors had to fuss at their own expense). Even among the encyclopedia casuals, for whom editing was a pin-money job and not a lifetime vocation, it was obvious there was professional pride at stake as well as dollars and cents. The working conditions and culture of editors are defined by a combination of economic and professional considerations, and these are both factors in explaining, if not the phenomenon of hyperstandardization itself, then at least how and why it is enacted so zealously in everyday working practice.