‘Ware man-traps: rethinking an OED entry

By James McCracken, Principal Editor, OED

When revising an OED entry, one often ends up thinking about the meaning not only of the word in question but also of the entry itself—what, in other words, was intended by the editor who originally put it together? Even the shortest entry may have involved dozens of editorial decisions, and, under the scrutiny of the revision process, these century-old trains of thought can occasionally reveal themselves.

In the case of mantrap (or Man-trap, as the first edition spells it), one important editorial decision must have been that of splitting the entry into two senses (unnumbered, but clearly differentiated): a literal sense first and a figurative sense following:

First Edition (1905)

An early draft gave a definition for the second sense: ‘fig. A person or thing which is likely to prove a pitfall or to cause injury to persons.’ But this was later crossed out and replaced with the more open-ended ‘transf[erred use]. and fig[urative].’ marking.

The second edition entry adds some more recent quotations but, characteristically, leaves the basic arrangement of the entry untouched:

Second Edition (1989)

There’s an interesting point to note about this arrangement: it makes it appear as if the literal sense gave rise to the figurative sense, in spite of the fact that the former is documented fifteen years later than the latter. Henry Bradley, the editor who approved Man-trap for the first edition in 1905, had James Murray’s backing for this. Murray stipulated that senses should be ordered chronologically according to the earliest evidence for each, except when this contradicted what one could reason about the order in which senses should have developed: ‘That sense is placed first which was actually the earliest in the language: the others follow in the order in which they appear to have arisen.’ So there are numerous entries like Man-trap in which the ordering of senses does not follow their chronology.

It’s easy to see why Murray wanted to keep both options open. The readers who collected quotation evidence for the first edition were impressively skilled and productive, but they could by no means guarantee to have uncovered the earliest example of every word, or of every sense of a word. Like palaeontologists tracing the evolution of dinosaurs from an incomplete fossil record, Murray’s team of editors had to keep considering whether an incomplete quotation record could be relied upon to set out accurately the history of a word’s evolution. Now and then they evidently decided that it could not.

Man-trap appears to be just such a case. Bradley must have reasoned to himself that since one can’t use a word in a figurative sense before it exists in a literal sense, the evidence collected so far must give an incomplete picture. With only fifteen years in it, he probably felt that this wasn’t chancing his arm too far.

When we came to revise the entry for the third edition, using much more extensive research resources than were available to the original editors, we were able to antedate both these senses. We found an example of the literal sense from 1775 in the comic opera The Rival Candidates. We also found an example of the figurative sense from 1726 in Charles Johnson’s comedy The Female Fortune-Teller. (Johnson was a minor Grub-Street dramatist: a scathing footnote in Pope’s Dunciad has him spending his days in a Covent Garden coffee house, ‘a martyr to obesity’.) In The Female Fortune-Teller there is a scene in which the credulous Astraea questions a mysterious all-knowing brass head. Brazen as well as brass, the head cynically disabuses Astraea of various cherished ideals, informing her that so-called truth is nothing but opinion, that love is a mere effect of bodily constitution, and finally that marriage is ‘a man-trap’.

These antedatings significantly widened the gap between the two senses identified in the first edition entry. As we filled in more of the picture, then, it seemed that the evidence had been pointing in the right direction all along: the figurative sense now appeared to be almost half a century earlier than the literal one. It was starting to look as if Bradley’s reasoning had misled him somehow.

Could a word exist in a figurative sense before it existed in any literal sense? It’s not impossible for certain kinds of words, for example those borrowed from other languages; but it is unusual. Perhaps, then, in the case of mantrap it is inaccurate to identify the two senses as ‘literal’ and ‘figurative’. Or perhaps the word doesn’t really fall into these two senses at all.

When we re-examined Bradley’s six quotations with these doubts in mind, we began to see an awkwardness about the plan. The last quotation, from William Greener’s Science of Gunnery, sits ill with the other ‘transf. and fig.’ quotations: Greener uses mantrap in more or less the same way we now use deathtrap, and so has little in common with Goldsmith and Dickens, who use mantrap in the context of sex and marriage.

With a relatively small array of evidence to go on, one can readily understand Bradley not making a meal out of this. But the quotations added for the second edition entry make this tension within the ‘transf. and fig.‘ sense much harder to overlook. The Shaw, Joyce, and Hubbard quotations revolve around marriage and sex (the quotation from Ulysses refers to a brothel). Chandless and Faulkner, meanwhile, follow Greener in using mantrap to mean deathtrap. What emerges is a sense division running along a line quite different to that which Bradley’s entry draws.

Picking through the quotations in this way made us realize that the flaw in Bradley’s reasoning had to do with the meaning of man. As the OED entry for man demonstrates, the various senses of the word fall into two camps, some referring to people generally and others referring to males specifically; these two branches of meaning have coexisted as far back as we can trace the word. The principal distinction to draw in analysing mantrap is not that between figurative and literal uses, but rather the same distinction that is drawn with man itself: something which traps people, and something which traps males. Within these two branches there may well be uses which are more or less literal and figurative, but distinguishing those would be a secondary concern.

It’s interesting to look back and note that the original definition reads ‘A trap for catching men, esp. one for catching trespassers in private grounds.’ Had the drafter put people instead of men—for gins and legholds are undiscriminating, and surely even in 1905 women were known to trespass now and then?—one wonders whether Bradley might perhaps have paused at the Goldsmith and Dickens quotations and been led to reconsider mantrap along the lines of gender.

Just as the two branches of man run in parallel through the historical record—one does not develop from the other—so we can suppose the two branches of mantrap to be parallel and perhaps independent formations. Looking at things this way, there is no reason not to place first that sense which is documented earliest: so first place in the entry can be given to the brass head in The Female Fortune-Teller; and Charles Johnson, derided by Pope as a plagiaristic hack who ‘means not, but blunders round about a meaning’, gets his moment of lexical glory:

Third Edition (2000)

In a very deliberate tightening of Murray’s rules for the first edition, our policy for the third edition is always to respect the chronology of the quotations as we find them: that sense is placed first which the evidence tells us was earliest. The picture that results may sometimes contradict our initial suppositions, but by presenting the quotation record as we have it, and attempting to explain it rather than explain it away, we can ensure that truth in lexicography is more than just opinion, whatever the brass head may say.

At the same time, we have to recognize that pinning everything on the quotation record makes the ordering of each entry provisional to some extent, liable to be overthrown by the discovery of new evidence—an even earlier quotation, for example. It’s always possible that Johnson’s apparent coinage may turn out to be just another of his cribs, and Bradley may have the last laugh after all.