Release notes: words in ‘un’ – patterns and surprises
Revising words beginning with un– has brought up one particular issue many times: sometimes the evidence that we have for the “negative” word beginning with un– is earlier than the evidence that we have for the corresponding “positive” word. Investigating these more closely brings some interesting issues to light.
We chose to work on “un-words” that correspond to parts of the alphabet already revised for OED3, so all of the corresponding “positive” words have been edited as part of the dictionary’s revision programme. In many cases, the entries for the corresponding “positive” words were edited some years ago (it is now 14 years since OED3 entries at the beginning of the letter M were first published online), and working on un– now has prompted us to take another look at whether some of these words can be antedated or otherwise revised in the light of resources (especially electronic text databases) that have either become available or been improved since the OED3 entries were first published. We frequently update the documentation for already-revised OED3 entries as new evidence comes to light (including work that is prompted by information from readers), but the concentrated work on un– in this quarter’s release has prompted an unusually high number of such changes..
To take a couple of examples:
Our earliest evidence for macadamized in OED3 (revised entry published in 2000) was from 1827, but researching unmacadamized brought to light a first example from 1826:
1826 Literary Lounger Jan. 45 Heavy and unmacadamized is the road that leads to literary attainment.
This prompted us to make a fresh search for earlier evidence for macadamized, which turned up in an 1823 example (identified through an online database of historical newspapers):
1823 Morning Post 2 Dec., The unsightly buildings at the east end‥have been removed, and a very handsome avenue formed, leading from Pall-mall to the Park, with a Macadamized road.
This is in turn slightly earlier than OED3’s existing earliest example of macadamize (from 1824), so we made a fresh search for this too, and found it earlier in 1823 in the same newspaper:
1823 Morning Post 23 July, We are informed it is finally arranged that the Pavement of St. James’s-Square is to be immediately taken up, and the whole Macadamized.
In some cases the changes have been much more dramatic. For instance, working on unmated meaning “having no mate, match, or pair” (OED3’s unmated adj.2) made us curious about the evidence we had for both mated and the verb mate with reference to animals pairing, copulating, or being paired, and this has prompted a number of quite large changes at both entries. Perhaps the most striking is at what was sense 5a of mate v.3, “Of birds and other animals: to pair for breeding. Of an animal: to copulate with a mate.” The earliest example we had of this was from 1853:
1853 E. S. Dixon in Househ. Words 14 May 248/2 The perverse pair of pigeons..seem determined to ‘mate’ according to their own, rather than in obedience to Bob’s ideas of a suitable match.
But fresh research using online databases that were unavailable at the beginning of the 2000s has enabled us to provide a set of examples going right back to the late 1500s (and to re-order this sense as 2a in the chronological sense sequence at mate v.3):
1581 J. Maplet Diall Destiny f. 57, So is ye Doue‥, whose kindnes to eche other as they bee matched or Mated together, (as it were in a certaine forme of Wedlocke) euery dayes experience doth teache vs.
1607 J. Day et al. Travailes Three Eng. Brothers sig. F3v, The Princely Lionesse disdaines to mate But with a Lyon.
1695 H. Killigrew tr. Martial Epigrams xii. 290 She me prevented with an amorous Kiss, Such as soft Doves, when mated, make their Bliss.
1785 Scots Mag. Mar. 145/2 Mating with his feather’d fair, She alone is all his care.
1789 R. Brookes Art of Angling (ed. 7) 148 They pair as Birds do; and as soon after they are mated as the Male Fish can find a proper Place,‥he, and not the Female,‥forms the Hole or Nidus in which the Spawn is to be deposited.
In other cases, the “un-word” remains earlier than the “positive” even after such searching has been carried out.
Sometimes the gap is small. For instance, maintained has recently been antedated from 1571 to 1536, but this is still three years later than our earliest evidence for unmaintained (1533).
Sometimes the gap is larger. unmountable c1550, mountable 1596; unmeasured a1398, measured 1440; unmeaningness has a first quotation dated 1728, meaningness 1753; and the list could be extended very considerably.
In some cases we may suspect that the difference in date is no more than an accident of the historical record, although in nearly all cases it is quite possible that the “un-word” really did exist earlier, and this is what is normally assumed in the etymologies given for these words. For example, unmountable is etymologized directly from un-, mount verb, and –able; unmeaningness is taken to show suffixation of unmeaning in –ness; and so on.
Finally, not all “un-words” are in a particularly close relationship with a “positive” counterpart; this can be the case even when both words exist, and have OED entries. The adjective unmistakable is first attested a1643, mistakable 1646; the difference in date is very small, and could be no more than an accident of the historical record (as reflected by our currently available resources). However, closer inspection brings several interesting things to light. Firstly, unmistakable is a far more common word than mistakable, and this is reflected by a greater extent and range of illustration in OED. Its usual modern sense “that cannot be mistaken for something or someone else; clearly recognizable; obvious; utterly distinctive or characteristic” is recorded from 1822 onwards; it could have arisen from mistakable, “able to be mistaken, misapprehended, or misunderstood”, which existed at this date, but even if so it has long since shaken off the shadow of its positive counterpart. The now-obsolete first sense of unmistakable is “that cannot be mistaken; infallible”; it is this sense that is first recorded a1643, and it does not correspond to any sense recorded for mistakable. The fact that unmistakable and mistakable are first recorded so close to one another in date is interesting, but the two words have very distinctive histories, and unmistakable would be very poorly summed up as “the negative of mistakable”.