I came across the following passage in a book I was reading this morning:
Did you notice the typo? (And in the title?) Typing that for than seems to be a very common slip. It appears in all sorts of prose, edited (see above, and the fourth paragraph proper here) and unedited. It appears occasionally in my own writing before I fix it. If you Google “bigger that”, “more common that”, etc., and ignore the false positives, you’ll get a hint of the extent of this mistake. Anecdotal evidence further suggests its prevalence.
For such a widespread and apparently simple typo, its cause is rather mysterious. It’s not like typing my name as “Stab” or “Stabn”, which I often do, and which is a simple misstroke resulting from the adjacency of B and N on a QWERTY keyboard and the mechanical imprecision of my typing. T and N are not adjacent, and that-for-than is not an error of omission, duplication, transposition, or repetition. Nor do that and than overlap in meaning. So whence this ubiquitoustypo?
Last year on Language Log, Mark Liberman wrote a very interesting post about typing slips. He included a link to a detailed study (PDF, 1.6 MB) that analysed skilled typing and categorised typos. Its authors, David E. Rumelhart and Donald A. Norman, describe as “capture errors” those that occur ‘when one intends to type one sequence, but gets “captured” by another that has a similar beginning’. Along the same lines, Arnold Zwicky has written about “completion errors“. Typing that instead of than seems to fall into this category, but there are several kinds of capture/completion error, each with its own idiosyncratic and often elusive causes.
Contagion and imitation might be responsible for some wrong thats, especially among learners and non-native speakers, but we can set this aside for now. The words’ strong typographical resemblance to each other is an obvious factor in their confusion. Only in the last letter do they differ; even there, /t/ and /n/ are both alveolar consonants, so they sound quite similar to our mind’s ear. (If you’re unconvinced, say them slowly and notice where your tongue goes.) Both that and than often follow adjectives, which may be another reason the slip slips by: on a cursory glance, this particular wrong word just isn’t conspicuously wrong enough. Since that is a legitimate word — unlike, say, *tham or *thaj — it is more likely to go unnoticed during proofreading.
That-for-than substitution could also have something to do with muscle memory and the commonness of the word that. When fingers type T, H, A in rapid succession, and “expect” to finish the word with one more letter, some kind of functional groove seems to be momentarily adopted, a pattern-recognition program that bypasses Central Cognitive Control and completes the desired word automatically and sometimes erroneously. A lapse is all the more likely because the left forefinger (usually) is already hovering over T and requires little encouragement, less effort, and no lateral shift to drop once more. It does not question the manoeuvre, because it is just a finger, and perhaps a tired one at that.
The comments on Liberman’s post make for an intriguing read. One contributor writes: “I think it’s similar to certain errors I make while playing the piano: my fingers are used to one sequence of movements rather than another, so that’s what comes out when I’m not paying attention.” I have noticed this too. It also seems more likely to occur when I’m tired, or still learning a piece. One hand might take a wrong turn, and if this happens again, it becomes even more likely to happen yet another time. Likewise with all sorts of repetitive behaviour, including certain typos, e.g. Stab. (This phenomenon harks back to what is sometimes called Hebbian learning, after the great Canadian psychologist Donal O. Hebb, and which is often phrased as: “neurons that fire together wire together” — even if the “wrong” combination is firing and wiring.)
While I’m on a binge of jargon, parapraxis is a fancy word for a blunder, what is commonly called a “Freudian slip”, or as the OED puts it, “the faulty performance of an intended action, esp. (psychology) as indicative of a subconscious motive or attitude.” Leaving aside Freud’s probable* over-interpretation, and focusing on our neurological fallibility: our minds play tricks on us all the time — they positively excel at it — often for our own good. But mistakes, when caught, allow us a useful glimpse of the strange machinery within. There is much to be learned from analysing such stumbles, though the that-for-than substitution is a very modest example.
And did I fall foul of Muphry’s Law?
Update: Lots more examples of that for than in my follow-up post, ‘Even stealthier that I thought‘.
* Originally typed as probably!