Slips of the finger vs. slips of the tongue

« previous post | next post »

There's an interesting and understudied way that typing errors and speaking errors are different. From Gary Dell, "Speaking and Misspeaking", Ch. 7 in Introduction to Cognitive Science: Language, 1995:

One of the most striking facts about word slips, such as exchanges, anticipations, perseverations, and noncontextual substitutions, is that they obey a syntactic category rule. When one word erroneously replaces another, most of the time the target and substituting word are of the same syntactic category. Nouns slip with nouns, verbs with verbs, and so on.

In other words, we're NOT likely to say something like "When one word erroneously replacement another, …" or "exchanges, anticipation, perseverations, and noncontextual substituted […] obey a syntactic category rule".

But errors of this type are fairly common in typing. They seem to be cases where we've started to type the right thing, but as our attention shifts to the following material, our fingers follow a familiar but incorrect path.

Here's (what I take to be) a recent example — Christina Cabrera, "Trump Taps Dow Chemical Lawyer To Lead EPA’s Response To Toxic Spills", TPM 2/3/2018:

The White House announced on Friday President Donald Trump’s chosen nominee to run the Environmental Protection Agency’s Office of Solid Waste: Peter C. Wright, a corporate lawyer from The Down Chemical Company.

My own errors of this type are generally toward the ends of longer words, but I can see myself typing "The Down Chemical Company".

The obligatory screenshot:

My ad hoc explanation ("our fingers follow a familiar but incorrect path") makes sense, and is broadly similar to more explicit and better grounded explanations that have been offered for slips of the tongue. But why do our tongue's familiar but incorrect paths generally obey the syntactic category rule, while our fingers' familiar but incorrect paths don't? (At least their keyboard paths — I would predict that sign language errors are  more like speech errors.)

I haven't been able to find any discussion (or empirical evaluation) of this point in the psycholinguistic literature. And in the course of exploring the question, and hoping to present a Breakfast Experiment™ rather than merely a hypothesis, I discovered something else that surprised me — none of the many collections of speech and typing errors are apparently available. There's a (rather antique) search interface at MPI for a collection of speech error datasets, but the interface is useless for exploring this question, and there's apparently no way to get the underlying data.

Update — DM wrote with a link to "The correctly spelled wrong word", 3/4/2018, explaining that "Quite coincidentally, I posted a less academic discussion of pretty much the same thing, a few hours before you did", and noting that "While typing this brief email, I typed 'academy' instead of 'academic'".



  1. Ralph Hickok said,

    March 4, 2018 @ 7:17 am

    A typing error that I have consistently made for as long as I can remember (which is well over 60 years) is adding a "g" to the end of any word that ends in "in."

    Unfortunately, aspiring doesn't relieve the headache that results from such errors.

  2. Vicki said,

    March 4, 2018 @ 8:16 am

    "Down" for "Dow" may not be a violation of, or counterexample to, that pattern. "Dow" is a name, sitting in a context where both adjectives ("International Business Machines") and names, sometimes in adjectival form, are common: Ford Motor Company, American Telephone and Telegraph.

    "Down" can be (at least) a preposition, an adjective ("I'm feeling down this morning"), a noun ("she's riding across the downs"), and a verb (an airplane can be downed, as can a cup of coffee). I speculate that this is why whatever mental "no, that's wrong" checker that often catches this sort of error didn't catch this, even if it would notice a place where "beside manner" was mistyped as "bedside manner."

  3. Tim Leonard said,

    March 4, 2018 @ 8:20 am

    Pardon my thick-headedness. Where is the typing error in the example? It seems grammatically unproblematic to me.

  4. Sidney Wood said,

    March 4, 2018 @ 8:23 am

    "But errors of this type are fairly common in typing. They seem to be cases where we've started to type the right thing, but as our attention shifts to the following material, our fingers follow a familiar but incorrect path."
    Surely attention is in the mind/brain. And straying fingers are obeying the brain. Unless you're fumbling, like hitting CapsLock instead of "a".

  5. Ellen Kozisek said,

    March 4, 2018 @ 8:39 am

    I think that the difference is that typing errors of this sort are a thing of motor memory. It's about control of our fingers, and where they move when. That based on my own experiences of making this sort of error. My thinking is the mental mechanism behind the speech errors is different, and has to do with retrieving the wrong word in the brain, rather than the motor pathways controlling the mouth falling into a familiar pattern.

    I wonder what kind of word substitution errors we see in handwritten writing.

    [(myl) "Motor memory" is a better, or at least more technical-sounding, term for what I had in mind. But this doesn't explain why motor memory involving the fingers should be different from motor memory involving the vocal apparatus.]

  6. Frans said,

    March 4, 2018 @ 8:46 am

    I think I suffer from this in writing much more than in typing. For years I accidentally added " on top of a y when writing in English because in Dutch it's almost always written as a ligature-ish ij, or sometimes I'll accidentally complete an a into a g. And while I don't think I've ever done it, I could much more readily see myself completing dow into down while writing than while typing.

    @Tim Leonard: it's the Dow company, not the Down company.

  7. Steve Anderson said,

    March 4, 2018 @ 9:02 am

    An esteemed former colleague of mine, Japanese, used quite often to interchange r and l … in typing.

  8. Jeff Carney said,

    March 4, 2018 @ 9:25 am

    My college freshmen frequently type "and" where there they mean "an." This strikes me as a case where Mark's ad hoc explanation probably fits.

  9. D.O. said,

    March 4, 2018 @ 10:43 am

    Reading an article in The Economist and found this nice typo:
    Judt issued a dying warning: “We have entered an age of insecurity: economic insecurity, physical insecurity, political insecurity.”

  10. John From Cincinnati said,

    March 4, 2018 @ 10:46 am

    I suppose everybody here besides me is too young to remember writing personal checks. Every new year without fail, many of us would miswrite several checks using the old year's number. Certainly seems a muscle memory phenomenon. I don't think I heard similar substitutions of the spoken year, however, and Mark discerningly wonders what distinguishes the finger motor memory from the vocal motor memory.

    I had, at the same time, an extra muscle memory fault to overcome. My initials are JAB, and during the first month I would sometimes find that I had written my initials instead of JAN.

  11. Robert Coren said,

    March 4, 2018 @ 11:07 am

    Might "Down" for "Dow" be an uncorrected auto-correct?

  12. Matthew McIrvin said,

    March 4, 2018 @ 11:16 am

    With modern technology it's hard to distinguish typos from cupertinos, and indeed there may be no hard boundary between the two.

  13. David L said,

    March 4, 2018 @ 11:16 am

    D.O.: What is the typo?

  14. D.O. said,

    March 4, 2018 @ 11:29 am

    Dying for dire.

  15. David L said,

    March 4, 2018 @ 12:00 pm

    No, it was a warning issued by Judt as he was dying (of ALS, I think).

  16. ja schreuder said,

    March 4, 2018 @ 12:10 pm

    I have the same question as Tim Leonard:

    Pardon my thick-headedness. Where is the typing error in the example? It seems grammatically unproblematic to me.

  17. ktschwarz said,

    March 4, 2018 @ 1:03 pm

    myl: why motor memory involving the fingers should be different from motor memory involving the vocal apparatus

    What about sign languages? Are signing errors more like speech errors (preserving syntactic category) or typing errors?

  18. Garrett Wollman said,

    March 4, 2018 @ 1:24 pm

    Perhaps one salient difference between typing and speech production errors is that we often type open-loop, while looking at some other written material, whereas for most people speech production is inescapably intertwined with hearing. (Which suggests an avenue for investigation: do deaf people make the same kinds of substitutions in speech as hearing people do?)

  19. Jason M said,

    March 4, 2018 @ 1:44 pm

    @ David L – yes, he died of ALS and wrote some poignant columns in NY Review of Books Those dying columns included dire warnings for us who were to continue living as well as being remarkable feats to execute, given how slowly he had to produce text with muscles that barely were strong enough to keep him drawing each breath. I used to wonder about how carefully he must have had to be about each letter. A typo might have set him back hours.

    On the subject of tupos, most of mine come these days from typing on an electronic jeypad where I muss the letter I aumef for.

  20. Bob Ladd said,

    March 4, 2018 @ 4:04 pm

    There definitely is a difference of the sort Mark has identified. I think it somehow has to do with the fact that speech errors are made "higher up" in the planning process, which is why they often respect part-of-speech categories, etc. Typing errors seem to operate at a lower level – you have already planned what you want to say, and the error arises in converting the intended words into a sequence of keystrokes. And that's why well-rehearsed keystroke sequences can intrude without regard to part of speech (my own personal example is frequently typing the word glad as gladd).
    However, there's at least one kind of typing mistake I've observed that does kind of respect higher-level factors, namely cross-linguistic typos usually involving function words. If I'm typing in another language, I often type and instead of German und or French et, and I've received emails in English written by Dutch speakers where English or is replaced by its Dutch equivalent of. I've observed some others of the same sort but can't think of them now. These would seem to involve a process more like speech errors.

  21. Frans said,

    March 4, 2018 @ 5:27 pm

    @Bob Ladd

    I've received emails in English written by Dutch speakers where English or is replaced by its Dutch equivalent of.

    I do that occasionally. I'll definitely be thinking or, but my fingers secretly type of. It's different from a regular typo because those tend to enter my conscious mind, even if sometimes only after correcting it. The stray ofs tend to stand out on a quick read-through though.

  22. Tim Leonard said,

    March 4, 2018 @ 5:56 pm

    Thank you, Robert Coren. I now see the substitution of "Down" for "Dow".

  23. Ralph Hickok said,

    March 4, 2018 @ 6:42 pm

    @John From Cincinnati:
    I still write quite a few personal checks!

    Your post jogged my memory of a couple of laughable check-writing errors. I once stopped in at my bank to cash a check on my birthday…and I wrote my actual birthdate, November 9, 1938, which caused the teller to do a triple or perhaps even quadruple take before she politely pointed out my error and asked my permission to write in the correct year.

    Just a few weeks ago, I had to write a check for $76.00 and, on the line where you spell out the amount, I wrote "Seventy-Six Trombones" :)

  24. Jen in Edinburgh said,

    March 4, 2018 @ 7:31 pm

    John from Cincinnati: My mum was writing a cheque for a cousin of mine with a January birthday, so she carefully checked that she had written 18 rather than 17, and checked again – only to have the bank reject it because she had dated it January 1918!

  25. John Roth said,

    March 4, 2018 @ 8:19 pm

    To Mark and Bob Ladd

    Speech and finger manipulation occupy different parts of the motor cortex. In fact, I'm not at all sure that the coordination required to speak is part of the motor cortex at all, although some pieces certainly are (like the lips and the tongue).

    The other point is that speech is a lot older than typing. In fact, speech is a lot older than alphabetic writing, and spelling errors are a very common problem, especially with English.

  26. TIC said,

    March 4, 2018 @ 9:44 pm

    Off topic, I know, but this reminds me of a mistake I made so often when editing and tweaking text I'd just keyed that it made me suspect there must be some (common?) brain-wiring glitch involved… After positioning the cursor and then keying a word I wanted to insert, I'd all-too-often realize that I'd accidentally placed the insertion *after* the word that I'd intended to place it before… Often — but not always — this was the mis-insertion of an adjective after the noun… And no, sadly, I have no familiarity with languages where that's the norm… Always wondered whether this was a unique(ish)-to-me phenomenon…

    Also, adding an inappropriate past-tense 'd' at the end of present-tense verbs ending in -de (such as decide and provide) was an all-too-common error for me… Don't know how widely shared this one might be, either…

  27. Anne Cutler said,

    March 4, 2018 @ 11:33 pm

    Motor memory indeed. The typical errors I persistently make in handwriting do not overlap with the typical errors I persistently make in typing.
    MYL – apologies for whatever responsibility I bear for the state of the speech error database at MPI. It partly reflects the fact that there was supposed to be a much larger database at UCLA of which the MPI one was a subset. I take it that searches now don't reveal the UCLA one either. I'll see what I can find.

  28. rosie said,

    March 5, 2018 @ 2:07 am

    @TIC Typing "decided" instead of "decide" is like a sort of typo I often make. I often find that I've typed "eset" instead of "est", or "thte" instead of "the". I don't believe that an involuntary tremor caused a finger to move up and and down again and press its key a second time. Perhaps a bug in how the keyboard's firmware interprets key rollover?

    Naturally typing errors are different from speech errors — the text must be typed one character at a time.

  29. maidhc said,

    March 5, 2018 @ 2:59 am

    I just noticed one today. "Where" instead of "we're".

    The one I make the most is "the" for "to". A remnant of my long-ago Greek studies?

    I didn't learn touch-typing in school. I did a computer course when I was in my thirties because I realized I was typing so much. I'm fair at it. There are some words I consistently mess up, like "archictechture" ("architechture"?) For some reason my fingers think there must be "tech" in there.

  30. Jon said,

    March 5, 2018 @ 4:41 am

    Many years ago, just after being promoted to the grade of HSO (higher scientific officer) I had to write my grade on a form, and wrote H2O.

  31. Rubrick said,

    March 5, 2018 @ 6:15 am

    @myl: But this doesn't explain why motor memory involving the fingers should be different from motor memory involving the vocal apparatus.

    This strikes me as a deceptively profound observation. Somehow, speech doesn't "feel" at all like muscle memory, even though that's clearly what it is. And I'd back your hunch that sign-language errors will mimic speech errors rather than typing errors.

  32. Ursa Major said,

    March 5, 2018 @ 6:38 am

    Your explanation, which Ellen Kozisek expands on as motor memory, is what I would have proposed. Anyone who performs repetitive physical tasks will eventually reach a point where they can do the actions without thinking, and it takes concentration to modify them. Athletes specifically regularly practise drills so they keep the correct form when exhausted or otherwise distracted.

    The only example of this typing error I'm aware of, which I do every time I type the word, is "ration" for "ratio". The ending -ation is just so common my fingers move through the pattern without being able to stop early.

  33. Jen in Edinburgh said,

    March 5, 2018 @ 7:49 am

    My main one is adding an extra 't' to the end of any word ending 'gh'.

    The wrong date in the new year thing also happens to teachers writing the date on the board and students writing the date in their book/notes, but it's hard to tell if this is muscle memory or a genuine memory slip – that is one overlap case where people might also go on mentioning the year as 2017.

  34. Laura Payne said,

    March 5, 2018 @ 8:25 am

    Back in 2010, on a blog I no longer update, I posted about some bizarre typing errors I noticed in my own typing. My typos were the exchange of voiced for voiceless consonants and vice-versa. I still find myself doing this today, and I still have no explanation as to why this would happen.

  35. Joyce Melton said,

    March 5, 2018 @ 10:21 am

    There's a typing error I frequently make that I have not seen discussed here. My ring fingers do not seem sure which hand they are connected to. I frequently type s when I mean to type l, and o when I mean to type w.

    Note the direction of the mistakes, the more common letter is substituted for the less common one, usually but not always. Swmetimes it goel the other way round.

    If the mistake happens to spell an actual word, I may not notice. It's easy enough to confuse too and two without this happening, so I put a little extra effort into proofreading for that particular error.

  36. Thomas Rees said,

    March 5, 2018 @ 8:25 pm

    The link in the OP update has lost a leading “h”.

  37. Jerry Friedman said,

    March 5, 2018 @ 8:51 pm

    Bob Ladd: I often type of for or, and I don't speak Dutch.

  38. Phil Jensen said,

    March 10, 2018 @ 1:27 am

    In the update to the o.p., the link to Never Pure … is missing its initial "h" (it starts with "ttps://..").

    I was recently looking at my copy of the hardcover of Hesse's The Glass Bead Game (the Winston translation, 1969) and noticed the following passage (starting line 5, page 480):

    "Instead of a demoralized horde of madmen, there now stood a reverent populace prepared for sacrifice and penance, each one benefiting, each one encouraged by now having to lock his horror and fear of death within himself, or bellow it crazily for himself alone. Each now fitted into his place in the orderly chorus of the multitude, keeping to the rhythm of the exorcistic ceremony."

    Note that there are three occurrences of "now", but the second is a typo for "not". I suggest that the upcoming legitimate "now" interfered in the typesetter's mental pipeline.

  39. Chas Belov said,

    March 10, 2018 @ 1:55 am

    For most of my life I have occasionally mixed up "t" "d" and surprisingly "s" when hand printing.

  40. Luke B said,

    March 13, 2018 @ 4:52 am

    Implicit in many comments, but worth spelling out (so to speak), is that the higher-level syntactic-category-respecting errors may well exist in typing, but rarely or never make it to the realization stage. This would be presumably be due to more processes and presumably more time intervening between lexical selection and motor realization in typing than in speech, thereby giving the error monitor more time to do its thing at the relevant higher level.

RSS feed for comments on this post