Word Weirding

« previous post | next post »

Rick Rubenstein asks

Has there been any research done on the familiar phenomenon wherein a word which is repeated over and over begins to look misspelled, or even like complete gibberish?

Yes, lots. The phenomenon was described by Edward Tichener in A Beginner's Psychology, 1915:

Repeat aloud some word  — the first that occurs to you; house, for instance — over and over again; presently the sound of the word becomes meaningless and blank; you are puzzled and a morsel frightened as you hear it.

The earliest serious study known to me is M.F. Bassett and C.J. Warne, "On the Lapse of Verbal Memory with Repetition" (Minor Studies from the Psychological Laboratory of Cornell University XLIV), The American Journal of Psychology 30(4) 1919:

It is well known that if a familiar word be stared at for a time, or repeated aloud over and over again, the meaning drops away. In this paper we report the result of experiments whose aim was the determination of the number of repetitions required for monosyllabic nouns to lose their meaning.

The terms "verbal satiation" and "semantic satiation" came to be used to refer to this loss of meaningfulness. Thus Donald Smith and Alton Raygor, "Verbal satiation and personality", Journal of Abnormal and Social Psychology, 52(3) 1956:

Satiation may be described as the reduction in effectiveness of a stimulus with continued exposure. The concept has been invoked to account for a number of diverse phenomena, loss of word meaning with repetition, visual alternation during fixation of ambiguous figures, boredom, alternation behavior of rats in a T maze, and diagnosis of brain lesion.

Wallace Lambert and Leon Jakobovits proposed a clever way to measure (and to conceptualize) this change "(Verbal satiation and changes in the intensity of meaning", Journal of Experimental Psychology 60(6) 1960):

Several investigators have demonstrated that Ss experience a change or loss of meaning for words which have continuously repeated or fixated for a certain period of time. For instance, Bassett and Warne (1919) reported that the meanings of familiar nouns which were repeated aloud "dissipated" for their Ss within 3 or 4 sec. More recently, it was found that if Ss fixated a word exposed on a screen for 20 sec., their first association to the word is uncommon as measured by the Kent-Rosanoff Word Association Test (Smith & Raygor, 1956). [...] Before the ful implication of this concept for theories of learning and meaning can be determined, it is necessary to develop a method for reliably measuring the extent of meaning change which be attributed to the satiation experience. [...]

Osgood, Suci, and Tannenbaum (1957) have proposed an objective and reliable instrument for the measurement of certain aspects of connotative meaning. This instrument, the "semantic differential," consists of a series of scales each representing a 7-point, bipolar dimension. The meaning of a word, such as "father," is given by its position on an evaluative factor (its degree of goodness or badness), on an activity factor, and on a potency factor. The theory underlying this method assumes that the meaning of a word has a place in a multidimensional semantic space. A word without meaning would rest at the point of origin for all dimensions. [...]

In the present studies, the semantic differential is used as a method of indexing changes in meaning induced by means of verbal repetition of words.

They found that after saying a word aloud 2-3 times per second for 15 seconds, subjects' estimates of its position in this semantic space moved to a point nearer the origin. A corresponding period of silence had almost no effect, and repeating a different word had a much smaller (and not statistically significant) effect.

There's a different, but perhaps related, effect that was first described by Richard Warren. In "Verbal transformation effect and auditory perceptual mechanisms", chapter XVIII in Norman Lass, Ed., Speech and hearing science: selected readings (Volume 10), 1974, he wrote:

If, rather than repeating a word to oneself, one listens to a recording of a word repeated over and over, a different effect occurs. Abrupt illusory changes are experienced, frequently involving considerable phonetic distortion. It is usually difficult to believe that the stimulus is not changing, even when the listener knows that all repetitions are identical. This "verbal transformation effect" has revealed some rather unexpected characteristics of speech perception and has suggested mechanisms underlying the perception of connected discourse.

Here's his description of his first experiment:

The first systematic study of the VT effect employed British sailors as subjects (Warren, 1961a). Stimuli were prepared by recording single vocal statements which were then spliced to form a loop of tape. Each loop was played back and rerecorded on a conventional reel to give a sitmulus which could be played for 3 minutes. [...]

The subject listened to the repeated stimulus through headphones, tapping on the table and calling out what he heard as soon as he could [...]

An example of one subject's responses to [...] TRESS, played loudly and clearly for 3 minutes is as follows (all responses listed in the order reported): "stress, dress, stress, dress, Jewish, Joyce, dress, Jewess, Jewish, dress, floris, florist, Joyce, dress, stress, dress, purse."

You can find plenty of other research reports by searching Google Scholar for "verbal satiation", "semantic satiation", or "verbal transformation effect". A small and somewhat random sample of the hits:

Lee Smith and Raymond Klein, "Evidence for semantic satiation: Repeating a category slows subsequent semantic processing", Journal of Experimental Psychology: Learning, Memory, and Cognition, 16(5) 1990.

M Pilotti et al., "The effect of presemantic acoustic adaptation on semantic 'satiation'", Mem Cognit 25(3) 1997.

T Ditzinger, B. Tuller, and J.A.S. Kelso, "Temporal patterning in an auditory illusion: the verbal transformation effect", Biological Cybernetics 77(1) 1997.

Mark Pitt and Lisa Shoaf, "Linking verbal transformations to their causes", Journal of Experimental Psychology: Human Perception and Performance 28(1) 2002.

Anahita Basirat et al., "Parieto-frontal gamma band activity during the perceptual emergence of speech forms", Neuroimage 42(1) 2008.



  1. Gregory Dyke said,

    December 8, 2010 @ 6:30 am

    I get this a lot when I'm programming, using a statically typed language. Most typically on the words String and Tree (names of types variables are declared as), which suddenly get me "puzzled and frightened" as the first quote so accurately described.

  2. Bob Couttie said,

    December 8, 2010 @ 7:06 am

    Indeed it is part of some religious practices

  3. Scott Y said,

    December 8, 2010 @ 7:33 am

    I get it with 'taught'. Demon child of a word.

  4. Twitter Trackbacks for Language Log » Word Weirding [upenn.edu] on Topsy.com said,

    December 8, 2010 @ 7:36 am

    [...] Language Log » Word Weirding languagelog.ldc.upenn.edu/nll/?p=2833 – view page – cached December 8, 2010 @ 6:02 am · Filed by Mark Liberman under Psychology of Tweets about this link [...]

  5. Randall said,

    December 8, 2010 @ 8:34 am

    Ah, my favorite (read: most detested) example of the verbal transformation effect is the youtube video "Navi: Insane Edition".

  6. Kapitano said,

    December 8, 2010 @ 8:36 am

    The 'satiation' effect – or something like it – also occurs with loops of music, but the 'transformation' effect does not.

    Satiation is a big problem in music production, where you might hear the same recording hundreds of times – it quickly gets to the point where you can't tell whether an instrument is too loud, or even off-key.

    I suspect if the transformation effect applied to music, all the religious and cultural practices which rely on musical repetition simply wouldn't work.

  7. Telofy said,

    December 8, 2010 @ 9:01 am

    Hecko native speakers,

    how do you perceive the “presently” (in the sense of “soon”) in the second excerpt?

    An Irish native speaker I know, while she was perfectly aware of the two senses of “presently,” perceived it as odd to use it the way I did here: “Once the freshmen have grown accustomed to university routine, the momentum of their academic progress presently wanes, …” The Random House dictionary writes “The two senses are rarely if ever confused in actual practice. Presently meaning ‘now’ is most often used with the present tense (The professor is presently on sabbatical leave) and presently meaning ‘soon’ often with the future tense (The supervisor will be back presently).”

    From this I concluded that there had to be a surprisingly clear cut complementary distribution of the meanings—as it were—between present and future tense (or the modal constructions denoting futurity). Or is “wane,” in this regard, different from “become” in some way?


    [(myl) This is one of many, many examples, in every language, of a polysemous word that must be disambiguated in context. (See here for some earlier discussion.) The only slightly unusual thing here is that one of the senses is archaic, and so its use (as in this case) is more likely to be noticed. Please don't add 50 comments giving your own re-phrasing of the dictionary gloss that this commenter has already looked up.]

  8. J. Goard said,

    December 8, 2010 @ 9:08 am


    You had me at "hecko".

  9. Adrian Morgan said,

    December 8, 2010 @ 10:14 am

    I have never been able to get such a thing to work for me. I experience no remarkable psychological effects, no matter how many times I repeat a word. The closest I get is a drift of attention, which is unremarkable. Attention drifts. So what?

  10. Stephen R. Anderson said,

    December 8, 2010 @ 10:58 am

    A variant of this affecting linguists specifically was once christened (I believe) by Barbara Partee as "scanting out." As I understand it (I got there just too late to be able to speak first hand), back in the 1960s linguists from the MIT Department were paid to work in the summer at MITRE, where they were supposed to study aspects of English syntax. Barbara is said to have found that after looking at a rather small number of sentences involving the word "scant" her judgments totally dissolved.

  11. David L Rattigan said,

    December 8, 2010 @ 11:14 am

    Fascinating. I've always been intrigued by the way words start to look completely foreign the more you look at them. Didn't occur to me that anyone had seriously studied the phenomenon.

  12. Jonathan Mayhew said,

    December 8, 2010 @ 11:56 am

    I get the "scanting out" effect often reading language log, where thinking too hard and long about whether something is grammatically acceptable to me turns my powers of judgment to mush.

  13. David L Rattigan said,

    December 8, 2010 @ 11:58 am

    I wonder how this phenomenon is related to that of loanwords from other languages appearing "normal," even though on analysis they're aberrations, eg, "shampoo," "trek," "amok."

  14. Chris said,

    December 8, 2010 @ 1:22 pm

    I just read the opening paragraphs of the Pitt & Shoaf paper on VTE and it occurs to me that this is a separate phenomenon from experiencing a familiar word as unfamiliar, but not through repetition (they used 2wrds/sec). For me, this other experience occurs most commonly when I'm bored and suddenly the word house sounds odd. I say it aloud and it is unfamiliar, even funny sounding. This is far more difficult to study systematically since it's difficult to re-create the experience on command. Could it be that boredom correlates with low brain activity and the network associated with recognizing house just can't quite muster enough activation to fire…this sounds like serious neuro-BS now that I've written it, ignore that last part…

  15. Xmun said,

    December 8, 2010 @ 1:45 pm

    I remember lying in bed one night at the age of about eight and pondering what a very peculiar word "and" was. Just its spelling was what interested me. I had no notions then of word use. For some reason the spelling "and" struck me as a weird array of letters.

    Why do I still remember this some 66 years later?

  16. Barbara Partee said,

    December 8, 2010 @ 2:36 pm

    @Stephen — I had forgotten about 'scanting out', but it sounds like me and it rings true, so I'll take it if no one challenges your memory!

  17. Chandra said,

    December 8, 2010 @ 2:37 pm

    I do this with my own name sometimes, which produces a very weird psychological effect because not only does the series of phonemes no longer make any sense, but I start to feel sort of disconnected from my own sense of self too.

  18. MB said,

    December 8, 2010 @ 3:50 pm

    "Who is Kim–Kim–Kim?" Rudyard Kipling, 1901.

  19. Aaron Toivo said,

    December 8, 2010 @ 3:57 pm

    Is this effect related to the one where, in reading any sort of text (but especially heavy material), you suddenly find you did not glean any meaning from a sentence and are forced to re-read it several times before it finally comes clear (which it doesn't always even do)? I'm not talking about passages you wouldn't understand in any case, but those times where your semantic engagement with the material unexpectedly goes on the blink.

    When I'm caught in one, I've found I can still judge the grammaticality of the sentence, I just can't tell what it means! The mind either struggles to make anything of it, or resolves a few referents but quits there, no proposition formed. That loss of semantics is why I wonder if this effects is related to verbal satiation.

  20. J Lee said,

    December 8, 2010 @ 4:55 pm

    I obeyed and repeated 'house' aloud rapidly (2-3x/s). Predictably the H didn't survive its adjacency to the fricative and I almost immediately recognized the S as the onset of the ridiculously low-frequency "sow." So perhaps rather than the first words that occur to us our examples should only be ones whose phonetic forms aren't distorted by repetition (a monosyllabic English word ending in a velar nasal?). More to the point, though, I think it is not sheer repetition but simply an appreciation of/mild fascination with the arbitrary sound-meaning correspondences that sort of temporarily bleaches a word — obviously even during a deep introspective episode on [haus] it will be a perfectly accessible and practical lexeme.

  21. Clayton Burns said,

    December 8, 2010 @ 5:33 pm

    Perhaps my comment is ellipsoidal to the intent of the post, but I would say that you have it backwards. The impression that words make on their face that they are coherent is an illusion.

    If words are real (which they are not), their meanings are enantiomorphic to what we see. At best.

    I wonder who arranged it so that it seems to us as if words make sense?

    Descartes was onto it, but lost the trail.

  22. ~flow said,

    December 8, 2010 @ 6:47 pm

    i once had an epiphany when i was maybe four, five or six years old: back in our home i stood in the alcove of our flat in hb for some reason, looking at my beloved roudy ™ anorak for some reason, when it occurred to me this piece was actually called anorak. anorak..? what could that mean? ano- meant nothing to me, -rak meant nothing to me, and there was just this one, anorak, anorak from the karstadt department store, the anorak, ano? rak? ano? rak? rak? ano? ano? rak? anorak? anorak?anorak? anorak?anorak?anorak? anorak?anorak?anorak? anorak?anorak?anorak???????…… got me completely mezmeryzed. beautiful anorak, let's go out into the autumn winds…

  23. Jeff R, said,

    December 8, 2010 @ 7:45 pm

    Occasionally, I become utterly unable to accept that the word 'school' is spelled correctly. Is this the same phenomenon, or a related one? It doesn't always require that much repitition

  24. Margaret L said,

    December 8, 2010 @ 8:51 pm

    It seems to me this is a straightforward example of detector fatigue, similar to the waterfall effect, or spatial frequency or orientation aftereffects. The comparison to alternation of percepts for ambiguous figures is also spot-on. (The comparison to alternation behavior in T-mazes is just bizarre.)

  25. Troy S. said,

    December 8, 2010 @ 9:47 pm

    Space Ghost did this with the titular word in "Punch @7:19".

  26. Clare said,

    December 8, 2010 @ 10:33 pm

    Exploited to great effect also in Little Britain:

    "Dust. Anyone? Anyone? Dust. High in fat, low in fat? Dust. Anyone? Dust."


  27. Jon said,

    December 9, 2010 @ 4:40 am

    And there was the famous review in Melody Maker, when the reviewer was given a pre-productiion copy of John Lennon's 'Wedding Album', that had a constant test tone on the B side. Despite the absolute constancy of pitch, he heard it as varying, and reviewed it.

  28. Breffni said,

    December 9, 2010 @ 5:31 am

    Arnold Zwicky did a Language Log post on "scanting out" (in which he attributes the term to Haj Ross). He describes "scanting out" on the "seem fit" construction, but by the end of the post I had the satiation experience with the word "fit". The way it struck me was not so much that I had lost my grip on the word's meaning, as that I had become unusually aware of its sound.

    (Off-topic, but Arnold says in that post that "scant" "can't be used predicatively". But evidence of "evidence is scant" isn't scant. You can also find "rainfall is scant" and "money is scant", and there are other examples in the OED. And now I've got "scant" satiation.)

  29. Nick said,

    December 9, 2010 @ 7:10 am

    I got smocked by Hobbes.

  30. Peter Sattler said,

    December 9, 2010 @ 7:33 am

    I assume that this phenomenon has been noticed for millennia and is at the heart both of meditative mantras and of Gertrude Stein's poetry. However I do wonder when it was first described or noted in a serious way.

    Wassily Kandinsky uses the satiation effect to describe what he sees at work in abstract painting, which evacuates the representational meanings of images and allows what he calls the spiritual and emotional content of the form to come through:

    The apt use of a word (in its poetical meaning), repetition of this word, twice, three times or even more frequently, according to the need of the poem, will not only tend to intensify the inner harmony but also bring to light unsuspected spiritual properties of the word itself. Further than that, frequent repetition of a word (again a favourite game of children, which is forgotten in after life) deprives the word of its original external meaning.

    This is from 1910's Concerning the Spiritual in Art. Are there any 19th-century (or earlier) considerations?

    (I'm sorry that I cannot access the PsycNET links to check this myself.)

    [(myl) The quote from Kandinsky is interesting -- thanks for that!

    Psychology didn't really exist as an experimental science until the second half of the 19th century -- Wundt founded his lab in Leipzig in 1879; James Cattell founded the psychology lab at Penn in 1887; Titchener (who studied with Wundt) started the psychology lab at Cornell in 1892. Before that, experiments in what we would now call psychology were carried out by physicists, physicians, or general polymaths. Though verbal "satiation effects" are interesting, they're rather far down on the list of phenomena that are likely to have seized the attention of such people.

    So the existence of such effects may very well have been noted briefly by some ancient Sanskrit, Greek, Latin, Chinese or Arabic writer, but I would be surprised to find any systematic studies in earlier times.]

  31. Alix said,

    December 9, 2010 @ 10:36 am

    Have experienced this phenomenon personally– also a phenomenon wherein a particular word starts to 'look' like what it describes: "angry" begins to 'look' peeved, if not outraged; and "hungry" simply cries out to be fed.

  32. Ellis said,

    December 10, 2010 @ 9:24 am

    'Have you ever tried the experiment of saying some plain word, such as "dog," thirty times? By the thirtieth time it has become a word like "snark" or "pobble." It does not become tame, it becomes wild, by repetition. In the end a dog walks about as startling and undecipherable as Leviathan or Croquemitaine'. – G.K. Chesterton, 'The Telegraph Poles', which was first collected in 1910 (Alarms and Discursions); the first publication would, knowing GKC, have been not much earlier.

  33. Ray Dillinger said,

    December 11, 2010 @ 3:22 pm

    It seems to take much more than 30 repetitions for me. When fully awake and alert, I don't lose track of a word's meaning or correctness in the first half-hour of repeating it. In an altered state of consciousness (meditation, hypnosis, etc) it seems to come on nearly spontaneously.

    If I start by convincing myself that it'll happen, it happens faster or more easily. I wonder how much of this is confirmation bias or self-hypnosis of the experience happening because the people expect the experience.

  34. Fergus said,

    December 13, 2010 @ 12:24 am

    Silent repetition of one's own name. Here is a footnote from the 'Mysticism' chapter of William James' The Varieties of Religious Experience. I suppose that the repeated word was 'Alfred'…

    "I have never had any revelations through anaesthetics, but a kind of waking trance – this for lack of a better word – I have frequently had, quite up from boyhood, when I have been all alone. This has come upon me through repeating my own name to myself silently, till all at once, as it were out of the intensity of the consciousness of individuality, individuality itself seemed to dissolve and fade away into boundless being, and this not a confused state but the clearest, the surest of the surest, utterly beyond words – where death was an almost laughable impossibility – the loss of personality (if so it were) seeming no extinction, but the only true life. I am ashamed of my feeble description. Have I not said the state is utterly beyond words?"
    Professor Tyndall, in a letter, recalls Tennyson saying of this condition: "By God Almighty! There is no delusion in the matter! It is no nebulous ecstasy, but a state of transcendent wonder, associated with absolute clearness of mind." Memoirs of Alfred Tennyson, ii. 473.

  35. John said,

    December 13, 2010 @ 1:32 am

    Back in the '60s, while teaching TOEFL in Bangkok, a class repetition exercise–"The nice witch washed the fresh bats first"–sent me for a loop on about the 20th cycle. After that, I found a variety of words would trigger a strange state.

  36. Chandra said,

    December 13, 2010 @ 7:20 pm

    @Fergus: That is indeed an interesting description of the sensation produced.

    For curiosity's sake, I tried doing this with a word from a language I learned later in life ("hermano", Spanish for "brother"). Interestingly, no matter how many times I typed out the word or repeated it in my head, I could not get it to disassociate from its definition. On the other hand, once I repeated the exercise with "brother", it started looking and sounding foreign to me after only a few repetitions.

  37. brb; being a chrome extension junkie. « company of three, black peppermint tea said,

    December 17, 2010 @ 10:50 am

    [...] [article] Word Weirding The first systematic study of the VT effect employed British sailors as subjects (Warren, 1961a). Stimuli were prepared by recording single vocal statements which were then spliced to form a loop of tape. Each loop was played back and rerecorded on a conventional reel to give a sitmulus which could be played for 3 minutes. […] [...]

  38. baylink said,

    December 17, 2010 @ 11:24 am

    I call this Pepsodent Syndrome (for the inferential reason) and noticed it at age 12 or so.

  39. George said,

    December 18, 2010 @ 2:09 pm

    This reminds me of what sometimes happens when I try to define a word for someone. The definition requires the use of other words which in turn call for definition. It seems that process always leads back to the original word…so the definition of the original word ends up being the same word. Yikes!

  40. CONSCIOUSNESS « Cole Johnson said,

    June 29, 2011 @ 3:48 pm

    [...] Some of the difficulties in this exercise could be attributed to the phenomenon of semantic satiation but I think it neatly illustrates the contradictions of arriving at an understanding of [...]

  41. Why are tongue twisters tricky? (Part 2) | said,

    May 21, 2012 @ 2:23 pm

    [...] (Is lexical access starting to sound like a fake word yet? That process, by the way, is known as semantic saturation.) [...]

RSS feed for comments on this post · TrackBack URI

Leave a Comment