10
$\begingroup$

I was really surprised how fast I was able to read and comprehend following text:

I cnduo't bvleiee taht I culod aulaclty uesdtannrd waht I was rdnaieg. Unisg the icndeblire pweor of the hmuan mnid, aocdcrnig to rseecrah at Cmabrigde Uinervtisy, it dseno't mttaer in waht oderr the lterets in a wrod are, the olny irpoamtnt tihng is taht the frsit and lsat ltteer be in the rhgit pclae. The rset can be a taotl mses and you can sitll raed it whoutit a pboerlm. Tihs is bucseae the huamn mnid deos not raed ervey ltteer by istlef, but the wrod as a wlohe. Aaznmig, huh? Yaeh and I awlyas tghhuot slelinpg was ipmorantt! See if yuor fdreins can raed tihs too.

Does this phenomenon relate to the language structure, i.e. is specific to English?

Would trick works for languages with longer words, like German?

Does the same phenomenon exist for Asian, non-Latin alphabet based type languages like Mandarin?

UPDATE:

I have tried to scramble longer words as suggested in answer below and it becomes more difficult to read: "Ceefrconne of iantrintaeonl enatldocatiiuss" (Conference of international educationalists).

Moreover, if letters just pair-wise permuted - it easier to read than random permutation. My initial feeling is that it should be certain metric that define closeness(distance) of the words to its right spelling. Then "easiness" and speed of comprehension will be proportional to that distance.

$\endgroup$
  • $\begingroup$ As for the last question, I have seen such an example in Chinese. In words which were formed by two or more characters, their order was swapped. I do not remember whether characters outside of a single word context changed places, but anyhow, the text remained completely readable. $\endgroup$ – user8304 May 11 '15 at 14:36
  • $\begingroup$ Regardless of this being an interesting question, it (and the text posted in this question) seem to be an urban legend/internet meme called typoglycemia. It is unclear what the purported 'research at Cambridge university' was, but it could have been an unpublished PhD. $\endgroup$ – Steven Jeuris Feb 9 '16 at 22:55
8
$\begingroup$

The scrambled words game is very useful in persuading the less sophisticated to take a passing interest in their own cognitive processes! it is intriguing and also rewarding as it shows we can do something apparently rather difficult more easily than expected. However, the difficulty will be greater for second language learners at earlier stages of study.

Thinking about the unscrambling process is also a useful illustration of certain established principles of cognitive psychology, many of which have a close relation to information-theoretic measures like message redundancy and channel noise. Such matters may well be further detailed in the references given by Leo, so here I will just mention a few obvious ones without going into subtleties. Those setting cryptology-based puzzles, like anagrams or crosswords, have probably given considerable thought to many of the same themes.

Most examples of word-scrambled texts I have seen conform to the case where individual words have the usual separators (spaces, commas etc.). It is a reasonable hypothesis that primary word-segregation is a (pre-conscious) analysis, fairly well automated.

Topical familiarity, general word-frequency

Another fairly automated effect, though quite dependent on individual experience

Letter-order

The edge-effect enhances perception of letters at the beginning and end of a word. Scrambled words with the first and last letters intact also greatly benefit because these strongly-perceived letters are precisely the ones often generating superior cues to the associative thesaurus, perhaps including a subvocal phonetic component.

Word length

Whilst short words are objectively easier, word-length itself provides significant information, and in longer words this may partially compensate for increased complexity.

Rarity

Less familiar (or exotic) letters are more readily noticed in scrambled text due to an analogue in selective attention to effects of feature-contrast long known of in studies of short-term memory. Such letters assist also because they restrict search, due to their presence in a much smaller subset of lexical items.

Semantic expectation

This is dependent on the ongoing robustness of the perception of meaning. A text which would be challenging even unscrambled will offer much greater difficulty if scrambled.

Elimination via syntactic constraint

English, whose case-endings and verb conjugations have largely been smoothed away is particularly dependent on word-order in phrases as a cue to the function of a particular word (as verb, noun, adjective, adverb &c)

Internal morphemic articulation of words

Various issues partly encapsulated in the German example cited by Leo.

Orthographic articulation

In longer words characteristic high-frequency constituents (things like "-ness", "-tion", "-ily", "pre-" &C) may be cued simply by the presence of the letters in the word. This is similar to, but distinct from, word-frequency effects.

Distraction

Certain combinations of letters may increase the perceptual effort required because they have strong internal coherence, but happen to be misleading in the particular case.

Before pronouncing on the likely differences between languages, I think it would be more worthwhile to obtain a clear and thorough understanding of the roles of the above factors, and similar ones, in the process of sentential scrambled word-perception.

$\endgroup$
8
$\begingroup$

The neologism used to describe this phenomenon is Typoglycemia.

It relates to the cognitive processes behind reading written text.

Randomising letters in the middle of words have little or no effect on the ability of skilled readers to understand the text.

Because the human mind does not read every letter by itself but as a whole, it would work for any given language where we can apply the same randomisation and keep the words separated in the same order.

Let's take the notorious German example Donaudampfschiffahrtsgesellschaftskapitän, meaning Danube steamship company captain:

  • Our noun Donau remains the same
  • It is possible for our noun dampfschiffahrts to be reformed into dampfschifffahrts
  • Our noun gesellschaft becomes the adjective gesellschafts
  • Then our noun kapitän remains the same

Let's see what happens when we scramble the compound lexeme:

  • Unscrambled German: Donaudampfschiffahrtsgesellschaftskapitän
  • Scrambled German: Dsnegapcfssamspfkaihdoftctflhreulasihatän
  • Unscrambled English: Danubesteamshipcompanycaptain
  • Scrambled English: Dbipiutpaccmmaapeansytoeshann

Virtually impossible! Now let's see when we scramble each lexeme, then compound it:

  • Original German per lexeme: Donau-dampfshiffahrts-gesellschafts-kapitän
  • Scrambled German per lexeme: Doanu-dhsffhmftpiacras-glhcselefasts-kipatän
  • Scrambled German composition: Daonudhiahscfmffatrpsglhcselefastskipatän

  • Original English per lexeme: Danube-steamship-company-captain

  • Scrambled English per lexeme: Dnbaue-stahmseip-camonpy-citapan
  • Scrambled English composition: Dnbauestahmseipcamonpycitapan

The composition being nearly just as readable as per lexeme!

Feel free to try another here.

Now let's take an example in Mandarin.

Script:

  • 研究表明
  • 漢字的順序並不一定能影響閱讀
  • 比如當你看完這句話後
  • 才發現這裡的字全都是亂的

Pinyin:

  • Yánjiū biǎomíng
  • Hànzì de shùnxù bìng bù yīdìng néng yǐngxiǎng yuèdú.
  • Bǐrú dāng nǐ kàn wán zhè jù huà hòu
  • cái fāxiàn zhèlǐ de zì quándōu shì luàn de.

Now let's invert 6 disyllabic characters.

Script:

  • 研表究明
  • 漢字的序順並不定一能影閱響讀
  • 比如當你看完這句話後
  • 才發這現裡的字全是都亂的

Pinyin:

  • Yán biǎo jiū míng
  • Hànzì de xù shùn bìng bù dìng yī néng yǐng yuè xiǎng dú.
  • Bǐrú dāng nǐ kàn wán zhè jù huà hòu
  • cái fā zhè xiàn lǐ de zì quán shì dōu luàn de.

This rather tame example now makes little sense, but does not quite appear completely disordered (with the exception of the preserved third line).

Further materials:

$\endgroup$
  • 1
    $\begingroup$ I don't think your comparison in German is a fair one. You used a word that combines several roots, and thus has in it a grammatical structure that the inside of an English words lacks. What happens if you scrambled while respecting that structure? i.e. keep the first and last letter of each root and the connectives as-is and randomize the letters inside each root. Will the word then be readable by a skilled German reader? $\endgroup$ – Artem Kaznatcheev May 7 '15 at 1:02
  • 1
    $\begingroup$ Agreed, thank you for your suggestions Artem. I have elaborated further, let me know what you think. $\endgroup$ – Leo May 7 '15 at 10:43
  • 1
    $\begingroup$ @LeoTM, Is typoglycemia exist in asian languages? What about permutations/variations in structure of single Chinese alphabet character? $\endgroup$ – hOff May 7 '15 at 12:36
  • $\begingroup$ Hi @LeoTM - sorry but I disagree with part of your answer (it made me write my own!). Actually, the human brain does read every letter by itself and read the word as a whole - both. $\endgroup$ – marsei May 11 '15 at 23:33
  • $\begingroup$ Hi @macduf - Thank you for your answer! I do believe Graham Rawlinson meant "the human mind does not [only] read every letter by itself but as a whole". $\endgroup$ – Leo May 16 '15 at 12:20
7
$\begingroup$

The ‘jumbled word effect’ is due to the special way in which the human brain encodes the positions of letters in printed words.

Psycholinguists investigate this effect with a procedure called masked-priming where a target word is primed with a briefly presented stimulus (usually a mix of target's letters). This lead to a model of word recognition that includes several consecutive steps from letter feature extraction by the occipital cortex to the word activation in the visual word form area. Within this model, all letters are individually recognised in their actual printed positions (position-dependant phenomenon) and are further combined in groups of letters. These groups of letters are not encoded with their absolute positions but their relative positions to each other. It is formalised in a model as:

  • letters of other are coded as 'o', 't', 'h', 'e', 'r' in this particular order,
  • but groups of letters are coded as 'er', 'he', 'ot', 'th' where the order of the groups is not important.

When we read, we automatically switch from an absolute-position code to a relative-position code, and this is where the jumbled word effect arises. The scrambled word activates the same groups of letters.

enter image description here

Does this phenomenon relate to the language structure, i.e. is specific to English?

This phenomenon has been observed in languages that use an alphabetic script, English included. There is an on-going debate on the locus of such effect (purely visual attention or related to literacy).

Would trick works for languages with longer words, like German?

Yes, to some extent as discussed by @David Holden

Does the same phenomenon exist for Asian, non-Latin alphabet based type languages >like Mandarin?

It has been observed in Thai.

Ref:

  • On the Flexibility of Letter Position Coding During Lexical Processing: The Case of Thai. Perea et al. (2012)
  • Modeling letter position coding in printed word perception. Grainger & van Heuven (2003).
$\endgroup$
  • $\begingroup$ Hi @macduf, note typoglycemia has also been observed in Mandarin. $\endgroup$ – Leo May 16 '15 at 13:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.