Sunday, December 16, 2007

Cambridge Word Scramble Study: It's Fake Already!

It seems the bogus "Cambridge University Study" concerning shuffled words is making the rounds again, and this is one bit of crapola I never tire of debunking. Here is a typical depiction:

"From Cambridge University .

O lny srmat poelpe can raed tihs.
I cdnuolt blveiee taht I cluod aulaclty uesdnatnrd waht I was rdanieg. The phaonmneal pweor of the hmuan mnid, aoccdrnig to a rscheearch at Cmabrigde Uinervtisy, it deosn't mttaer in waht oredr the ltteers in a wrod are, t he olny iprmoatnt tihng is taht the frist and lsat ltteer be in the rgh it pclae. The rset can be a taotl mses and you can sitll raed it wouthit a porbelm. Tihs is bcuseae the huamn mnid deos not raed ervey lteter by istlef, but the wrod as a wlohe. Amzanig huh? yaeh and I awlyas tghuhot slpeling was ipmorantt! if you can raed tihs psas it on !!"


That this piece of obvious claptrap continues to impress is testimony not only to the ever sagging education level of Americans, but also to our ever-growing anti-intellectualism that gives more and more benefit of the doubt to anything that promises to validate ignorance and basic intellectual laziness. It is flawed in every way. I've never been able to confirm that such a study was ever done at Cambridge, but if it was, and this description accurately portrayed their findings, everyone associated with the study out to be booted from serious intellectual circles.

If we truly read words as a whole, then why must the first and last letters be fixed? Why can't the entire word be scrambled? And what exactly does "as a whole" mean anyway? How can one see a word as a whole without seeing the letters in it?

In a way this is a cheap magician's trick, because the only reason people can read the scrambled words is because they aren't very scrambled. Fixing the first and last letters means 2 and 3 letter words don't change at all, and 4 letter words just swap the middle letters. That's the bulk of our vocabulary. Try making a sentence with very long words, and our ability to read words "as a whole" mysteriously vanishes. To wit:

Bblaaesl pryleas pnmrrioefg sllaimy aeoulltsby dvrseee clbrpmaaoe tteenmrat.

is incomprehensible, because now every word is truly scrambled, with the first and last letters being an insignificant proportion of the total. So sorry all of you that thought you had academic backing to your poor spelling and grammar skills. They do matter, because baseball players performing similarly absolutely deserve comparable treatment.

44 comments:

  1. I love it! Thanks for clearing that up.

    Note: You mis-misspelled 'similarly.'

    I think this just goes to show that the meaning is harder to extract if the right letters aren't all there.

    Also, consider the following:

    My berad is gary.

    Do I keep it or throw it out?

    ReplyDelete
  2. As someone who does the "daily jumble", I can attest that it is sometimes difficult to unscramble words. :-)

    Of course I agree with you about this bogus "study", but I do think that there are some interesting problems there:

    1) how difficult would it be to write an unscrambling algorithm that would make sense of words that were scrambled to the degree that the given example was?

    2) What type of scrambling would be the hardest to decipher, and why?

    There are tons of lovely questions here!

    ReplyDelete
  3. 1. That is not a real study
    2. It was not truly random.
    3. But it still somewhat valid. If reasonable easy reading passage is truly scrambled most people would be able to read it. The odd word may stump us, and it may a bit slow, but you'd get it. Try it, there are online scramblers that do this.
    4. Why must the first and last letter be fixed? Because when reading we tap into our shape recognition capabilities. The first and last letters provides an outline of sorts.
    5. Our brains perform some type of disambiguity on the words. Like a Bayesian filter, it suddenly "recognizes" the word if a given threshold is passed. That is why the scrambled words are unscrambled so suddenly. There is no conscious process invovled. Try it. The harder words may take a bit longer, but you just look at and all of a sudden it "appears" (match found) or not at all. Your brain is doing the equivalent of a massive parallel "hacking" attempt.

    In summary, this "study" is a hoax, but it does reveal a lot about how we read.

    ReplyDelete
  4. I do apologize for posting a comment to a very old entry, but today I'm searching about this hoax study and stumbled upon your blog.

    I've also found this one, http://www.mrc-cbu.cam.ac.uk/~mattd/Cmabrigde/ The page was written and last updated in 2003, but the owner was trying to track down this "study."

    Personally, this hoax study would have been much more interesting it it wasn't claimed to be conducted by Cambridge. I think lying about its origin just to impress or convince readers is a lame attempt.

    Nevertheless, how the brain works in this case of reading is an interesting fact.

    ReplyDelete
  5. While the study may be fake, it is true that we do not read letter for letter or word for word. As such, letter and word order can be mixed up and not necessarily prevent reading and the comprehension there of. The key is whether our brain can engage in the phenomenon of "closure," or making whole the parts we can discern.

    Closure causes us to a circle when there is only a line segment drawn in the shape of a circle with the end points close togehter, but not touching. Closure is the reason we do not always recognize typos: the order and sequence of the letters is organized into a known pattern even when some of the letters required are missing or misaligned.

    Closure is likely what caused "similarly" to be come "misally." Both are real words, but the latter makes no sense in the unscrambled version of the scrambled sentence provided:

    "Baseball players performing misally absolutely deserve comparable treatment."

    The "correct" scramble might be: "Bblaaesl pryleas pnmrrioefg smiirally dvrseee clbrpmaaoe tteenmrat."

    Closure is also what causes us to now recognize the formerly incomprehensible sentence:

    "Baseball players performing similarly absolutely deserve comparable treatment."

    Thought the study is fake, it is not false that our brains engage in closure to create order in all that we perceive.

    ReplyDelete
  6. Although I am able to read and unscramble words rather well, I read slow always, so I wonder how well I create this so-called closure link. I did however see that the word out was used where the word ought was intended from "I've never been able to confirm that such a study was ever done at Cambridge, but if it was, and this description accurately portrayed their findings, everyone associated with the study out to be booted from serious intellectual circles." Additionally, I too realized that the scrambled version of similarly was missing the letter r, but it is actually a letter i as well, which I missed the first time.

    ReplyDelete
  7. The Cambridge word scramble study has been proven fake, please read this article on how the actual Cambridge scientist debunks the chain e-mail:

    http://www.foxnews.com/story/0,2933,511177,00.html

    ReplyDelete
  8. If we truly read words as a whole, then why must the first and last letters be fixed? Why can't the entire word be scrambled?

    Because holding the first and last letters constant means that the relative ordering of many of the digraphs are kept, enough to
    recognize the word.

    How can one see a word as a whole without seeing the letters in it?

    Just as we see a whole painting without seeing each stroke.

    ReplyDelete
  9. Your first point is also mine: we aren't reading words as a whole, we are mentally unscrambling the letters in the word. That's how we recognize it, and that's made far harder if the end letters are moved as well.

    I don't think the painting analogy holds. We see a painting without seeing each stroke because the colors blend. What we can see as a blue dot next to a yellow dot appears to become a green dot if we back up far enough. Letters don't blend like that, and certainly not if they are out of order.

    ReplyDelete
  10. Taht tihs pciee of obuoivs ctaaplrp cetniouns to ipresms is tsetinomy not olny to the eevr sniggag eoictaudn leevl of Aeimacrns, but aslo to our eevr-gownrig atni-iesanitltuecllm taht gievs mroe and mroe bfeeint of the dobut to annhtiyg taht pisoemrs to viladate ignaoncre and baisc illtcteeanul lianzess. It is fealwd in evrey way. I’ve neevr been albe to crfoinm taht scuh a sduty was eevr dnoe at Cmbardgie, but if it was, and tihs ditesrocpin acelautrcy peyarotrd thier fginidns, evryonee aiostescad wtih the sutdy out to be bteood form sioures inacelteltul circels.

    Recognize that? How about scrambled with out the first and last preserved?

    Atht tsih iepce fo vobouis lcaptpra esnotnuci ot pmressi si tytnoisem nto noyl ot eth vere snggaig nocudiate vlele fo Aearcinms, ubt osal ot rou erve-oiwnggr tnia-lnttliaislumcee thta egivs roem nad mroe enieftb fo hte otudb ot nitnagyh ttha omsreisp ot ladatiev ngriencoa nda acbis ieltlanlucte nziasels. Ti si wdflea ni rveye yaw. I’ev rvene bene albe ot niofrmc thta uhcs a dusty aws erve edon ta Biadcgerm, tub fi ti aws, dan tish icdosnipert catauycrel doypeartr ierht fdigsnni, roeenyve stdicoeasa thwi eht sydut uto ot eb boeodt omfr sresuoi uncleitlleat rccslie.

    ReplyDelete
  11. I forgot to mention:

    http://www.lerfjhax.com/scrambler

    This fellow has a nice ajax script that does scrambles of each type.

    ReplyDelete
  12. Sorry,but uh. Incomprehensible is a pretty long word,right? Well,you can scramble it,keeping the first and last letter the same,and still have people that are able to read it. You can read this,right?
    icohmnpesrnielbe,ssclaupreflgaiirsticixldpeaioluoics,dsntonleasiasim, and istutanoizlitis?
    I think it has more to do with the words you've seen before,not length. If you can somewhat recognize the word,then you can read it if scrambled.
    As for your sentence,
    it's not so hard to read.




    Well,can you?

    ReplyDelete
  13. http://www.mrc-cbu.cam.ac.uk/people/matt.davis/Cmabrigde/

    Here is the Cambridge link.

    ReplyDelete
  14. I think you are all missing the point. Stop being so over analytical about it and just see it for what it is. Something to amuse other people. The paragraph isn't scramble with the intent of "stumping" people. It's just pointing out how neat it is that the brain processes "familiar" words as a whole whether they are spelled right or not. Get a grip people...

    ReplyDelete
  15. you guys are all way, way too smart. too many big words. if you keep things simple, this game is easy. very few words containing over four letters log to get my point across.

    smile... or slime.

    ReplyDelete
  16. The mind of someone who has read for many years has what is called "word shape recognition" with the common words they read. This is why articles written for general consumption are to be written at a '6th grade level'.

    Another value of writing at this level is something called "Saccadic eye movements" which states that while reading the average person can only "see" 3-5 letters at a given time.

    Combined with "word shape recognition" if the words are greater than 6-7 characters long, the mind will lose some of it's ability to recognize the word.

    Morons....

    ==================================

    The mnid of soemone who has raed for mnay yeras has waht is caelld "wrod shpae recoingtion" wtih the cmomon wdros tehy raed. Tihs is why arcitles wrttien for geaernl cmnsuoption are to be wriettn at a '6th garde level'.

    Anethor vlaue of wrtiing at tihs lveel is someihtng caelld "Saciadcc eye movtmenes" whcih stetas taht wihle raeding the arevage porsen can olny "see" 3-5 lteters at a gevin tmie.

    Comibned wtih "wrod spahe ricogniteon" if the wodrs are graeter tahn 6-7 chatacrers lnog, the mnid wlil lsoe smoe of it's abliity to recognzie the wrod.

    Mroons....

    ReplyDelete
  17. The good thing about your information is that it is explicit enough for students to grasp. Thanks for your efforts

    in spreading academic knowledge.

    ReplyDelete
  18. Very nice and impressive article you have posted. Its very helpful, i have read and bookmark this site and will recommend it to more other peoples.

    ReplyDelete
  19. This blog is written on one of the divergent topics that writers have derelict composing a long while ago.

    ReplyDelete
  20. cool article providing a real information.

    ReplyDelete
  21. Even if the "Cambridge study" is invalid, there is still something to word shape recognition that should not be dismissed—especially with text set with a roman serif typeface (as opposed to a sans serif typeface).

    Typefaces like "Schoolbook" were specifically designed for use in elementary school primers, and have the characteristics (x-height, serif choice, stroke weights) that help a reader to see the whole word, as opposed to seeing a word as a group of individual letters.

    As a printer for over 40 years, I was trained that it is good typographical practice in the printing world to use roman serif typefaces for prose, and sans serif typefaces for "look-ups" like names in a phone book or a lead word in a dictionary entry. While the definition may be in a roman serif face, the lead word in the dictionary entry is usually sans serif, so it is easier to scan those lead words letter by letter.

    An experienced reader is more capable of interpreting those "jumbled Cambridge words" because of their general shape, which is improved by keeping the first and last letter intact. Having a more developed vocabulary is also an asset in this recognition.

    Yes, it is true that many words are short, and they do not get jumbled very much. And it is probably true that some jumbled words get derived by the context of their location in the sentence. I would also agree that the contention of the "Cambridge" study is not absolute proof of the "power of the human mind", but rather the result of repetitious experience of reading large amounts of text over time—especially those of a visual learner.

    Another thing to consider is that the rules for typesetting for print are different than the rules for typesetting on the web. While print can display typical reading type (roman serif) at 10-12 points in size very accurately and sharply, the resolution commonly used on the web makes usage of roman serif type at that same 10-12 point size not as sharp. Web text at that same size are sharper with sans serif faces, which reduce the ability to detect word shape. As a result, I would argue that a person's word shape recognition ability is better on a printed sheet in a roman serif typeface than on a web page with a sans serif typeface.

    In conclusion, just because a story has a fabricated source, it does not destroy all of those truths that are valid. It is sad that those word shape recognition truths are discredited by trying to make more out of this legitimate human perception.

    Historically speaking, if there is no validity to word shape recognition, there are a lot of type designers, typographers, and graphic artists that have been wasting their time for centuries—well before the "Cambridge study" was allegedly done!

    ReplyDelete
  22. The only challenging part of this project is creating the lid and inside compartments. The good news is, they follow the exact same process

    ReplyDelete
  23. Those words that are too long can probably be pretty well guess based on context. Thanks for this!

    ReplyDelete
  24. Thanks for the nice post. I am expecting some extraordinary idea from your side. You always place for some new thinking in your post

    ReplyDelete
  25. You completed a few fine points there. I did a search on the matter and found nearly all folks will agree with you.

    ReplyDelete
  26. Hello, good luck for me I was looking for this information for several months ago. Finally I got it all here. Thanks a lot!

    ReplyDelete
  27. Tihs is toltaly ture bcausee it is so esay to raed. The rseacerh at Cmadgbre Uinvreitsy is porevn ture. If you can read that. There is your proof.

    ReplyDelete
  28. How can you say that scramble study is fake?
    DWI laws

    ReplyDelete
  29. Wonderful article,thanks for putting this together! "This is obviously one great post. Thanks for the valuable information and insights you have so provided here. Keep it up!"

    ReplyDelete
  30. you're right, it may be fake but it still is amazing that the mind can read all that without messing up

    ReplyDelete
  31. So what if the study was a hoax? It was nonetheless a clever discovery and an empowering thought that the human mind is capable of decoding partially scrambled words.


    @freako

    Win.



    P.S. I'm atheist but I would never start a blog against religion *freedom of religion*

    ReplyDelete
  32. Have you stopped to consider that when you move the first and last letter also, that you remove the frame of reference needed to determine what word it is supposed to be? For example, the word "Nerve" could be misinterpreted as "Never" if it were not for the "N" and the "e" remaining in their proper place.

    I think that you need to think things out more before you write about them...

    ReplyDelete
  33. Ok, here is a question. Why can an 8 year old that is a poor reader read the jumbled words faster than the same words in the correct order? Puzzling.

    ReplyDelete
  34. Amazing that most posts are spelt wrong, in the wrong context or just purely wrong!!!!! And this is from a confirmed statemented DYSLEXIC eheheh.....IRONY!

    ReplyDelete
  35. I'm an Artificial Intelligence guy.... so here is my observation.

    Our brains are not auto de-scramblers. We are skilled context resolvers.

    As pointed out, the 3 and 4 letter words make up the bulk of our everyday vocabulary. Many 5 letter words are easy as well.

    We should all be able to read: “My mom siad to me taht tish sduty is wnorg”

    As the author of this blog pointed out, a sentence with all big words is next to impossible to read in the same way.

    The reason the phenomenon still works with some larger words thrown in there is because of our strong contextual resolution and neural firing. A related ability to resolve unknown words in context will serve as a good example of this.

    Consider this passage: “I was testing the hottest new PETCO product with my dog Fido. I took the new Pachuchu and hurled it with all my strength. Fido jumped into the air, grabbed it with his teeth, and brought it back to me.” In this passage, most of us would understand that a Pachuchu was some sort of toy or object like a Frisbee or stick. It is some gadgetry of sorts that is somehow like other things we use to play catch or fetch with our dogs. The context of the passage helped us understand that new word.

    Now consider the previous scrambled example with a few larger additions:

    “My mom siad to me taht tish sduty is wnorg. It is a lie. She siad it is flsae to bieleve in teshe slliy gmaes. Tehy are ricuuidlos and utnure.”

    Most of us can still read this even though there are words of 5, 6, 7, and even a 10 letters. This passage as a whole contains sufficient information to "fire" contextual neurons that can help us "resolve" a scrambled word (in context) without actually having to do much work. In a way, it sort of functions as an unknown word until we can 'test' the solution and proceed forward reading the sentence. The easy words such as “wrong”, “lie” and “false” put into context the bigger words “believe”, “ridiculous” and “untrue”. Those related “in context” neurons are firing at low levels before we actually encounter them and they can more easily be resolved because they are already in context (firing at low levels in our brain).

    I have not seen this explanation in any white papers or studies. I am simply sharing my observation.

    Cheers.

    -lkw

    ReplyDelete
  36. Lovely I'll have to come back for more! I just wrote about this (lately) but I totally agree for some of your reasons and some of mine it is a fake. Wikipedia is a bunch of trash too, you need to pick through it and then you don't get anything worth it anyway if you need facts. ♥

    ReplyDelete
  37. try scrambling longer words on a per syllable basis, the way dyslexic peoples brains do it. you can read at pretty much a normal rate with even very large words scrambled by syllable, so as you all seem to know how the human mind breaks down words (or at least are taking pretty educated stabs at it, like most science involving the brain is doing anyways) humans have funny glitches. anyways here is the program I found that actually scrambles any length of word so its still readable (ie by syllable) www.dafk.net/what/

    ReplyDelete
  38. Awesome blog. I enjoyed reading your articles. This is truly a great read for me. I have bookmarked it and I am looking forward to reading new articles. Keep up the good work!

    ReplyDelete
  39. I love it when things get improved, thanks for putting up a great guide. Hope it will help me out in my future projects

    ReplyDelete
  40. Well first off I want to say that I can read that sentence with the suposably hard scrambled words. If u can't read it your either under the intellectual level of such grammar or your trying to hard. When u let your brain do the work it comes much easier just don't over think it. Not saying that all the words came extremely easy to me but I'm not saying I had to go into with a pen and some paper. Your rain uses a decoding sort of action that you use for contemplating actions.this is how your brain is working. So no Cambridge isn't complete bogus , it's just not as easy as they make it seem:)

    ReplyDelete
  41. Actually this is the game of brain and subconscious mind.

    ReplyDelete
  42. mikayla:)

    a. Reading doesn't require a pen and a pad. (your explanation proved my point)

    b. You presented a 'false dilemma' (a logical fallacy) with your if-then-either-or. (http://en.wikipedia.org/wiki/False_dilemma)

    c. If you are nicer to others they will likely be nicer to you.

    ReplyDelete
  43. You might like to try the following link. It provides guidance, with worked examples, on understanding what the examiner seeks and writing in an academic manner. If in doubt can always ask tutors whether it's advice applies in your discipline and institution.
    UK Essay Writing

    ReplyDelete
  44. It's "ought to be booted," not "out to be booted".

    ReplyDelete