Forgot your password?
typodupeerror
Science News

Method of Reading Discovered 181

Posted by ScuttleMonkey
from the going-crosseyed dept.
Scientists have discovered that the method our eyes use to process letters on a page is different than previously believed. Instead of assimilating one letter at a time our eyes actually lock on to two different letters simultaneously about half the time. "The team's results demonstrated that both eyes lock on to the same letter 53% of the time; for 39% of the time they see different letters with uncrossed eyes; and for 8% of the time the eyes are crossing to focus on different letters. A follow-up experiment with the eye-tracking equipment showed that we only see one clear image when reading because our brain fuses the different images from our eyes together."
This discussion has been archived. No new comments can be posted.

Method of Reading Discovered

Comments Filter:
  • Frsit Psot (Score:5, Funny)

    by Harmonious Botch (921977) * on Monday September 10, 2007 @01:09PM (#20541039) Homepage Journal
    If yuo corss yuor eeys smoetemis, you sohlud be albe to raed tihs qitue eailsy. I terid it, and it mdae all the sepllnig msitaeks on salsdhot go aawy. Hvewoer, it ddi not ipormve Sttucle Mkoney's eitding.
    • by biocute (936687)
      I wished it was true, but this discovery is about cross-eyeing letters, not words.
    • by kalirion (728907)
      Wow, I actually read that pretty easily. Except for Sttucle, which caused me to look at the editor's name.
      • Re: (Score:3, Interesting)

        by Paradise Pete (33184)
        Wow, I actually read that pretty easily.

        Yaeh, it's prttey amaizng taht as lnog as the begnning and the enidng of the wrod are coerrct taht you can raed it at alomst full speed.

        • Re: (Score:3, Informative)

          by ThosLives (686517)

          Yeah, I can tell what it's intended to say, but it still doesn't mean I'd accept stuff like that. It's almost as bad as text-message writing.

    • by steelfood (895457)
      Actually, there was a research article about scrambling words a long time ago. The jist of the research was that there's a very small difference in reading speed if you scrambled all the letters in a word except the first and last, or something along those lines. However, if you completely reversed the order of the letters, it would take a lot longer to read.

      This is an interesting followup to that research.
  • by WiglyWorm (1139035) on Monday September 10, 2007 @01:13PM (#20541089) Homepage
    Aoccdrnig to rscheearch at Cmabrigde Uinervtisy, it deosn't mttaer in waht oredr the ltteers in a wrod are, the olny iprmoetnt tihng is taht the frist and lsat ltteer be at the rghit pclae. The rset can be a toatl 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.

    In other words, this study was flawed in the first place. Our eyes don't look at individual letters, they look at groups at a time. I learned this in high school....

    • by Verteiron (224042) on Monday September 10, 2007 @01:19PM (#20541205) Homepage
      A_d y_t t___e e___a l_____s a_e i_______t.

      And yet those extra letters are important.

      Bt yu cn lv ot innr vwls and stll be mstly rdble.
      • by nganju (821034) on Monday September 10, 2007 @01:44PM (#20541687)
        Very true. Interesting discussion of that whole "Aoccdrnig to a rscheearch at Cmabridge" thing below, which largely debunks their claims. [typewriting.org]http://typewriting.org/2003/09/14/Aoccdrnig_to_a_rscheearch.../#content [typewriting.org]
      • by cp.tar (871488)

        Bt yu cn lv ot innr vwls and stll be mstly rdble.

        That's been known since cuneiform times at the very least.

      • by vertinox (846076)
        Bt yu cn lv ot innr vwls and stll be mstly rdble.

        BTW, ancient Hebrew has no vowels.
        • by Dun Malg (230075)

          Bt yu cn lv ot innr vwls and stll be mstly rdble.

          BTW, ancient Hebrew has no vowels.
          modern Arabic still doesn't
      • by Plutonite (999141)
        And it gets even more complicated. The letter/shape grouping thing is well known in cognitive psychology (it's called Gestalt for Gates' sake) but how exactly the brain parses the words is a difficult question. First you have to consider the fact that context plays an important role, the brain uses it's knowledge of correct syntax to predict the next term - so if I wreit thnigs lkie this then (A) bad grmamar sihthaed (/A) you will realize that after (A) your brain takes much longer to recongnize the words s
    • by Stamen (745223)
      Oh really? Kindly link to said study at Cambridge University (or Cmabrigde Uinervtisy if you prefer) that are referring too.
      • Re: (Score:3, Informative)

        by stoolpigeon (454276)
        Here's a decent rundown of the thing [cam.ac.uk] it made the front page here at the dot - though I'm having a tougher time tracking that down.
    • There is a minor exception to the first and last letter rule: on three letter words, the second and third letter can be switched.
    • Baloney. If we can read that it's because we are already good readers. "Whole Language" is where good readers end up, but that's not how we learn to BECOME good readers.

      Seriously, try this one, Mr. Wizard:

      "Atluds nveer tkae tmie to tnihk aoubt how to pcnuonore the iaudividnl wdros; tehy jsut sacn anolg at a vrey fsat cilp and triehr bniars tkae crae of the "bnikaerg-dwon" of the pmargonohs allacitamotuy and aletaruccy. Hevewor, ttha's atluds who lenraed to raed wtih pcinohs. Atluds who rley olny on shgit-rgnidaeg teuqinhces rleray gnia mcuh foitcnun, and boy, deos taht sohw in our steicoy tadoy, wtih rlevitaley low lleves of lcaretiy cerapmod to gnoitarenes psat. Cerdlihn tadoy, who dno't hvae pcinohs ioitcurtsnn, are bllacisay gnisseug at waht wdros maen, and it swohs in enihtyrevg form sezidradnatd tset serocs to lcaretiy deicneicifes in the wcalpkroe."
      From http://www.gobiged.com/wfdata/frame265-1059/pressrel45.asp [gobiged.com]

      Y Hole Langwidg Seams OK

      [...] Adults never take time to think about how to pronounce the individual words; they just scan along at a very fast clip and their brains take care of the "breaking down" of the phonograms automatically and accurately. However, that's adults who learned to read with phonics. Adults who rely only on sight-reading techniques rarely gain much function, and boy, does that show in our society today, with relatively low levels of literacy compared to generations past. Children today, who don't have phonics instruction, are basically guessing at what words mean, and it shows in everything from standardized test scores to literacy deficiencies in the workplace.

      • Sorry, I mistyped a word:

        Atluds who rley olny on shgit-rgnidaeg teuqinhces rleray gnia mcuh foitcnun
        should be

        Atluds who rley olny on shgit-rgnidaeg teuqinhces rleray gian mcuh foitcnun

        In my defense, I must say it was really hard to proofread.

        • by Kamots (321174)
          Given that this is slashdot and all... I've got to wonder why you didn't whip up a quick script to do your internal reversing? :P
      • The article isn't talking about "learning to read" though, it's talking about how we read. It makes no distinction at all about how we learn to read, and therefore has absolutely nothing to do with "whole language" method versus phonetic method.

        By the way, I've never seen your example before and the only word I had trouble with was "Adults" in the first sentence. Once I saw it repeated, however, the context made sense and it was fairly obvious what the paragraph was saying. As for children "basically gue
        • by Reziac (43301) *
          'As for children "basically guessing at what words mean" that's been the case for as long as children
          have been learning to read.'

          Er, well, not for those taught phonics. To us, ALL words consist of recognisable parts, and we almost never have to make a wild guess at the meaning; rather, we can judge probable meaning by those parts we CAN decipher. Which means we're never entirely lost, even in a sea of unfamilar words.

          • by bkr1_2k (237627)
            I'm guessing you haven't done too much reading out loud with children, but I can guarantee you that every kid I've ever read outloud with (which is an admittedly small number of 20-30) was indeed guessing words regularly. Especially when reading new things.

            Sounding out a word doesn't automatically create some sort of magical understanding of meaning, though it can occasionally help. As for "we're never entirely lost, even in a sea of unfamiliar words" I'd wager that's about equivalent for children taught
      • by Reziac (43301) *
        In my observation, "whole word recognition" (or "see first three letters, make WAG at rest of word") is how many dyslexics read (actually, ALL those I know personally and have watched reading do WWR of some sort). WWR simply teaches everyone to read at the minimal level achieved by untutored dyslexics. IOW, it makes everyone equally crippled!

        When I RTFA, my first thought was -- Oh, that explains "letters crawl around" dyslexics; their brain doesn't re-integrate the letter groups properly.

        I'm also reminded o
      • My mom once told a story about the one time that my cousin was helping her with some baking. She was like, "Q, read off the recipe to me", and he said, "blah blah blah, Pecans, blah blah". My mom was like, "pecans? This recipe doesn't call for Pecans. ???"

        So she went and looked at the recipe herself, and it called for walnuts. She was like, "Q, this says 'Walnuts', not 'Pecans'." Cousin Q responded that they were both nuts, and didn't get why it was important.

        My mom talked to her mother in law, who got
        • by Skreems (598317)
          I hate to break it to you, but not understanding why you need to make a distinction between walnuts and pecans in cooking isn't a problem with reading. That kid's just stupid.
    • So you don't really understand what I'm saying when I say.

      Where did you get the form from
    • Re: (Score:3, Informative)

      by necro2607 (771790)
      Ahh OK for the record, Cambridge University didn't do any of this alleged research [cam.ac.uk], according to Matt Davis, a "cognitive neuroscientist interested in language" working in the Cognition and Brain Science unit at Cambridge. Read the link for further details, and a lot more interesting analysis/discussion on this same phenomenon in other languages and whatnot. :)
    • by hackstraw (262471)
      Our eyes don't look at individual letters, they look at groups at a time. I learned this in high school....

      Actually, it appears as though our brain does like the inverse of how a fractal is generated. Fractals get more detail until you quit. Our brain gets the outline of the text, and context, and other things, and then gets the meaning of what is said. THAT IS WHY CAPS ARE HARDER TO READ. The letters are the same, but the spacing between characters and their height helps us.

      Also, I thought this was old
    • by bkr1_2k (237627)
      And that was almost 20 years ago for me.

      This is nothing exciting and certainly isn't news.
    • I'd be curious if the effect works with braille. Is the information routed through the area of the brain needed to "unmix" letters or not? I would guess yes, but the result would be interesting either way. Some processes take place in the eye itself, but this doesn't seem like one of them.
    • by Yvanhoe (564877)

      In other words, this study was flawed in the first place. Our eyes don't look at individual letters, they look at groups at a time. I learned this in high school....
      There are quite a lot of layers between the eyes and the upper levels of consciousness. This study is interested in very low level layers : eye movements. The fact that eyes scans one letter at a time or two letters simultaneously does not make assumption about how words are treated in the next layer.
    • I learned this in high school....

      I learnt about evolution in primary school. That doesn't mean it is an invalid subject for scientific investigations. Especially when the results directly contradict what you thought you knew:

      The team's results demonstrated that both eyes lock on to the same letter 53% of the time

      Maybe the fact the eyes do something differently to the higher brain functions is important, maybe not. I was aware of the same corny jokes as you, but had no idea they had such detail on what

  • "lock on to two different letters simultaneously about half the time."


    ...half the time, every time.
  • I could have sworn we knew this was where dyslexic came from, that you see two letters that don't end up in the right order in your head.

    • by davidsyes (765062)
      How about Spoonerisms? I haven't read all of this topic, but, in the 7th grade, what got me hooked on Spoonerisms was a friend saying:

      "Miss on YOU pister. You aren't so MUCKing FUTCH. Why don't you go in your jack yard and back off."

      Of course, that got me into trouble a few times. Once, I hailed out to my mom that this TV movie, starring Elizabeth Montgomery, was starting. I said, "Ohh, mom, it's that lady from Webitched".

      She scrambled over to me and yelled that she had told me switching words around would
  • by natpoor (142801) on Monday September 10, 2007 @01:18PM (#20541193) Homepage
    That's pretty cool, but what about non-alphabetic systems, such as Chinese, Japanese, or Korean? Does the physical act of reading depend at all on the unit of meaning we are scanning with our eyes? Not that the researchers should have done this in the same experiment, they're in England, so it makes sense for them to stick to the native language.
    • They should study how a person with only one eye reads also. Or one eye closed.
      • It would be interesting to compare tactile reading to visual reading. How much is pre-processed by the eye(s), how much is handled by the brain, and how much is routed around the brain?
      • I'm blind in one eye and am also a very fast reader. I know I personally tend to see sections of the page, focussing on whole words and a word or two before and after. I certainly don't read the whole words, I more "predict" them from the first couple letters and the general shape.

        I have no trouble with the scrambled-up examples higher up, except for a couple words where they got overzealous.

        It's also worth noting that I'm a computer programmer and avid reader, so I read all day, every day. I might not be t
    • Re: (Score:2, Interesting)

      by someme2 (670523)

      That's pretty cool, but what about non-alphabetic systems, such as Chinese, Japanese, or Korean?

      Korean (Hangul) is an alphabetic system. The study might be really interesting there because Korean letters are always aggregated in blocks of two or three letters. It's part of the way they write. I have no idea if Koreans read these blocks as one.
      It's also a cool system because it was designed from scratch and follows a number of logical rules that makes it comparatively easy to learn (the alphabet... no

    • by bkr1_2k (237627)
      Hate to disappoint you but Korean and Japanese both have alphabetic systems. Yes they also use Chinese characters (rarely for Korean) but they do have alphabets and they're both completely phonetic. IE the same letters always make the same sounds within a given structure.
  • by Dausha (546002) on Monday September 10, 2007 @01:21PM (#20541243) Homepage
    I've known for about 20 years that we don't focus on one letter. There are numerous books that show that we (at least those using Latin alphabets) look at the shape of the top half of the word rather than each letter. All this does is break down literacy to crossing eyes, etc. Not really new.
  • by juuri (7678) on Monday September 10, 2007 @01:24PM (#20541315) Homepage
    You remember all that, the letter order doesn't matter when reading bit? (http://www.languagehat.com/archives/000840.php for a refresher).

    It's always seemed pretty apparent to me that we don't reach letters in "correct order" by focusing only on a single one at a time. If that were the case things like speed-reading and scanning for content would be nearly impossible. Outside confirmation of this is nice however.

    The real question is how much redundancy can we remove from printed words for faster information dispersal while still expressing things clearly. Sure, having everything spelled correctly and in long form is great for books for pleasure (art) but do we really need it for basic information sharing? Especially if doing so increases the time spent needlessly?

    • by p3d0 (42270)

      It's always seemed pretty apparent to me that we don't reach letters in "correct order" by focusing only on a single one at a time. If that were the case things like speed-reading and scanning for content would be nearly impossible. Outside confirmation of this is nice however.

      Well then, despite your remarkable intelligence, you missed the point of the article, which was contained in its first two paragraphs.

      It's not about where you "focus". It's about the fact that your two eyes look at different letters simultaneously while you read.

    • by cp.tar (871488)

      The real question is how much redundancy can we remove from printed words for faster information dispersal while still expressing things clearly. Sure, having everything spelled correctly and in long form is great for books for pleasure (art) but do we really need it for basic information sharing? Especially if doing so increases the time spent needlessly?

      I think that would be redundant, or even counter-productive.

      Our brains eliminate superfluous information automatically. However, something you find redundant may be necessary to me, e.g. because I speak a different first language (pulling a parallel with phonetic systems). Therefore it may mean a bit less work for some, and a lot more work for everyone else.

    • by Reziac (43301) *
      But it's not really redunancy. Think of normal words as uncompressed data, and "redundancy removed words" as *compressed data*. It's more work to "unzip" the word, then read it, than it would have been to just read the uncompressed word in the first place.

    • by bobdotorg (598873)
      Sure, having everything spelled correctly and in long form is great for books for pleasure (art) but do we really need it for basic information sharing? Especially if doing so increases the time spent needlessly?

      I'm left with a tough choice for the more appropriate response:

      1. f u.

      2. GTFOML!!!

      Ugh. IM speak.

      I don't know why, but I value capitalization and complete words / sentences in my reading. Usenet was my first exposure to 'u' substituting for 'you'. It drove me nuts, and it just got worse from th
      • by otomo_1001 (22925)
        Karma burn time.

        so y r u so srios abut abbr and LC typing????!?!?!??!?!?!?!?!?!

        Honestly, it is a wonder she didn't clock you upside your head with the shift key gift. Talk about being just a skosh insensitive. And this coming from a guy.
  • duh (Score:2, Insightful)

    by jm.one (655706)
    [i]"A follow-up experiment with the eye-tracking equipment showed that we only see one clear image when reading because our brain fuses the different images from our eyes together."[/i] Wait... they neeeded a follow up experiment to discover something that is so well known that it s rather common knowledge? I mean.. the other stuff isnt actually news either but this... and how does eye-trracking lead to a RESULT about what the brain does. I mean... an eye tracking experiment leading to a thesis.. or support
  • by Dhrakar (32366) on Monday September 10, 2007 @01:25PM (#20541337)
    Very interesting. I wonder if this could be a contributing factor to why folks get headaches when reading on some computer screens. That is, computers, unlike books, are constantly redrawing the screen so not all of the letters may actually be visible very well at any one time. Your brain starts straining because it can't scan multiple letters (or entire words?) very well due to the flicker. Do eletronic book readers have a high refresh rate?
    • Most people's monitors are excessively bright (especially if they use the same brightness day & night), which causes eye strain.
  • by LOTHAR, of the Hill (14645) on Monday September 10, 2007 @01:26PM (#20541357)
    Reading is a process of pattern recognition. We recognize and assemble patterns of letters/symbols and then associate those patterns with meaning. Some people can recognize larger patterns at a time, other people can only recognize shorter patterns. Most people move past the "processing a single letter at a time" stage of pattern recognition at a young age. Personally, I read whole multiple words or even short sentences at a time.

    This has been known for a very long time.
    • by matt me (850665)
      >Reading is a process of pattern recognition.
      What's the regex syntax like?
    • by davidsyes (765062)
      Kind of on/off topic?

      Take this example...

      "Reading is a rocess of pattern recognize. We recognition and assembly patterns of letters/symbols and then association tho patterns with meaning. Shome people can reconi larcher patterns at a time, other people can only reconi shorter patterns. Mot people moo passed the "processing a single letter at a time" stage of pattern reconize at a young age. Personally, I read whole multiple words or even short sentences at a time."

      That modifying I did to your dialog is how
  • Ligatures (Score:3, Insightful)

    by Biff Stu (654099) on Monday September 10, 2007 @01:28PM (#20541393)
    Typographers have used ligatures for ages. Now we have a scientific explanation.
  • by DeeVeeAnt (1002953) on Monday September 10, 2007 @01:29PM (#20541415)
    It seems that technical documentation is often optimised to take advantage of this phenomenon. For instance, recent tests on IBM's Tivoli Access Manager docs caused my eyes to cross 130% of the time.
  • Fusing images (Score:4, Insightful)

    by BorgDrone (64343) on Monday September 10, 2007 @01:33PM (#20541477) Homepage

    A follow-up experiment with the eye-tracking equipment showed that we only see one clear image when reading because our brain fuses the different images from our eyes together.
    No it doesn't.

    There is no internal 'viewscreen' that the brain displays the images on. (a so called "cartesian theater" [wikipedia.org] ) after all, if that happens, who is watching the screen and how does that work ?

    Instead of an internal 'framebuffer' I think* it's more like a MVC kind of system. Instead of pasting parts of images on an internal framebuffer to make up a whole, the individual parts are used to fill the datamodel of the world you've got inside your head. You 'see' the datamodel.

    * - This is all just a bit of philosophizing on my side, I may be completely wrong.
    • Re: (Score:3, Interesting)

      by Artifakt (700173)
      I'm glad you put quotes around 'see' there - it sums things up nicely. The datamodel we experience when reading contains lots of interlocking sensory cues. A simple concept, such as the word 'cow', may trigger a visual image of a cow, the sound of the spoken word cow, or for some rare people, even a smell cue. Sometimes a reader may become aware of a related sensory or logical image, i.e. first thinking of the sound of the word 'cow' may trigger the brain recalling the sound a cow makes and adding that to t
    • There is a framebuffer. It's called persistence of vision. I agree with you about everything else though. Some people do wait for each letter or pair to come into focus. Others pattern match entire concepts and associate based on that and move on without taking the time to bring all the symbols into focus. Trouble can happen when what you expected to see is not what is written but the process is adaptive and you slow down on new ideas and clever turns of phrase. Interestingly the latter method is bett
  • Dyslexia (Score:3, Interesting)

    by Dan East (318230) on Monday September 10, 2007 @01:43PM (#20541683) Homepage Journal
    The article only touches on this (last word in the entire story), but this should have ramifications in studying and treating dyslexia. At first glance, it would seem very strange how people could suffer from dyslexia. Why would they perceive pairs of letters and numbers as flipped, if we read in a serial fashion? If both eyes aren't even looking at the same letter then the physiology begins to make more sense - somewhere along the way the information isn't being assembled in the proper order.

    Dan East
    • At first glance, it would seem very strange how people could suffer from dyslexia ...

      You've reminded me of a law firm where I once worked. It was one of those international prestigious, original-art-on-the-walls kind of places. One of the regular receptionists, it turns out, suffered from dyslexia. Not a quality one would want in someone whose job it is to take and give out names and numbers. To make things worse, she was a native or Rochester, New York. The sound of her accent was something like you'
  • Can you type a little more slowly? I'm having trouble keeping up ...
  • No. NO. (Score:4, Informative)

    by wonkavader (605434) on Monday September 10, 2007 @01:51PM (#20541801)
    OK, I'm hoping the real article is not nearly as silly as this blurb. I'm sure it talks about this situation in more reasonable terms. But the blurb focuses on eye position and implies that it has meaning on a letter level.

    While the issue of eye position is interesting, we are NOT focusing on a letter. We are not reading letters, much less looking at them.

    Hold you arm out. Raise your thumb. Look at it. The space of the back of your thumb, at that distance is special. That's your fovea -- the area of your eye which has the greatest acuity. When you read, depending on font size and text distance, that area covers multiple lines of text, and usually more than one word. Focusing on a letter means picking that letter as a point in the text, and seeing the areas around it.

    A strong reader is picking up both the words below and left and right of the word he/she is reading at that fraction of a second.

    Yes, it's interesting to ask where we fixate. Yes, it's VERY interesting that we go crosseyed and that begs the question of whether we do it systematically to reduce the amount of new data which is common in both foveas, either to increase speed by processing both independently, or to reduce the amount in common and thus reduce the load that reading takes (you'd possibly see that in a "difficult" or unfamiliar word). However, we do NOT look at letters. They're just a spot.

    Someone asked here about other languages, do we do the same thing for Kanji, Hangul, etc.? Is suspect that things might be different there, as I suspect that this behavior that they've found is strongly connected with syllable boundaries in English. However, eye-trackers are notoriously inaccurate (unless you're willing to have a coil surgically implanted in your eye, and even then, it ain't fantastic) and so their letter accuracy information must come from AVERAGES ACROSS MULTIPLE OBSERVATIONS. This should lead us to ask what their dataset was and what behavior they saw on specific character clusters. (That, in turn leads us to question if they got enough data to get much accuracy on those clusters.)

    It would be nice to see the original article, as opposed to this fluff piece.
    • It's probably more likely that our eyes just are not able to accurately position on a page of text. The brain tells both eyes to 'look at' the same exact place, but due to variations in muscle response and delays getting the command sent to the eye they end up at slightly different places.

      You could probably reduce this error though by putting features on the page of text that let the eye track better than vast areas of white. For instance if the same shape and size text was engraved in natural wood the ey
  • Before we get yet another "I always knew I didn't look at one letter at a time! I read whole words at a time! hyuk hyuk", please go read the first two paragraphs of the damn article. Or, for that matter, the summary right here on Slashdot. I don't think I have yet seen even one response I didn't write that understood they were talking about the reader's eyes looking at two different things at the same time.
  • Anyone else end up with a headache while trying to independantly prove or disprove the findings of the study?

    I need to go home now.
  • Words kind of look like cars in a train. Some are longer and shorter than others. Some have big "wheels" (descenders) or thingees sticking up. I have word shapes memorized for standard modern fonts, but if see a strange font or handwrighting, then I go back to letter mode. I thing the "shape" mechanism is why the scrambling the letters inside a word, but not the first and last character, is generally readable.
  • How does this change the dynamics? Is there an upper limit placed on max reading speed, I read faster than most people do, but I know that the few people read faster than I do dramatically overshoot any speeds I could hope to achieve.
  • Maybe our eyes are looking at multiple letters simultaneously a lot of the time in order to find where one ends and the next begins. How can we find the separate individual letters without inspecting them to see if they're not just a single letter?

    This is interesting research, but the conclusions seem hasty.

Computers are unreliable, but humans are even more unreliable. Any system which depends on human reliability is unreliable. -- Gilb

Working...