as long as the first letter and last letter.....

[Shig]

Surely some of you have seen this or something similar on facebook or, well, anywhere really: "ACDRNOICG TO RSAREECH AT CMABRIGDE UINERVTISY, IT DSENO'T MTAETR WAHT OERDR THE LTTERES IN A WROD ARE, THE OLNY IPROAMTNT TIHNG IS TAHT THE FRSIT AND LSAT LTTEER BE IN THE RGHIT PCLAE. TIHS IS BCUSEAE THE HUAMN MNID DEOS NOT RAED ERVEY LTETER BY ISTLEF, BUT THE WROD AS A WLOHE. IF YOU CAN RAED TIHS, PSOT IT TO YUOR WLAL. OLNY 55% OF PLEPOE CAN"

Can someone please help me explain to someone that it really is not the mind-blowing statement that they feel it is? Also, the 55% seems hard to find as being accurate when looking for actual research on the topic. If it is as amazing as some seem to think, please explain that, but it just simply seems goofy to me. I really think that anyone that is literate could handle figuring it all out but just think that most are able to do it quickly. I personally think it would take longer for me to read a book written this way, because even though I really had no problems in reading it did definitely catch myself re-scanning a word with my eyes once or twice.

[automagic]

You altlcauy hvae to cleralufy pcale the lteetrs to mkae it lleigbe. You aslo hvae to cooshe yuor wdros cullerfay owhtresie it deos not wrok.

You actually have to carefully place the letters to make it legible. You also have to choose your words carefully otherwise it does not work.

[Otara]

http://www.snopes.com/language/apocryph/cambridge.asp

Doesnt seem to really have a definitive study attached to it.

Otara

[Mangetout]

Also, it doesn't work very well with block caps.

I believe what's happening here is that most people read whole words at a time - interpreting them as discrete shapes garnished by the letters they comprise, rather than as a sequence of individual characters.

[si_blakely]

Well, they made the thing harder to read by sticking it in CAPS. Part of the reason the word/letter order thing works is that the outline shape of the word does not change much as long as the first and last letters are correct and the rest of the letters are randomised.

Acdrnoicg to rsareech at Cmabrigde Uinervtisy, it dseno't mtaetr waht oerdr the ltteres in a wrod are, the olny iproamtnt tihng is taht the frsit and lsat ltteer be in the rghit pclae. tihs is bcuseae the huamn mnid deos not raed ervey lteter by istlef, but the wrod as a wlohe. if you can raed tihs, psot it to yuor wlal. olny 55% of plepoe can.

Someone has posted a similar text with the first and last letters correct, but all the other letters reversed - this was much harder to read. I am pretty sure that if you put all the tall letters first or last you would have similar difficulties. A good link discussing this meme is here.

Si

[Shig]

As always, I love this site and you guys are great. Thanks so much for the replies.

[Richard Pearse]

The blurb as posted is a bit contradictory. To start with it says it doesn't matter what order the middle letters are in and then it goes on to say only 55% of people can read it, so for 45% the order of the letters does matter.

[xoferew]

I think this phenomenon contributes to learning to read being both harder and easier than it might be. I'm studying Japanese, and I can look at many characters and know from the general shape and line density what it says, so my eye moves on to the next, but when it comes time to write the character I have no idea. Similarly there are kids in my preschool class who can "read" all their classmates' names, but can't begin to write them. They can form the individual letters, but have little idea which letters are present after the first one. Yet if we couldn't predict and move on, reading would take forever.

[BobLibDem]

I doubt the 55% figure. I've never heard of anyone who couldn't figure it out.

[Hairy Bob]

I've seen that in the past, sans the 55% nonsense at the end. My guess is someone just tacked it on when they reposted it 'cause everything looks more plausible with a percentage value attached.

[Elendil's Heir]

I've heard that 37.5% of all statistics are made up.

[pancakes3]

the consonants are in the right place. the filler vowels can all pretty much be inferred. it's a crock of poo.

CMABRIGDE is a lot easier to construe than CARDIGEBM

[Markxxx]

It's like trying to read from right to left it's not that hard after a bit of practice

ecitcarp fo tib a retfa drah taht ton s'ti tfel ot thgir morf daer ot gniyrt ekil s'tI

[Ludovic]

on eth tohre hnad, ti nis't taht ahrd ot edcihpre a emssgea howse frist elttrse rae rongw fi ouy hoosec htem arefcully.

[Chronos]

I think this phenomenon contributes to learning to read being both harder and easier than it might be. I'm studying Japanese, and I can look at many characters and know from the general shape and line density what it says, so my eye moves on to the next, but when it comes time to write the character I have no idea. Similarly there are kids in my preschool class who can "read" all their classmates' names, but can't begin to write them. They can form the individual letters, but have little idea which letters are present after the first one. Yet if we couldn't predict and move on, reading would take forever.

I certainly remember a time as a child when I could not yet actually read, but would guess what each word was based on its first letter and the context. Sometimes I got it right, sometimes wildly wrong.

[Randy Seltzer]

A perennial favorite:
http://boards.straightdope.com/sdmb/...d.php?t=213599
http://boards.straightdope.com/sdmb/...d.php?t=211689
http://boards.straightdope.com/sdmb/...d.php?t=436487
http://boards.straightdope.com/sdmb/...d.php?t=325854
http://boards.straightdope.com/sdmb/...d.php?t=211907
http://boards.straightdope.com/sdmb/...d.php?t=213361

I know I missed at least one, because I remember arguing with someone about this in one of these threads, and that argument isn't in any of the ones listed above.

[Saint Cad]

Originally Posted by pancakes3

the consonants are in the right place. the filler vowels can all pretty much be inferred. it's a crock of poo.

CMABRIGDE is a lot easier to construe than CARDIGEBM

Also notice that the letters are transposed with their neighbor and not random. If the urban legend were correct, you should be able to read these three words with no problem.

irntieor cioorclde aaltotigr

[pancakes3]

also, anagrams would be a lot easier to solve. and the game textwist would be utterly pointless.

[Colophon]

Originally Posted by pancakes3

the consonants are in the right place. the filler vowels can all pretty much be inferred. it's a crock of poo.

CMABRIGDE is a lot easier to construe than CARDIGEBM

Yeah, but CARDIGEBM doesn't have the first and last letters in the right place.

But, as Richard Pearse says, the "only 55% of people can read it" statement totally contradicts the rest of it. (I've never seen it with the stat attached, which appears to be total bkloclos.)

[Malacandra]

As to random text, I wrote a randomizing program once that left first and last letters alone, and I found the resulting text quite easy to read.

[glowacks]

I thnk ts fr mr ntrstng tht rmvng vwls frm wrds tht hv t lst n cnsnnt dsnt mk th txt mpssbl t rd, xlthxxgh xt xs xxsxxr xf yxx knxw whxrx thx vxwxls xrx.

Smhw I nvr ntcd hw hrd "easier" s t dcd whn dng ths.

[Saint Cad]

Originally Posted by Malacandra

As to random text, I wrote a randomizing program once that left first and last letters alone, and I found the resulting text quite easy to read.

Then what 3 words did I write above?

[Randy Seltzer]

Originally Posted by Saint Cad

irntieor cioorclde aaltotigr

Originally Posted by Saint Cad

Then what 3 words did I write above?

Not that I can really read them, but some scrutiny and head-scratching reveals:

SPOILER:

interior

SPOILER:

crocodile

and

SPOILER:

alligator. Or rather "aligator"

[SciFiSam]

The reason it shouldn't be that surprising is, if everyone were looking for the individual forms of individual letters, then we wouldn't be able to read handwriting. It doesn't mean we don't need to worry about spelling - which is how I've seen some people use it - any more than scrawling illegibly is A-OK.

Originally Posted by Saint Cad

Then what 3 words did I write above?

It's rather more difficult out of context - I have no idea what the last word is.

[Chessic Sense]

Originally Posted by Randy Seltzer

Not that I can really read them, but some scrutiny and head-scratching reveals:

SPOILER:

interior

SPOILER:

crocodile

and

SPOILER:

alligator. Or rather "aligator"

Yep. St. Cad drives a Chevrolet movie the-Ater, too.

[spoiler]Just Google it, folks[/quote]


ETA: I'd just like to point out, too, that keeping the first and last letters of a 6-letter word means that the word is at least 33% correct.

[JWT Kottekoe]

I just wrote a little program to scramble words like this. After experimenting a little bit, I think the original claims are a bit over dramatized. If you fix any two letters, all 1,2, and 3 letter words are unchanged. 4 letter words have two letters swapped half the time and are easy to figure out. By the time you get beyond 8 letters or so it starts getting more difficult, you often have to figure it out from the context. Most text (including the original sample) has a high frequency of short words. There are lots of clues in the context and it does help to know that the first and last letter are correct. Here is a sample from a randomly selected paragraph in the New York Times. The original example was easy to read, but in this sample, for me at least, some of the words are stumpers and can't simply be read on sight. Our brains are pretty good at using the textual clues to figure out others.

The Oifcfe of Csiaeorsognnl Etcsih, an idneenepndt ietvgivtisnae unit that serevs as a prceot'rsuos ocefif, had abeslemsd waht it bileeved was cllnoiepmg poorf taht the ovrepaiplng vteos and fanciainl insrtudy fariinnu-dsg paierts crateed the acneaaprpe of a clnoifct of isttenre, eevn if no daels were mdae to canhge vteos as a rluest of the dnosotnai. But the etichs ciettmmoe, now led by Rrieptntavesee Jo Bnonre, Riaulebpcn of Ablaaam, edeosrnd a rropet by its own saftf that clcnueodd that these dnernis and ctokcail peratis were just ritunoe entesv, oaenrizgd by piofsresonal fdiauinns-rg costntlanus who work ietpelneddnny from the lwkrsmaae’ Hsuoe pocily staff. “hTe oarelvl rerocd darenmesottd that three wree no acepaeparns of iyrm,perptio” the rporet colduednc.

[JWT Kottekoe]

Here's another one, just for fun. It's really easy once you recognize it.

Fuor socre and seven yaers ago our fatrhes bogrhut fotrh on tihs cnnnioett, a new nitnao, cievcenod in Lybrite, and deedcitad to the ptoiorospin that all men are caeertd elaqu.

Now we are eagnged in a gaert cviil wra, tteisng whetehr taht ninota, or any noaitn so cecvineod and so detcidaed, can lnog edenur. We are met on a garet blei-leftatd of taht wra. We hvae come to daetidce a ptiroon of that fdeli, as a fianl rnstieg plcae for toshe who here gvae tiehr lives that that notain mhigt live. It is ahoetlgetr ftintig and poperr that we soulhd do tsih.

Btu, in a larger sesen, we can not dticeade -- we can not ctesracone -- we can not hollaw -- this goudnr. The bvare men, linivg and ddae, who sugrgletd heer, hvae centsrcaeod it, far aobve our poor pwoer to add or dcettra. The wolrd will lttlie neot, nor lnog remeebmr waht we say heer, but it can neevr feogrt what tehy did here. It is for us the linigv, rtaerh, to be dciedaetd hree to the unfhsinied wrok whcih they who fguoht hree have tuhs far so nboly aadnvdce. It is rthaer for us to be hree dtcieaedd to the graet tsak rnenaimig bfoere us -- taht from these hooernd dead we tkae icnseared dootiven to that cuase for wcihh they gave the last full msueare of dtoivoen -- taht we here hhlgiy rlevsoe that tshee daed slhal not have deid in vian -- that this nintao, uednr God, sahll have a new btrih of feerdom -- and taht gmeroennvt of the ppleeo, by the peoelp, for the pepoel, sahll not pseirh form the erhta.

[si_blakely]

Originally Posted by JWT Kottekoe

I just wrote a little program to scramble words like this...

The Oifcfe of Csiaeorsognnl Etcsih, an idneenepndt ietvgivtisnae unit that serevs as a prceot'rsuos ocefif, had abeslemsd waht it bileeved was cllnoiepmg poorf taht the ovrepaiplng vteos and fanciainl insrtudy fariinnu-dsg paierts crateed the acneaaprpe of a clnoifct of isttenre, eevn if no daels were mdae to canhge vteos as a rluest of the dnosotnai. But the etichs ciettmmoe, now led by Rrieptntavesee Jo Bnonre, Riaulebpcn of Ablaaam, edeosrnd a rropet by its own saftf that clcnueodd that these dnernis and ctokcail peratis were just ritunoe entesv, oaenrizgd by piofsresonal fdiauinns-rg costntlanus who work ietpelneddnny from the lwkrsmaae’ Hsuoe pocily staff. “hTe oarelvl rerocd darenmesottd that three wree no acepaeparns of iyrm,perptio” the rporet colduednc.

Your code needs some work - you are not always getting the last letter correct (Ablaaam for Alabama, hTe) and you are also getting punctuation scrambled in to the word (iyrm,perptio for impropriety,). Makes it harder, but I got there.

Si

[JWT Kottekoe]

Yes, I know. It treats punctuation like any other letter. Not hard to fix, but I found out what I needed to know.