10.19.2005

Jumlbed Txet is Esay to Read

It's been said that people can read jumbled text fairly easily when the middle letters in each word are rearranged, as long as the first and last letters of each word remain the same.

This idea gained attention not long ago from an example floating around the Internet that cited "research at Cambridge University" (acutally, it cited "rscheearch at Cmabrigde Uinervtisy".) That claim appears to be false and is refuted by Cambridge's Matt Davis in a writing posted at www.mrc-cbu.cam.ac.uk/~mattd/Cmabrigde/, in which he dissects the Internet piece and references some real research in this field. There is, however, evidently some truth to this idea, at least in English and other languages where words are similarly structured.

In any case, it's fun to play with. By way of example, here's a jumbled version of the first sentence of this blog entry:

"It's been siad that peolpe can raed jbumeld text farily esliay when the mlidde lrtetes in ecah word are raegarnerd, as lnog as the frsit and lsat lteters of ecah word reiamn the smae."

Did you have much difficulty reading that? Clearly, some jumbled words are easier to read than others. Among other things, the degree of difficulty depends on the length of the word and the extent to which its middle letters are jumbled. In the example above, I find the jumbled version of the word "rearranged" more difficult to read than some of the other, shorter words because it's longer and the letters are pretty well mixed up.

What I find interesting is that jumbles involving the trasnposition of two or even three adjacent letters are so easy to overlook. This is evidently because our brains compensate for the errors in such a way that we don't even notice them (the errors, not our brains).

This is probably one reason why we sometimes fail to see typos in our own work, which of course has implications for software developers. For example, while working in Visual FoxPro I sometimes type 'emtpy' when I mean 'empty', but I usually don't notice it.

BTW, developers using Visual FoxPro 9.0 have an advantage in this regard over developers using previous versions of VFP. The new background compilation feature in VFP 9.0 highlights incorrect syntax as you type by underlining or changing the color of the incorrect syntax. This helps catch errors right away instead of having to wait until compile time or run time. You can choose your preferred style for this feature from the Background Compile drop-down list under Tools | Options | Editor.

Just for fun, I wrote a VFP class called JumbleTron to do this kind of jumbling on any block of text. You can download it from my website at www.ita-software.com/foxpage.htm. Look for the link to JumbleTron under "Just for fun" in the downloads section. By design, the class accepts only 4K-byte blocks of text or smaller, but you can increase this if you want to.

Usage is simple: instantiate a JumbleTron object and pass some text to its GetJumbledText method. For example, copy some text to the clipboard, then in the VFP command window do the following to display the jumbled text on the screen.

ox=NEWOBJECT( "jumbleTron", "jumbleTron.prg")
?ox.GetJumbledText( _cliptext)

Or, to return the jumbled text to the clipboard:

_cliptext = ox.GetJumbledText( _cliptext)

Play around with the JumbleTron and decide for yourself how easy or difficult it is to read jumbled text. Have fun!

P.S. Outside of the deliberate examples, there are two typographical errors in this blog entry. Both are simple transpositions of two adjacent letters. Can you find them? Did they jump out at you as you were reading this, or did your brain take them in stride without notifying you?

1 comment:

Andrew MacNeill said...

Found them - very neat trick.