QWERTY stops jamming?

[Workoft]

Hi,

I've switched to Colemak a while ago and I love it. Most people know that the reason for QWERTY is not ergonomics. I've read a couple of things why QWERTY is the way it is, the most commonly quoted one being that QWERTY places the most common bigrams far apart to stop typwriters from jamming. This sounds like a reasonable explanation, until you actually check for the most frequent English bigrams. TH is the most frequent one, but the letters are fairly close. The fourth and sixth most frequent ones are ER and RE, respectively. ES and ED are also rather common.
If the claim that QWERTY arranges the keys to minimize jamming is true, why aren't these freqent bigrams further apart?

Cheers,
Workoft

[Tym]

I think it was designed in a time when people were less intelligent ?

[TotalChaos]

QWERTY layout was designed to make ppl type more slowly than a good layout.  QWERTY was not meant to be absolutely perfect at slowing ppl down.   In order to achieve your perfect bigram suppression this would have taken actual work and studying to achieve.  They simply wanted to throw something together quickly without having to do much actual work.

Colemak is what you get when someone spends actual time studying the concept of typing.  Colemak is what you get when someone actually cares.  Colemak is what you get when you cheat and use computers to help you analyze typing. 

COLEMAK Forever! 

[jwaz]

COLEMAK Forever! 

word!

It's a gnarly process to switch but worth it.

[Workoft]

QWERTY layout was designed to make ppl type more slowly than a good layout.

This, I've heard, is a myth. I heard it was placing frequent bigrams apart so you could actually type faster without jamming. It doesn't really make much sense, either. Why would someone need the keyboard layout to make them slower if they know the typewriter would jam if they type too fast. QWERTY may be evil, but it's not that evil.

[Proword]

It's not a myth.  This academic paper, presented to a conference of the Printing Industry Research Association (PIRA) in September 1977 is quite  up front about the design of the QWERTY layout, in the third paragraph of the introduction.

http://www.maltron.com/media/lillian_kditee_001.pdf

"...a team of people spent one year developing this [QWERTY] layout so that it should provide the greatest inhibition to fast keying ..."

Joe

[davkol]

Interesting, but... ss there any _original_ sauce for that information? Or yet better, an analysis of the layout? Because it appears those guys did a poor job whatever was their goal.

[Proword]

Did you read the entire paper?  I only linked to the first page.

http://www.maltron.com/keyboard-info/academic-papers.html

The "sauce" for the information is  the Printing Industry Research Association and the paper presented to it.

https://www.smitherspira.com/About-us.aspx

Though it's got a new name, 

"Smithers Pira was originally established in 1930 as PIRA, the initials standing for Printing Industry Research Association. The aim of the organisation was to be "a technical research bureau for the pooling of technical information and to conduct scientific investigation of technical problems" for the printing industry."

I know of no other organisation, past or present, with such a brief.  In other words these are THE experts.

Since this is an academic paper, it must have been refereed (anonymously) before presentation, by qualified people in the field, and if they had any difficulties with it, it would not have been presented.  Published in 1977, I first read it in 1986, before I bought my first Maltron.

Joe

[Proword]

I should add here that if you read my initial post, I was simply answering the implied question as to whether it was a "myth".  I'm not arguing as to the accuracy or otherwise of Malt's statement.  I've never been able to understand why people devote so much time to the reason behind the design of QWERTY.  There is almost universal agreement that it's not a good design - end of story for my money.

Albert Einstein once remarked to the effect that no amount of experimentation could prove him right; a single experiment could prove him wrong.  I'd submit that Malt's paper is the single experiment which proves that the claim is NOT a myth - it has academic backing.   

Joe

[davkol]

Only quickly, thanks anyway.

[tp4tissue]

QWERTY layout was designed to make ppl type more slowly than a good layout.  QWERTY was not meant to be absolutely perfect at slowing ppl down.   In order to achieve your perfect bigram suppression this would have taken actual work and studying to achieve.  They simply wanted to throw something together quickly without having to do much actual work.

Colemak is what you get when someone spends actual time studying the concept of typing.  Colemak is what you get when someone actually cares.  Colemak is what you get when you cheat and use computers to help you analyze typing. 

COLEMAK Forever! 

Colemak has a fatal flaw just like most other layouts..

namely the HOME ROW is NOT resting position for most fast typists   awefjio' are

So if anything they should've moved awft nuyo  into those positions...


There's not a huge improvement one could make with Layouts, because ultimately you're not limited by travel distance, rather the amount of "burst" muscle memory you've accumulated.

[jwaz]

QWERTY layout was designed to make ppl type more slowly than a good layout.  QWERTY was not meant to be absolutely perfect at slowing ppl down.   In order to achieve your perfect bigram suppression this would have taken actual work and studying to achieve.  They simply wanted to throw something together quickly without having to do much actual work.

Colemak is what you get when someone spends actual time studying the concept of typing.  Colemak is what you get when someone actually cares.  Colemak is what you get when you cheat and use computers to help you analyze typing. 

COLEMAK Forever! 

Colemak has a fatal flaw just like most other layouts..

namely the HOME ROW is NOT resting position for most fast typists   awefjio' are

So if anything they should've moved awft nuyo  into those positions...


There's not a huge improvement one could make with Layouts, because ultimately you're not limited by travel distance, rather the amount of "burst" muscle memory you've accumulated.

That is a problem with keyboard staggering not with the Colemak layout. See ergodox, kinesis etc.

[tufty]

Albert Einstein once remarked to the effect that no amount of experimentation could prove him right; a single experiment could prove him wrong.  I'd submit that Malt's paper is the single experiment which proves that the claim is NOT a myth - it has academic backing.

Lots of things have academic backing, doesn't make them true.

Consider this.  At the time where QWERTY was introduced back in the 18<mumbly>s, there were many keying layouts, probably the most common of which was alphabetic.  What mattered was /not/ slowing down trained typists so they wouldn't jam the machines,  but to make the machines such that trained typists could type fast without jamming them. After all, time is money, and if you're paying a typist x amount per hour, you want them to be typing *more*, not less.

What QWERTY did was to make it such that the typebars didn't jam (as much) by moving the most common bigrams apart far enough that collisions were less common. As such, it's been anachronistic since the introduction of the golfball typewriter...

Sholes' patent is here.  You'll note the lack of comments regarding "slowing down typists".

[hoggy]

On a side note, the literature on the qwerty layout is peppered by people who just wanted to defend the 'market knows best' mantra.   

[jabar]

On a side note, the literature on the qwerty layout is peppered by people who just wanted to defend the 'market knows best' mantra.

shh... *"invisible hand" slaps hoggy link*

wikipedia, despite being usually terrible, has a good writeup of the history behind the original Sholes and Glidden typewriter layout: http://en.wikipedia.org/wiki/Qwerty

[Proword]

Lots of things have academic backing, doesn't make them true.

That's a pretty sweeping statement.  How is it relevant?

Joe

[sordna]

Among QWERTY's design considerations, was the requirement to be able to type TYPEWRITER on the top row :-)

[Proword]

With a single finger as I understand. )

Joe

[Hubbert]

During my research I saw a claim in a couple of places that Qwerty is no better than a “random”
layout. This is not a fair claim.
The most uncommon letters in English are PBKVXJQZ. In Qwerty's defense, note that of those
8 letters, BVXZ are on the bottom row, PQ are in the corners of the keyboard, which is not bad, and
only J and K waste important real estate. Only one very common letter, N, is on the bottom row.
Qwerty is better than an alphabetic layout which is effectively random (the order of the alphabet itself
is random, and even if it weren't, there's no reason to presume efficiency when you put it on a
keyboard.)

...
http://millikeys.sourceforge.net/asset/why-qwerty.pdf