May 13, 2009

Numeric Keypads Mystery

Karen asks an intriguing question: Why are numeric keypads grouped into two different designs? Phones and ATMs arrange the numbers from left-to-right in three columns in ascending order top-to-bottom. Keyboard number pads and calculators use a left-to-right ascending, three-column, descending-order arrangement. (You can see some photos she has diligently collected on her blog.)

I'm currently looking through my Donald Norman ( The Design of Everyday Things ) and Henry Petroski ( The Evolution of Useful Things: How Everyday Artifacts-From Forks and Pins to Paper Clips and Zippers-Came to be as They are ) for some reference to the original design of these objects, but nothing yet indicates how they came up with the layouts. Panati ( Extraordinary Origins of Everyday Things , The Browser's Book of Beginnings: Origins of Everything Under, and Including, the Sun ) is not helpful here, either.

The likely answer is going to be that one group of these devices evolved from adding machines and the other evolved from the original push-button phone from Bell Laboratories. If I had to hazard a guess, I'd say that since adding machines predate the push-button phone, early adding machine designs are key (pardon the pun) to present-day calculator layouts. This design would have been carried forward into new devices to take advantage of human experience with the top-to-bottom increasing order. Designers would have not wanted to upset potential customers by changing that layout, and they had no interest in determining whether it was the most efficient.

Take Charles Latham Sholes "qwerty" keyboard as an example. The original design was not the first keyboard design, but it was the most successful because it overcame a mechanical difficulty: the possibility of collision of the typebars, which would lead to jamming. By placing letters which were frequently juxtaposed in English words (like 'i' and 'e') far apart on the keyboard, typebar collision became less likely because the levers were approaching from different angles. With such mechanisms now in our distant past, the "qwerty" keyboard is perpetuated because of human experience with it.

Norman says that Bell Laboratories researched their keypad design (and everything else about the original digital phone) very carefully, and so they were not wedded to an earlier mechanical calculator layout. It is likely that they were also influenced by the fact that rotary phones put the "1" near the top of the dial, while the "9" was near the bottom. If they were not influenced from a design perspective, it's probably that their testers were, and this might have figured into their results.

My guess is that calculator manufacturers did not put much design thought into their keypad layout. Having some knowledge of calculator manufacturers, I have learned that other factors far outweigh human factors in their design consideration process.

Aha! A little more searching has produced this link. Some helpful person has pulled together a few references on Keyboard Trivia. I forgot to check my Feldman! (Why Do Clocks Run Clockwise?: An Imponderables Book )

It seems that there is not a lot of definitive information for why these two keypads are different, but it does boil down to Bell laboratories research finding that people could dial numbers more efficiently with their top-down ascending design. Several people note that Bell Labs scientists were surprised when they asked calculator designers for their prior research on keyboard layout and were answered with "What research?"

Also worth noting: the rotary telephone had already adopted a system where alphabetic characters were assigned to digits. "ABC" are assigned to "2", "DEF" to "3" and so on. If the engineers had rearranged the numbers similar to the adding machines and calculators, they would be rearranging the alphabet as well, because they could not change the associations that were already in popular use as a mnemonic for remembering phone numbers. The new alphabetic order would have been: "PQRS TUV WXYZ GHI JKL MNO [blank] ABC DEF." What a mess!


Did you know that the first private (home) telephone was installed in Somerville, MA?

Posted by James at May 13, 2009 8:57 AM
Create Social Bookmark Links

I knew I could count on you! :-) I still wonder why ATM manufacturers chose to copy telephone keypads instead of adding machines -- maybe because people are inputting passwords that are likely to have letters as well as numbers.

Posted by: Karen at May 13, 2009 9:22 AM

Right - it's likely to allow people to use the same mnemonics. Plus, people are more familiar with telephone keypads than they are with calculators.

And since Bell Labs actually did research before adopting their design, it makes some sense to follow from that research.

Posted by: James at May 13, 2009 9:46 AM

What bugs me is that not only are the keypads for phone and keyboard numerical input different, but there are no letters on the computer keyboard. This becomes significant when certain systems require a numeric-only password. I have an algorithm I developed for creating passwords years ago and am able to create different passwords for everything I use (more secure) without almost ever forgetting them (more useful). When I have to do a numeric-only password I usually just use the phone-letter system, but when I'm logging in I have to actively translate from letters-and-numbers to just numbers on a keypad that is opposite of the phone AND has no letters on the keys.

Posted by: David Grenier at May 13, 2009 10:17 AM

Yes, it is also just plain annoying trying to key in your ATM pin on a computers' numeric keypad.

Posted by: James at May 13, 2009 1:26 PM

Now that I think of it...

My son once asked my why "qwerty" is the way it is. I was tired and in no position to look it up, so I did what every good dad does...gave him a bullshit answer.

Now I need to go back and correct myself.

Posted by: Bull at May 13, 2009 2:15 PM

This design would have been carried forward into new devices to take advantage of human experience with the top-to-bottom increasing order.

I'm sure you're right. I used to use an adding machine every day at a job years ago, and I was damned fast. That translated into my zippy keying with the side numerical keypad on a keyboard. I still can't use the numbers atop the words with remotely the same level of accuracy as I can the side keypad.

Posted by: Patti M. at May 14, 2009 7:23 PM

Of course, what I meant to refer to are the numbers atop the letters on a keyboard--you know, those things that make up the words.

Posted by: Patti M. at May 14, 2009 7:24 PM

I knew what you meant. :)

Posted by: James at May 14, 2009 10:33 PM

Copyright © 1999-2007 James P. Burke. All Rights Reserved