Join 3,380 readers in helping fund MetaFilter (Hide)


Dial D for Design
June 8, 2010 10:46 AM   Subscribe

The question facing Chapanis and lab assistant Mary C. Lutz was deceptively simple: What should a push-button telephone look like?

In 1954, industrial psychologist Dr. Alphonse Chapanis, Mary C. Lutz, and Engineer D.L. Deiniger were assigned by Bell Labs to design a commercial push button phone keypad to replace the rotary dial. The team experimented with 16 keypad lay outs before settling on the one used today. Deiniger's paper on the process (PDF), published in 1960, became a classic in Human Factors Engineering.
posted by chrisulonic (28 comments total) 16 users marked this as a favorite

 
I've always found it curious that the arrangement of buttons on a phone is arranged top-to-bottom, while my computer's numeric keypad is the opposite (though both place zero under the 3x3 grid).
posted by The Winsome Parker Lewis at 10:58 AM on June 8, 2010 [2 favorites]


It seems crazy they didn't just use the numeric 10 key layout which dates to 1914. The journal article says they tested it and it was a bit slower than the telephone layout we use now. I guess numeric 10-key was enough of a specialist skill they didn't see any value in being consistent.
posted by Nelson at 11:04 AM on June 8, 2010


The first time I ever saw a TouchTone phone was at HemisFair '68.
posted by Daddy-O at 11:06 AM on June 8, 2010


I thought I read somewhere that the reason the phone keypad is different from a regular numeric keypad is because the tone system in the phone couldn't work as fast as people could use a numeric keypad. When they tested the phone with the setup with 1-3 on the bottom row, subjects could hit the buttons faster than the tones could make the call. So they reversed it to slow it down.

No idea where I read that, and it could be total bunk, but it sounds plausible.
posted by nushustu at 11:11 AM on June 8, 2010


I love the kind of design that isn't just making logos look glossy.
posted by DU at 11:16 AM on June 8, 2010 [1 favorite]


If you think her design was suboptimal, you should see the one Mary C. Lulz came up with.
posted by Shepherd at 11:20 AM on June 8, 2010 [7 favorites]


When they tested the phone with the setup with 1-3 on the bottom row, subjects could hit the buttons faster than the tones could make the call.

The order (rather than layout of the physical buttons) being a a major factor makes sense for something like a typewriter or cash register because the input is generally nonrandom, but isn't the digit distribution of telephone numbers relatively even? Aside from things like dialing 0 for operator or 1 before the number, I would think most of the digits entered by most phone users would be random enough that making certain buttons slightly easier to press would do next to nothing to improve entry times.
posted by burnmp3s at 11:24 AM on June 8, 2010


Someone didn't get the memo.
posted by mazola at 11:45 AM on June 8, 2010 [1 favorite]


They designed that in the mid-fifties but I never actually had a touch tone phone until the early eighties, almost thirty years later. I realize that my family was a late adopter but that's still an amazingly long roll out period for technology.
posted by octothorpe at 11:52 AM on June 8, 2010


And yet they still got it wrong.
The damn thing is upside-down, and that's all there is to it.
posted by SLC Mom at 12:05 PM on June 8, 2010 [1 favorite]


What always bugged me was the letters: If you're omitting letters, why Q and Z? Sure, they might not be used much, but it would make more sense to leave out O and I, which are a bit redundant when there's a 0 and a 1.
posted by Sys Rq at 12:26 PM on June 8, 2010 [1 favorite]


Why have letters at all? They should've followed up # and * with more punctuation and required everybody to spell in 1337$ΒΆ3@|<.
posted by The Winsome Parker Lewis at 12:32 PM on June 8, 2010


To expand on burnmp3s point, number keypads for calculators or cash registers are designed to optimize on numbers with lots of zeroes and ones (not to mention decimal points). Fitts' law suggests that putting the zero and one keys close together (and making them larger) would pay off, and experienced users would be able to cope with any confusion.

For a phone, though, the digits are evenly distributed, 1 is a rare/special number (long distance), and the users are novices. Ease of use is by far the most important criteria.

Good to hear some retrospectives on Chapanis, he's a real founding father of human-centred engineering.
posted by anthill at 12:34 PM on June 8, 2010 [2 favorites]


Now I want to know "why B-17 bombers kept crashing on the runway in World War II."
posted by gottabefunky at 12:35 PM on June 8, 2010


Now I want to know "why B-17 bombers kept crashing on the runway in World War II."

I don't know, but that's a great way to memorize 217-9944.
posted by Sys Rq at 12:39 PM on June 8, 2010


Error free dialing!*

---
* WTF?
posted by mazola at 12:45 PM on June 8, 2010


What always bugged me was the letters: If you're omitting letters, why Q and Z? Sure, they might not be used much, but it would make more sense to leave out O and I, which are a bit redundant when there's a 0 and a 1.

The letters originally weren't for spelling your company name with your phone number, as we do today. They were for specifying the exchange with a word instead of two numbers, presumably to make it easier to memorize.
Where I grew up, the exchanges were Jackson (JA, or 52) and Fulton (FU or 38). In movies and TV shows, the fake exchange was Klondike-5 (KL5 or 555).
posted by rocket88 at 12:48 PM on June 8, 2010 [5 favorites]


Now I want to know "why B-17 bombers kept crashing on the runway in World War II."

Apparently the methods to retract the flaps (a normal part of landing) and retract the wheels (a good way to seriously damage the aircraft during landing) were nearly identical. From this site:

In 1943 Lt. Alphonse Chapanis was called on to figure out why pilots and copilots of P-47s, B-17s, and B-25s frequently retracted the wheels instead of the flaps after landing. Chapanis, who was the only psychologist at Wright Field until the end of the war, was not familiar with the ongoing studies of human factors in equipment design. Still, he immediately noticed that the side-by-side wheel and flap controls-in most cases identical toggle switches or nearly identical levers-could easily be confused. He also noted that the corresponding controls on the C-47 were not adjacent and their methods of actuation were quite different; hence C-47 copilots never pulled up the wheels after landing.

Chapanis realized that the so-called pilot errors were really cockpit design errors and that by coding the shapes and modes-of-operation of controls the problem could be solved. As an immediate wartime fix, a small, rubber-tired wheel was attached to the end of the wheel control and a small wedge-shaped end to the flap control on several types of airplanes, and the pilots and copilots of the modified planes stopped retracting their wheels after landing. When the war was over, these mnemonically shape-coded wheel and flap controls were standardized worldwide, as were the tactually discriminable heads of the power control levers found in conventional airplanes today.


It's basically the same reasoning behind Murphy's Law, the point of which is that if you don't want people to make mistakes, you should make it difficult or impossible for them to do so.
posted by burnmp3s at 12:51 PM on June 8, 2010 [1 favorite]


Sometime in my design career, I realized that good design can be completely unremarkable. When something works well, you just don't notice that it works well - it just works. It's bad design that you notice, because it annoys or frustrates you every time you encounter it.

My ex wife became irritated with me after I pointed out why several local parking lots didn't work well. After that, she was ex post facto frustrated with the designer, but took it out on me. Yeah, shoot the messenger.

On a side note (perhaps better called a derail), T9 can really piss me off. Why in the world will it default to garbage random letter combinations, when there are perfectly good words that I want to use? And when in the world did "foot" and "font" take higher priority over "dont" ?? Yes, I know there should be an apostrophe in there, did you ever really try to use contractions in word mode?

AND PLEASE FOR THE LOVE OF GOD, LET ME DELETE "Yo8" FROM THE DICTIONARY WHEN I TRY TO WRITE "you".
posted by Xoebe at 1:01 PM on June 8, 2010


I'm just so insanely happy that a wonky "2" key (or a fat finger) means that dialing 9-1-215-... from the office results in a 911 call. It's the ejector seat analogy at work.
posted by kurumi at 1:22 PM on June 8, 2010


I remember seeing a phone with buttons arranged in a circle, with each number button located where the same number used to appear on a dial phone.
posted by Chocolate Pickle at 1:53 PM on June 8, 2010


"I'm just so insanely happy that a wonky "2" key (or a fat finger) means that dialing 9-1-215-... from the office results in a 911 call."

This is why our phone system is programmed to use "8" as the number to dial for an outside line.
posted by mr_crash_davis mark II: Jazz Odyssey at 2:10 PM on June 8, 2010 [2 favorites]


I'm just so insanely happy that a wonky "2" key (or a fat finger) means that dialing 9-1-215-... from the office results in a 911 call.

Yes, but that's 9-9-1-1...
posted by pointless_incessant_barking at 2:13 PM on June 8, 2010


Xoebe wrote: "Yes, I know there should be an apostrophe in there, did you ever really try to use contractions in word mode? "

Yes, it works very well, at least with the Nokia implementation. If you type "36618", you'll get "don't" by default. It also reorders the word list based on how frequently you use particular words. Some brands have a better T9 implementation than others.
posted by wierdo at 2:45 PM on June 8, 2010


I can't believe that they didn't try a grid layout that starts with zero. Just because it came after nine in the rotary world doesn't imply it should in they keypad world, did it?
posted by CheeseDigestsAll at 8:04 PM on June 8, 2010


I remember seeing a phone with buttons arranged in a circle, with each number button located where the same number used to appear on a dial phone.

I still have one of those. Bought it in 1981. Still works a treat.

I personally have always wanted to hack one of those old-fashioned plastic console-like baby toys into a phone dial. You know: you would punch a big red button for 0, twist a knob for 1, pull a lever for 2, slide a bead for 3, etc.

That, or one of those toy xylophones.
posted by trip and a half at 10:40 PM on June 8, 2010


I always thought it would be an easy trick in college, when someone was standing by the campus phone asking, "What's her extension?" to say "ninety-nine eleven."

(Yeah, I went to college before cell phones and before the phrase "nine eleven" was instantly recognizable in everyone's head.)
posted by straight at 8:02 AM on June 9, 2010


Yes, it works very well, at least with the Nokia implementation. If you type "36618", you'll get "don't" by default. It also reorders the word list based on how frequently you use particular words. Some brands have a better T9 implementation than others.

My biggest pet peeve from the T9 world is on my old RAZR - it also does the reordering of the word list based on how frequently you use particular words... But in the case of, for example, "on" and "no", which I use approximately equally, it is constantly swapping them back and forth. So I am constantly mixing those two up. There's another couple words that it does that for me, but that's the one that pops into mind immediately.
posted by antifuse at 9:25 AM on June 9, 2010


« Older Supreme Court Blocks Arizona Campaign Finance Syst...  |  Why are so many recent Young A... Newer »


This thread has been archived and is closed to new comments