John E. Karlin, Who Led the Way To All-Digit Dialing, Dies At 94

First time accepted submitter g01d4 writes "Who was John E. Karlin? 'He was the one who introduced the notion that behavioral sciences could answer some questions about telephone design,' according to Ed Israelski, an engineer who worked under Mr. Karlin at Bell Labs in the 1970s. And you thought Steve Jobs was cool. An interesting obituary in the NYT."

Re:upside down keypads?

[stevedog]

Apparently he did. From TFA... "The rectangular design of the keypad, the shape of its buttons and the position of the numbers — with “1-2-3” on the top row instead of the bottom, as on a calculator — all sprang from empirical research conducted or overseen by Mr. Karlin."

Re:upside down keypads?

[arth1]

No, the 1-2-3 on the top appears to be due to an R. L. Deininger [vcalc.net], and probably some Bell execs who figured (npi) that it would look better with ABC at the top instead of PRS.

What Mr. Deininger didn't realize was that the industries that already used keypads with higher numbers at the top weren't likely to change.

Calculators started out with 900, 90 and 9 at the top, and going down to 0 at the bottom. Later digital calculators continued with the high numbers at the top, because that's what calculators (the human ones) were used to. So 7-8-9 went at the top.

Similar for cash registers, which really were just narrow purpose calculators, but here there was also a mechanical reason. Registers popped up plaques with the numbers for the customer to see. The designs varied, but generally these were slotted in order from 0-9, with the 0 and 1 closest to the customer, to prevent fraud where the customer would see (and pay) a higher sum that what was entered. Having the low numbers at the bottom meant fewer mechanical crossings.

Then there's the elevator industry. Buildings in general go upwards, not downwards, and placing the top floors, i.e. high numbers at the top was natural.

So there were at least three examples of higher numbers at the top which Bell ignored.

What bothers me is that ATMs also appear to have 1-2-3 at the top. I cannot get this to make any kind of sense, as they're used to enter sums, not mnemonics.
Did AT&T perhaps "help" design early ATMs?

Re:upside down keypads?

[NixieBunny]

Elevators and cash registers did not have 7-8-9 keypads in 1960. Cash registers had 10 keys per digit, and elevators have always had one button per floor.

The only type of machine that had a 7-8-9 keypad was the ten-key machine, used by bookkeepers and accountants to total receipts.

Re:upside down keypads?

[tlhIngan]

Supposedly it is calculator keyboards that are upside down. Two reasons touch tone phones use the order they do:

Touch tone phones replaced rotary phones, which already had 123 at the top of the dial, and 789 at the bottom. So it made sense to keep the same order that millions of people were already used to, in order to make the transition easier.

Touch tone phones have the alphabet sharing the keys, starting with ABC on key 2. Thus the letters are alphabetic from top to bottom, which also properly follows reading order.

Apparently no real research was done in the choice of calculator keyboards having the numbers descending from 9 down. It just happened, and since calculator keyboard layout was more arbitrary (it had neither a predecessor like touch tone phones, nor the alphabet sharing the keys), it would have made sense for calculator designers to match the touch tone phone layout.

I don't know if any studies have been done, but I don't see any reason why one layout would be more intuitive than the other for pure numerical use to a human than the other. It's whatever you get used to. If calculators matched telephones from the beginning then today no one would feel something was inherently wrong with their calculator or that it is upside down from what it should have always been.

Sorta, kinda accurate.

The main reason actually relates to the position of the zero. On a rotary phone, the numbers go 7-8-9-0 (phone phreaks should know that dialing 0 generates 10 pulses - just like 1-9 generate 1-9 pulses, respectively).

On an adding machine and other such hardware, the zero is actually beside the 1-2-3. As at the time the numbers were in a vertical column, you'd see them as 0-1-2-3 ... -8-9.

So when they went to the key pad, the phone engineers decided that since the 0 was besides the 9 on every phone they made, it should stay close to the 9 on the final phone layout. Hence 1-2-3 on top, 7-8-9 on the bottom, and *-0-# on the bottom. (Or on old keypads, 0 aligned with either the 8 or 9).

LIkewise, calculator engineers saw that people who used adding machines expect the 0 to be near the 1-2-3, so they designed their keypads with that in mind as adding machine users expected 0 to be near 1.

And look at your keyboard to this day - the number row reflects the telephone layout (1-2-3 ... -7-8-9-0) while the numeric keypad reflects the calculator layout. Presumably, this was because the typewriter guys saw that the telephone kept the 0 near the 9 so they kept their 0 near the 9 as well (being that more people would've seen a phone at the time than a calculator. I'm certain back in the late 19th century when keyboards weren't standardized on QWERTY and the phone was for rich folks, they probably had 0-1-2-3 just as often as 1-2-3..-9-0.

Re:upside down keypads?

[rkww]

According to your reference, they measured the time taken to dial using the 7-8-9 and the 1-2-3 keypads, and the 1-2-3 was slightly faster: "arrangement I-A had an average keying time of 5.08 seconds, and arrangement IV-A had an average of 4.92 seconds." which is pretty much the point of the article: they measured this stuff.