Stupid Question ™
March 29, 2001
By John Ruch
© 2001
Q: Why do most clocks with Roman numerals use “IIII” to represent 4, even though we all learned in school that it’s supposed to be “IV”?
—Edson Freeman
A: That “IV” we learned in school is a medieval invention. The wide usage of IV postdates the invention of the dial-faced clock, which in its earliest forms always used the original Roman numeral IIII for 4.
The Romans had seven basic numerals: I for 1, V for 5, X for 10, L for 50, C for 100, D for 500 and M for 1,000. To signify a number, the proper signs were arranged in descending order of value and simply added up. Thus, 40 was XXXX; 4 was IIII. A few oddball cases aside, Roman numerals were based exclusively on this “additive principle.”
Halfway through the Middle Ages, Roman numerals were still being used and were increasingly considered awkward. To shorten them, a “subtractive principle” was introduced: placing a small-value numeral in front of a larger-value numeral meant the numbers were subtracted rather than added. Thus, 40 was XL; 4 was IV.
This subtractive principle was introduced in the 1200s, but in the case of IV took another couple hundred years to become widely used. It also competed with a variety of other methods, including one in which 4 was written as iiij.
Meanwhile, the sundial with hours marked in Roman numerals had already long been in use. All such sundials used the one-and-only Roman IIII for 4 o’clock. Sundials directly inspired the design of the first dial-faced clocks in the 1300s, which naturally used IIII as well.
The persistence of IIII on clock faces in the post-IV world is fundamentally a matter of tradition. (Indeed, you can find 24-hour clocks in Europe that use IIII for 4, but XIV for 14 and XXIV for 24.) Clockmakers have reinforced it with a body of wild folklore.
Some attribute IIII to various kingly decrees; to Romans finding IV blasphemous as the first two letters of “Jupiter”; or to the most efficient way of casting iron numerals. Some simply think IIII nicely balances the VIII (8) on the opposing side of the clock face.
Nonetheless, isolated uses of IV are known from as early as the 1480s. After 1700, IV became more common, especially in the US and England (where it appears on the Parliament clock tower housing Big Ben). But it remains a minority.
All of this raises the interesting question of why 9 has always been written in the subtractive form IX (rather than the original Roman VIIII). It’s IX on some of the earliest surviving European sundials and clocks on to the present day without variation.
We can only presume that IX was more rapidly accepted than IV—in time to earn its place on the clock face.
Showing posts with label clocks. Show all posts
Showing posts with label clocks. Show all posts
March 27, 2008
Snooze Alarms Set To Odd Times
Stupid Question ™
Oct. 15, 1998
By John Ruch
© 1998
Q: Why are most digital alarm clock snooze times set to nine minutes instead of a round number like five or 10?
—L.H.
A: Clocks made by Zenith, Sony and market leader General Time Corporation have nine-minute snooze times—except for General Time’s new adjustable-snooze clocks, which offer five-, 10- and 20-minute settings.
Most novelty clocks and build-your-own kits are 10 minutes, as are Emerson clocks. Timex is eight minutes. Canon’s CC-90 alarm clock calculator is five minutes. The 1970s Heath Company digital was seven minutes.
Why?
Corporate answers ranged from suspicious (a design mistake) to ludicrous (“five was too short and 10 was too long”) to hostile (Zenith said consumers are too dumb to understand the real answer).
Professor Kim Boyer of Ohio State University’s computer engineering department says the nine-minute snooze is just simpler and cheaper.
Most digital clocks use a vibrating quartz crystal as a timer. A chip slows the vibration down to 1 hertz, or one cycle per second. The clock turns these cycles into digital pulses and counts them as seconds of time (adding them up for minutes, hours and so forth).
If you want the snooze function to wait 10 minutes, you need it to count to 600 seconds (pulses). The catch is that digital circuitry “counts” with binary numbers. In binary, 600 equals 1001011000.
The circuitry to let the counter recognize the number and decode it into decimal takes “a fair amount of space on the chip, when the fact is this is not a very critical timing event,” Boyer said.
So instead, you pick a number of pulses that’s pretty close to 600 and that is also easily read in binary. The closest is 512 pulses, which equals eight minutes and 32 seconds, and which in binary equals 1000000000—a big number, but one that can be recognized simply by waiting for the number in the farthest left column to turn from 0 to 1 for the first time. That takes less analysis and less chip space.
This probably explains the early National Semiconductor clock chips that, in the fine print, said the snooze time was “between eight and nine minutes.” Probably explains Timex, too.
But my own clock goes off exactly nine minutes later. Which leads us to an explanation provided by the reader-answered “Collected Wisdom” column in the Toronto Globe and Mail, which tackled this question in August 1996.
It said the simplest way to run a snooze counter is to tell the alarm to ring again when the last digit in the time (say “9” in 12:09) is one less than it is now (i.e., nine minutes later). If you told it to go off when the last digit is the same, it would go off immediately.
Timex tells me a researcher for “Parade” magazine’s “Ask Marilyn” column recently called with the exact same question. What answer will Marilyn come up with? Time will tell.
Oct. 15, 1998
By John Ruch
© 1998
Q: Why are most digital alarm clock snooze times set to nine minutes instead of a round number like five or 10?
—L.H.
A: Clocks made by Zenith, Sony and market leader General Time Corporation have nine-minute snooze times—except for General Time’s new adjustable-snooze clocks, which offer five-, 10- and 20-minute settings.
Most novelty clocks and build-your-own kits are 10 minutes, as are Emerson clocks. Timex is eight minutes. Canon’s CC-90 alarm clock calculator is five minutes. The 1970s Heath Company digital was seven minutes.
Why?
Corporate answers ranged from suspicious (a design mistake) to ludicrous (“five was too short and 10 was too long”) to hostile (Zenith said consumers are too dumb to understand the real answer).
Professor Kim Boyer of Ohio State University’s computer engineering department says the nine-minute snooze is just simpler and cheaper.
Most digital clocks use a vibrating quartz crystal as a timer. A chip slows the vibration down to 1 hertz, or one cycle per second. The clock turns these cycles into digital pulses and counts them as seconds of time (adding them up for minutes, hours and so forth).
If you want the snooze function to wait 10 minutes, you need it to count to 600 seconds (pulses). The catch is that digital circuitry “counts” with binary numbers. In binary, 600 equals 1001011000.
The circuitry to let the counter recognize the number and decode it into decimal takes “a fair amount of space on the chip, when the fact is this is not a very critical timing event,” Boyer said.
So instead, you pick a number of pulses that’s pretty close to 600 and that is also easily read in binary. The closest is 512 pulses, which equals eight minutes and 32 seconds, and which in binary equals 1000000000—a big number, but one that can be recognized simply by waiting for the number in the farthest left column to turn from 0 to 1 for the first time. That takes less analysis and less chip space.
This probably explains the early National Semiconductor clock chips that, in the fine print, said the snooze time was “between eight and nine minutes.” Probably explains Timex, too.
But my own clock goes off exactly nine minutes later. Which leads us to an explanation provided by the reader-answered “Collected Wisdom” column in the Toronto Globe and Mail, which tackled this question in August 1996.
It said the simplest way to run a snooze counter is to tell the alarm to ring again when the last digit in the time (say “9” in 12:09) is one less than it is now (i.e., nine minutes later). If you told it to go off when the last digit is the same, it would go off immediately.
Timex tells me a researcher for “Parade” magazine’s “Ask Marilyn” column recently called with the exact same question. What answer will Marilyn come up with? Time will tell.