We have talked periodically about time zones, most recently earlier this year regarding the International Dateline and how 24 zones were originally established, one for each hour of the day, and then each zone was standardized by determining when the sun was above its midpoint, that then being considered valid for the entire zone (2009/7). The need for regularized zones came about because of the need to standardize rail schedules, since faster travel time over considerable distances was suddenly possible. Still, we talk about hours being in 24 time zones, but why do we have hours? And why 24?

It would be useful to investigate, to some extent through logic and conjecture, (1) how the structure of the day most likely came about sequentially, and what nomenclature we use for that structure (“afternoon”); and (2) how, as instruments to measure time developed (sundial, hourglass, clock), we developed, also sequentially, verbal expressions to describe time measurement (“It’s two-thirty”). Here we need to find out how the three units that subdivide the day also developed, hours, minutes, and seconds.

On this latter point, when a church bell rings once, it’s a nonverbal indication. On hearing that, the English speaker will SAY “It’s one o’clock”, the French speaker “Il est une heure”. This is the verbalization of time measurement, which no one seems to investigate, even while we’re being told how the mechanical devices developed. What could the very oldest term possibly be?

I’ve been interested in this for a long time. For year after year over my 28 years in teaching, I would teach beginning German students how to read a clock, and also Spanish students. Then, in the late Eighties and early Nineties, just before my retirement in 1992, I started to teach our seventh graders the Introduction to Computers course, and put together a simple teaching manual for our district, which relied heavily, in those years, on learning how to program using the BASIC computer language. For my own enjoyment, I wrote a program in BASIC to teach telling time in German, Spanish, or French. It was simple, but worked. One would choose a language to work in, and clock face would randomly show a time of day. One would type in words to fill the indicated blanks, and errors would be indicated as they occurred until the phrase was correctly written. This was never put to actual use, and wasn’t really intended to be; it was just a fun project for me, but I did present it to the Language Department during one department meeting. [Beverly was no slacker either at this—she wrote an attractive household budgeting program in BASIC which she actually used for several years to monitor our personal cashflow.]

Decimal, Duodecimal, Sexagesimal   We need to first discuss a little bit of highly simplified math. Everyday decimal calculation centers on 10, 100, 1000, and so on, with 5 as a significant halfway point, so that there are 1000 meters to a kilometer and 100 cents to a dollar, yet coins still appear at the 5- and 25-cent levels.

We also still deal in duodecimals, especially in telling time. Duo- is two, decimal is ten, so a duodecimal system is based on twelve. The halfway and quarter points are at 3, 6, 9, which become significant in time-telling. Above 12, two twelves are 24, five twelves are 60, and six sixties are 360. Six remains one of the significant numbers, and there are elements here of a sexagesimal (based on sixty) system as well, but let’s keep this simple. It should be pointed out that the sexagesimal aspect is essential to geometry. A circle is defined as having 360°, so that a right angle is 90° and a straight angle is 180°. If geometry had been decimalized so that a circle had 100 super degrees, a right angle would be 25 super degrees and a straight angle 50. But that’s not the case.

Neither has measuring time of day been decimalized, although it was suggested when the metric system was being established at the time of the French Revolution. A day would have had, in place of 24 regular hours, 10 super-hours, each being 2.4 regular hours long. Each super-hour would have been broken down into 100 super-minutes. Calculate with me: 2.4 regular hours come to 60+60+24=144 regular minutes, so each super-minute would have been 1.44 regular minutes long. While the super-hours would have been quite a bit longer than we’re used to, a super-minute would merely have been just under a minute-and-a-half.

But this never happened, so we have today the unique situation that familiar units of time are the only non-decimal calculation allowed to interact within metric calculation, such as in kilometers-per-hour (km/h) or meters-per-second (m/s), given that hours and seconds are not decimalized.

The duodecimal system also appears in having 12 items in a dozen; in the UK there used to be 12 pence to the shilling, and in one last remaining unenlightened major country there are still 12 inches in a foot, and, less frequently, 12 dozen in a gross.

The Day   Where must it have all started? Let’s consider it logically, in sequence.

DAY/NIGHT The most fundamental division is between “day”, a positive concept, and then negatively, the left-over time when the sun didn’t shine “night”. Unlike today, given the difficulties of illumination, nighttime activities were minimal to nonexistent.

MIDDAY (NOON) Between day and night we have the borders of sunrise and sunset. For purposes of our discussions here, we’ll standardize those at 6 AM and 6 PM (although we haven’t actually invented hours yet). From there, the next logical thing would have been to divide the working day in half, based on the sun being centered overhead, resulting in words such as midday, GE Mittag, DU & SW middag, FR midi, SP mediodía, IT mezzogiorno, PO meio-dia, RU полдень/polden’. Note that every one of these words is structured the same, and that that the English word “midday” has become far less frequently used than its synonym “noon”.

So saying “It’s noon” (or "It's midday")has to be the oldest designation we have for giving the time of day. Then, just as night follows day—literally--the negatively defined concept of midnight (merely the halfway point between two middays) would probably have been the second oldest.

In passing, we need to talk about the unusual word “noon”, unique to English. It’s related to the word “nine” (surprise!), and originally referred to 3 PM (wow!) and not 12. What are these duodecimals doing to us!? Good question, since the explanation is based 6 AM, so we’ve now got the whole series of 6, 9, 12, and 3.

It’s all based on Roman reckoning and the canonical hours of the Catholic Church, denoting the times of day for prayer. They figured (as we are also doing here) that sunrise was at about 6 AM, so that the “nona hora”, the “ninth hour” of daylight, would fall at about the equivalent of 3 PM, which became “the canonical hour of nones”. In English by about 1140, this developed into the word “noon”, but still referred to 3 PM. But starting in the 12C, the time of church prayers shifted from the ninth hour to the sixth, and the word “noon” moved to midday with this shift, which was completed by the 14C. An additional curiosity is that earlier on, the main meal of the day was also eaten at about 3 PM, and as prayers shifted earlier, so did this meal. Just think of it: a noon meal at 3 PM—which would have made sense at the time—then became a noon meal at 12.

In the modern world, some countries maintain the main meal in the early afternoon, and others have it in the early evening, but having it at 3PM might strike as odd. You wouldn’t expect to find a restaurant busy at that time. But we do indeed maintain that mid-afternoon dining custom when it comes to holiday meals, such as Christmas dinner, or even a traditional “Sunday dinner”. Some customs go back further than one realizes.

(FORENOON)/AFTERNOON We are now at the point when we can consider the day divided into four quadrants: dawn to noon; noon to sunset; sunset to midnight; midnight to dawn. The two daylight quadrants were very possibly named early on. The first quadrant was the forenoon, a word now very rare in English, and the second quadrant was the afternoon, which has remained a common word.

(MORN)/(EVE) We need now to go back to those border times, sunrise and sunset. They each got in English a very specific name, “morn” and “eve”. Both of these words have now been considerably altered and have shifted meaning, but do realize that each referred to the short time period of the sun’s appearance or disappearance, and to nothing more.

MORNING/EVENING Then the strangest thing happened: these two nouns became verbs, then nouns again. As the sun came up, the original expression would have to have been “It’s morn”, and later “It’s eve”. Then two verbs developed: “It’s morning” and “It’s evening”, meaning that it’s becoming morn or eve right at that point. As verbs, these might have appeared as “In the summer it morns earlier and eves later”. Finally these verbal expressions were considered to be nouns once again in the form of “morning” and “evening”, and the shift to the forms we use today was complete. This all happened by about 1250.

Note that “eve” originally ended in an N as “even” (GE Abend still has that N). The N was dropped before 1200 because it was felt it was some sort of grammatical ending, but can still be found in “evensong” and “eventide”. The N returned when “eve” developed into “evening”.

But what happened to the original words? “Morn” pretty much died out, and remains only in poetry, where you might find a reference to a “fair morn”. But what happened to “eve” is unusual. It remains a common word today, but with a radically shifted meaning. It now refers to the previous day, such as “the eve of my birthday” or “New Year’s Eve”. Not only that, it doesn’t just refer to the late part of the previous day, it refers to the entire day, as in “on Christmas Eve we had a big breakfast”.

Then both “morning” and “evening” altered what they referred to, “morning” taking the wilder ride. They both had referred specifically to the few moments of sunrise and sunset. Then “morning”, Godzilla-like, attacked. It spread in meaning to invade the territory of “forenoon”, the first quadrant of the day, from sunrise to noon. Then it went backward into the nighttime to also take over the fourth quadrant, from midnight to sunrise. Today, “morning” covers half the day, from midnight to noon.

The word “afternoon”, in contrast, held on to its territory, the second quadrant from noon to sunset. So we now have to decide on what happened with the third quadrant, from sunset to midnight. Any ideas? Just be careful, since this is also rather odd.

You might be tempted to say that “evening” has tiptoed beyond just sunset, into the third quadrant. Maybe halfway? I don’t agree. I don’t think that “night” yielded entirely. It’s held its ground in the third quadrant and SHARES the territory with “evening” in a very unusual way. Let me illustrate by using the greetings “Good evening” and “Good night” both of which are used in the same third quadrant. But look what’s happened. “Good evening” is used as a greeting, and “Good night” as a farewell. As guests arrive, you greet them referring to the evening, but as they leave, you bid farewell by evoking the night.

But the entire six-hour quadrant is shared equally by the words. Look at this extreme example as proof. If a party arrival takes sick and promptly leaves at 6:30, you say “Good night”, since a farewell is appropriate, even as arriving guests rate a “Good evening”. If a guest has just flown into town and makes it to the end of your party just before midnight, you still say “Good evening”, even though someone else may be leaving at the same time to whom you say “Good night”. The third quadrant is therefore uniquely a dual monarchy in which both “evening” and “night” rule the roost.

In sum, we today consider the day to start, not logically at dawn, but at midnight (???), with the four quadrants of the complete day being morning, morning, afternoon, evening/night.

Hours, Minutes, Seconds   We’ve been so far discussing time periods, the structure of the day, but as for specific times of day with corresponding phrases we have so far only two, “It’s midday/noon” and “It’s midnight”. These must both predate “It’s twelve o’clock”, which must have come later, since it refers to hours.

HOURS Over time, merely dividing the period of daylight into two parts on either side of noon became inadequate, and a further subdivision was desired. The Babylonians had devised dividing the day, duodecimally, into twelfths, but with a difference. They divided the entire day-night sequence into twelfths, meaning their hours were about twice as long as current ones. (What goes around comes around, since that idea seems similar to the decimalized day consisting of hours that are 2.4 current hours long.) The Greeks took the Babylonian concept, but divided only the daylight hours into twelfths, which the Romans also accepted. Why just the day? Possibly because sundials were originally relied upon as the most obvious measurement system. Only later on was the nighttime period finally considered important enough to also subdivide into twelfths. The original word in both Greek and Latin, “hora”, originally referred in those languages very generally to “time, season”, but was adapted into this new concept. The word “hour” finally reached English in about 1250, and has always been spelled with an H, even though there is no H in the word, “hour” and “our” being pronounced alike. (But then there’s really no H in “honor”, either, but that word, too, is just written with one.)

From the way this developed, it was inevitable that, when clocks were invented, the face only showed one set of twelve hours, to be utilized twice, which we clarify by identifying one circuit as AM and the other as PM. However, in modern times, to avoid potential confusion between similar times such as 9 AM and 9 PM, the 24-hour system was invented, changing the distinction to 9:00 versus 21:00. This is used on schedules in most countries, and is even creeping into the US, where I’ve seen it used in places such as phone bills. This system also works well on digital clocks, although the AM and PM system is used there, too. I use the 24-hour system setting on my computer, and on digital watches I’ve owned in the past.

But the numbering system of the hours is totally illogical. If you are measuring a new day, it is logical to start with the beginning of that day, right? If dawn (6 AM) is 0, then 7 AM would be hour 1, noon hour 6, and sunset hour 12. The dark hours would work similarly. But that never happened. Noon was considered 0, so the first hour, 1 PM is in the afternoon. But noon is not the beginning of anything. By definition it’s MIDday, or six hours into the new day. That’s even how the Romans called it, the sixth hour, so that 3 PM was the ninth. This is a very odd numbering system.

With the coming of hours, and given that few people had timepieces, least of all the farmer in the field, we would be in the era of church bells, or municipal bells on city halls, marking these hours. Just denoting noon was no longer sufficient.

I remember something interesting when watching someone bring in a very old clock on Antiques Roadshow. This clock seemed to be damaged, since it had only an hour hand. No so, apparently. The clock was old enough to date from the time when, although hours had been invented, minutes and seconds had not been, hence the one hand.

This brings me to the thought of how we underuse the hour hand on modern clocks that have three. We just look to see if the hour hand is between, say, 1 and 2, then look to the minute hand for additional information. It is perfectly possible on a modern clock to tell time relatively accurately from just the hour hand. You can judge the time quite well if the hour hand halfway between 1 and 2, or ¼ or ¾ of the way, or even 1/3 or 2/3. Maybe we should consider minute hands as being just for wimps.

HALF/QUARTER As we listen to the bells strike the hour, so that we can say, for instance “It’s one o’clock”, let’s consider the tricky question of just what time expressions might have come next. Actually, I’ve just given the answer away. No, it has nothing to do with minutes. When telling time on the hour eventually became not accurate enough, expressions developed such as “It’s half-past one”, “It’s a quarter past one”, “It’s a quarter to (of) two”. These expressions have to predate saying “It’s one-thirty”, since “thirty” refers to minutes, not yet invented. Auditory indications of time would also have advanced, and continue today with different public designations of time, each harking back to one of three traditions: the noon cannon from that era, to hourly bells, to what is now the most modern, hourly bells including the half- and quarter-hours.

Here I’ll make a reference to other languages, but only because this matter affects English as well. I will ask American readers to consider what they think is meant by “It’s half two”. This is an expression existing in Germanic languages, which is also used in Britain, but not in the US. Still avoiding for the moment saying “one-thirty”, since we don’t have minutes yet, “half two” is the same as “half-past one” (GE halb zwei; SW halv två; DU half twee). It might strike the uninitiated as confusing, but simply look at “half two” as meaning “half(way to) two”, and it should become clearer.

MINUTES/SECONDS In time, as more precision was needed, hours were divided into 60 minutes, and then minutes into 60 seconds. This ancient development sifted down to modern languages so that starting with this era we are able to say “ten after one”, and now “one-thirty” joins “half-past one” as an alternate, or “two-fifteen” joins “a quarter past two”. But do realize that there is only one way to write times of day in figures. For instance, if you say “five-twenty”, you logically write 5:20, but if you say “twenty past five”, you STILL write 5:20. Similarly, 7:30 can be read just as written, as “seven-thirty”, but it also can be read as “half-past seven”, or (careful, now) “half eight”.

It is known just how this much more precise and scientific move came about, and was named. It all goes back to Ptolemy, so let’s take a closer look at him. He was a Greek mathematician, astronomer, and geographer who lived in the 2 C of the common era. We don’t usually use his first name at all in English, but it would be Claude, since his name was Κλαυδιος Πτολεμαιος / Klaudios Ptolemaios (often Latinized to Claudius Ptolemaeus). It is significant that he was a mathematician, since he first devised the sexagesimal division of a circle into 60 minutes, and then those minutes further into 60 seconds. It is also significant that he was a geographer and published maps, since to this day, longitude and latitude is given in degrees, minutes, and seconds, because of Ptolemy. This subdividing into 60 minutes, then 60 seconds was only later applied secondarily to telling the time of day.

The words for the concepts sifted down into modern languages, particularly as clocks were finally developed that could actually measure something as small as seconds. The word “minute (MI.nit)” meaning a sixtieth of an hour (or a circle), first appeared in English in 1377. It wasn’t until about 1420 that the related variation “minute (my.NOOT)”, meaning small (related to “minor”) also appeared in English. It took a little longer for “second”, the version meaning a sixtieth of a minute, to appear, in 1391.

But how did these two words come about? A sixtieth of an hour (or circle) was, in Latin, a pars minuta prima, a “first small part” or “first minute (my.NOOT) part”, and a sixtieth of that was a pars secunda minuta, a “second small part” or “second minute (my.NOOT) part”.

It is not a coincidence that minute (MI.nit) coexists with minute (my.NOOT), or that second is both an ordinal number and unit of time. Plainly put, compared to an hour, a minute is a “small (or minute) unit” and a second is the “second level of smallness”. A bit of conjecture about those two three-word Latin phrases brings this “what if” to mind. What if the word minuta had been ignored, since both units involve smallness. What if instead attention had been given just to the numerical words prima and secunda? This might have developed into our calling the smaller units under an hour “primes” and “seconds”, or even “firsts” and “seconds”.

DECIMALIZATION NEVERTHELESS It is highly ironic that all non-decimal time measurements can be, and frequently are, decimalized. These include larger ones (2.4 years; 3.6 months; 49.5 weeks, 8.75 days), as well as the smaller ones we’ve been discussing, such as 9.25 hours (rather than 9 hours and 15 minutes) or 12.5 minutes (rather than 12 minutes and 30 seconds).

There are no time units smaller than seconds. A nanosecond is not a new unit, just a metric subdivision (a billionth) of a second. Also the second (not the hour or the minute) that the standard metric time unit upon which other time measurements are based. With no smaller units, seconds MUST be decimalized, most frequently to three places, such as a stopwatch reading of 25.375 seconds.