they didn’t have minutes or seconds

Measure of Time

Tarin: The part about dates and time. I don’t know if universally, but in the US, the rail system prompted the time zones. Otherwise trains would crash. When did we decide what day it is, and when was? I have a book about the invention of yesterday, and that’s the news.

Ilex: Yes—and your page points to an important distinction: knowing which day it is locally isn’t the same as giving that day an address everyone else recognizes. There wasn’t one worldwide decision. Several agreements accumulated.

You’re right about American railroads. On November 18, 1883, North American railroads adopted Standard Railway Time. Previously, local and railroad times differed; coordinating them was essential for schedules and safety. The railroads standardized the clock, not the calendar. [1]

For the calendar, there were several separate questions:

Which day and month? Local calendars, often tied to the Moon and seasons. Your passage says Greek communities could have different calendars—even more than one within a city.

Which year? A ruler’s reign or another locally meaningful reference. Egyptian documents, for example, counted years of a king’s reign. [2]

Where does the year count begin? Our AD numbering originated with the sixth-century scholar Dionysius Exiguus, working on Easter tables. It was a retrospective reference point—not a count everyone had been keeping since “year one.” [3]

Which calendar should we share? The Gregorian reform began in 1582, but adoption was staggered. Britain and its colonies changed in 1752. [4]

An example makes the distinction tangible: in the initial Gregorian changeover, Thursday, October 4, 1582, was followed by Friday, October 15. No lived days disappeared. The date labels jumped; the sequence of weekdays continued. [5]

That also answers your “when was?” Ancient people didn’t call their own year “480 BC.” Historians place their records onto our timeline afterward. Your author’s 481/80 BC example means a community’s year crossed parts of two of ours—not necessarily that historians are uncertain by a year.

So there are really two achievements here: agreeing how to meet tomorrow, and working out how everyone’s yesterdays fit together. The second requires translating between calendars, reigns, records, and shared events. A single timeline in a history book can conceal all that reconstruction.

Tarin: Thursday, October 4, 1582, was followed by Friday, October 15. This is when we started counting, either backwards or forwards to tomorrow.

Ilex: It was a recalibration of an existing count, rather than the beginning of counting. They already numbered days and years, and could date both past events and future appointments.

In the places adopting the reform then, people went to bed on Thursday, October 4, and woke the next morning on Friday, October 15. One night passed. The calendar advanced ten extra dates to correct its accumulated drift against the seasons. [5, 6]

But you’re identifying something real: once we establish a calendar and a reference point, we can extend that numbering backward and forward. We can even assign dates to times before that calendar existed.

The distinction is between when an event happened and the date we assign it. October 1582 changed the labels—not the succession of yesterdays and tomorrows.

Tarin: Ah, I missed the ten days.

So the text says we had days. We could count back and forth. But people didn’t agree on the names, or the number of days per period. And some people could understand both.

Ilex: Yes—that’s the distinction. They could count days without sharing the same calendar.

The differences weren’t just names. Communities could disagree about when a month began, when the year began, and when to add an extra month to keep the calendar aligned with the seasons. Your passage describes those local systems.

And yes, someone familiar with two systems could relate them—like translating between languages, although the calendar conversion wasn’t always straightforward.

So “we’ll meet in five days” could be understood without a shared calendar. “We’ll meet on the tenth day of this named month” required knowing whose calendar.

The difficulty for the historian is recovering that correspondence afterward, when some of the information needed to translate it has been lost.

Tarin: And then they didn’t have clocks, so they didn’t have minutes or seconds.

Ilex: Not quite—they didn’t have mechanical clocks, but the ancient Greeks did have sundials and water clocks. A shadow could indicate the time of day; flowing water could measure an interval, even without sunlight. [7]

Minutes and seconds introduce another distinction: being able to define a small division isn’t the same as being able to measure it reliably. Those divisions have ancient astronomical roots, initially used to subdivide angles—not as the familiar ticking units of everyday clocks. [8]

So your thought works better as: they could arrange and measure time, but not with our everyday access to minute-and-second precision. “How long?” and “At exactly what time?” were different practical problems.

Tarin: A water clock would have ticks to count. Not a sundial. A sand clock has lines, but it’s a reflection of minutes. Did they all agree on a 24-hour day? These different calendars were called different names but they were all days?

Ilex: Yes—the different calendars counted days: the recurring cycle of daylight and night. They organized and named those days differently. That didn’t require agreement about hours.

But 24 equal hours wasn’t a universal starting agreement. In ancient Egypt, daylight was divided into twelve hours and nighttime into twelve. That meant a summer daylight hour was longer than a winter daylight hour. Hellenistic astronomers later used equal-length hours. [2, 9]

And by “ticks,” if you mean marks on a scale, yes—a water clock could have those. It didn’t necessarily tick: you watched the water level move past the marks. A sundial also had marks, crossed by the shadow. A sandglass could simply measure one interval by emptying, without any lines. [7, 10]

The marks don’t inherently mean minutes. They mean whatever interval the instrument was calibrated to represent.

So your underlying distinction holds: the day was the shared natural cycle; its subdivisions, boundaries, and calendar address were human conventions.

Tarin: But minutes are degrees of something about the sun on the horizon. Then that means time passed and it’s what’s in between times.

Ilex: Yes—you’re connecting a change in position with the time elapsed between two positions.

One distinction: an angular minute is 1/60 of a degree; a time minute is 1/60 of an hour. They share the division into sixtieths, but measure different things.

Using the Sun’s average apparent motion around the celestial pole:

360° corresponds to 24 hours.
15° corresponds to one hour.
1° corresponds to four minutes of time.

That isn’t simply its height above the horizon, which changes differently depending on latitude and season.

But your central thought is right: one observation marks a moment; the change between observations lets us measure an interval. A sundial makes that visible through the shadow’s movement. The clock reading says when; the difference between readings says how long.

Tarin: Yes, and when or how did that movement become a measure of time, of that determined length? And the hour to have sixty of them. It’s like streaming. It was something else then, and now is watching on a tablet via the internet.

Ilex: Yes—your “streaming” analogy helps: you’re asking how a technical idea became the familiar thing we mean by the word today. I was separating angle from time without explaining that bridge.

It happened in stages:

First, a repeatable cycle supplied the reference. From one solar noon to the next is a day. Divide that cycle into 24 equal parts, and you have a fixed-length hour. Hipparchus proposed equal hours in the second century BC, though seasonal hours remained in ordinary use much longer. [11]

Then mathematics supplied the sixty. Babylonian astronomy used base-60 arithmetic. An hour could therefore be subdivided into sixty parts, then each part into sixty again. Sixty wasn’t something the Sun required; it was an inherited numerical convention, convenient for fractions. [9]

The names described the subdivisions. In the Latin terminology, minute meant the first small division; second came from the second subdivision. The same scheme could divide either a degree or an hour. [11]

Finally, instruments made those divisions everyday experience. Clocks displaying minutes appeared near the end of the sixteenth century. A mathematical subdivision became something you could read on a dial and arrange your activities around. [11]

So the minute’s length came from a fraction of an established cycle, not from discovering a separate natural event that lasted exactly one minute. Then clocks let us carry that fraction indoors, through the night, and eventually on our wrists.

Tarin: But you say the clock displays the minutes, but the made-up minutes, not the minutes in the angle of the sun and the horizon. It’s like, the solution of the equation for the particle in the box is a bell curve. And we imagine the particle making the shape of the curve. It’s what we see when we say minutes, a clock dial with ticks. But do they translate to the sun on the horizon?

Ilex: Yes—the clock’s marks display chosen units of duration, not little angles of the Sun’s height above the horizon. My earlier explanation blurred that distinction.

They can translate into solar position, but through a calculation—not a direct one-to-one reading:

One clock minute corresponds to about ¼° of mean solar hour angle—progress around the celestial pole, measured relative to your local meridian.

It does not mean the Sun rises or falls ¼° above the horizon. That change depends on your latitude, the season, and the time of day. Near solar noon, its height changes very little even while time continues passing.

And the minute hand itself moves 6° per minute. That angle belongs to the clock’s gearing and display, not to the Sun.

So your point about the representation is exactly relevant: the shape of the display isn’t the shape of the physical process. A clock could display numbers without any circle at all.

One physics adjustment: the standard infinite-box ground state is sinusoidal, with a sine-squared probability density, rather than a Gaussian bell curve. But your distinction holds: that plotted density is not the particle’s trajectory, just as the minute hand’s path is not the Sun’s path.

Tarin: So they are different minutes, the ones on the clock and the ones on a globe.

Ilex: Yes. Same name, different units.

Clock — A minute means: A duration: 60 seconds. Sixty of them make: One hour.

Globe — A minute means: An angle: 60 arcseconds. Sixty of them make: One degree.

Both mean a sixtieth, but a sixtieth of different things.

They can be related through Earth’s rotation, but they aren’t interchangeable. I should have made that explicit earlier.

Tarin: Back to Greece and Sparta. How did they see time then? They didn’t have clock faces, so only one minute or both types of minutes?

Ilex: For an ordinary person in that period, probably neither was a familiar unit. But they could still measure how long something took—or decide how long it should last. A water clock could measure an interval without anyone naming it in minutes.

Tarin: In hindsight? That much time had passed?

Ilex: Or beforehand: this much time is allowed. When the water runs out, you stop.

Tarin: So they had a practical measure. What they didn’t have was my way of naming it.

Ilex: Yes. And we should keep two questions separate: what their instruments could measure, and what units people ordinarily used. My “neither” made it sound as though they had neither.

Tarin: Did they have church towers then? Bells?

Ilex: Bells, yes. Church towers, no—not in fifth-century BC Athens and Sparta. This was centuries before Christianity; their religious buildings were temples and sanctuaries.

We have a specific example of a bell being used: Thucydides describes a bell carried between sentries at Potidaea during the Peloponnesian War. It accompanied a round checking the posts; the passage doesn’t establish that it signaled a change of watch. [12]

If you’re picturing a sound everyone could hear that told them what time it was, the familiar church clock striking twelve belongs to a much later setting. A bell can coordinate people—“act now,” “assemble,” “the watch is passing”—without giving a numbered time of day.

Tarin: How were watches organized? In Sparta for example. How did they know when one ended? These rounds. Someone had a water clock.

Ilex: Yes—a water clock was one way to determine when a watch ended. Your suggestion is explicitly supported by an ancient Greek military manual, though I can’t establish it as the particular Spartan practice in your book’s period.

Aeneas Tacticus, writing in the fourth century BC, recommends measuring watches with a water clock so the night’s duty is divided fairly. He even describes adjusting its capacity with wax as nights lengthen or shorten. So the aim was equal shares of that night, not necessarily a fixed number of our minutes. He also describes changes of watch coordinated by signals. [13]

In practical terms, someone could attend the timer and initiate the change; the sentries themselves didn’t each need a clock.

For Sparta specifically, Xenophon describes night sentries outside the camp, armed rounds, and orders to relieve outposts. But that passage doesn’t say how the night-watch intervals were timed. [14]

Tarin: So they didn’t have clocks but timers?

Ilex: They had both. But a timer is what you’re asking for when you ask how a guard knew his watch was over.

Think of three questions. A timer answers, “Has this much time passed?” A stopwatch answers, “How much time has passed since I started?” A clock answers, “What time is it?”

Tarin: A stopwatch is a timer inverted.

Ilex: Yes. One begins with a duration you choose; the other measures the duration you give it. Neither needs to know what time of day it is.

A clock needs a reference. A sundial gets its reference from the Sun. Our clocks belong to an agreed system. All three devices can use hours and minutes, but a sandglass can simply give you one glassful.

Tarin: And the sundial doesn’t need a theory before it works.

Ilex: The shadow moves anyway. You learn to read it. Explaining why it moves can come later.

Tarin: So the guard doesn’t need to know how many minutes.

Ilex: He needs to know when he can leave.

WE&P by: EZorrillaMc&Co.

Sources

[1] Library of Congress — Today in History: November 18⁠

[2] Metropolitan Museum of Art — Telling Time in Ancient Egypt⁠

[3] U.S. Naval Observatory — The 21st Century and the 3rd Millennium⁠

[4] Massachusetts — Happy New Year?⁠

[5] U.S. Naval Observatory — Julian Date Converter⁠

[6] U.S. Naval Observatory — Leap Years⁠