In OTL, time zones were introduced in the late 19th century when time tables for long-distance trains, for instance, began to require more precision and standardization across wider areas than a sundial could provide. The day had been divided into 24 equal hours long before that.

In an ATL, how would it have affected the development of mechanical and electronic clocks and watches – also time-keeping as a whole – if social custom required that local sun time with hours of flexible length was to be respected in the following way? How would such a society differ from ours?

  • Each calendar day is separated by local sunrise and sunset into two phases, light and night.
  • Each phase is divided into 12 hours – just for familiarity – of equal length (in a phase)! Therefore, the length of hours varies from day to day; light hours are longer than night hours in summer, but shorter in winter.
  • Sunset starts the calendar day and light at 00:00, night starts at 12:00. Consequently, there’s no need for daylight saving time – the very idea would be absurd.

A stop watch (absolute chronometer) and a clock (relative chronograph) would have to use different units of time from each other.

“Local time” may acquire a looser meaning over time, i.e. not the observed sunset and sunrise at the location of the clock would be decisive, but the nominal times from some kind of capital would be applied to a wider area.

With sundials, the length of hours may even have varied throughout a day, but let’s assume the difference is so small that people tried to implement hours constant per day when they started building mechanical or electronic clocks.

PS: Note that in OTL, (atomic) clocks are now so accurate that we occasionally need leap seconds to synchronize relative with absolute time, because a second is not exactly 1/86400th of any day.

  • $\begingroup$ Wait, how could you have two phases be decided by sunset and sunrise, but also have them be 12 hours long. That only happens on equinoxes. Which is the defining characteristic? $\endgroup$
    – DonyorM
    Jan 25, 2016 at 8:11
  • $\begingroup$ @DonyorM These hours are not a constant length, like we are used to. They differ by season and by latitude (but not by longitude). That’s a major point of the question. $\endgroup$
    – Crissov
    Jan 25, 2016 at 8:34
  • $\begingroup$ Ah I see. I didn't get that much the question. $\endgroup$
    – DonyorM
    Jan 25, 2016 at 8:40
  • $\begingroup$ How could such customs exist before precision clocks were available? $\endgroup$
    – JDługosz
    Jan 25, 2016 at 9:17
  • 1
    $\begingroup$ @OldBunny2800 OLT = Our/Original Time Line and ALT = Alternat(iv)e Time Line $\endgroup$
    – Crissov
    Jan 26, 2016 at 6:27

3 Answers 3


Most obviously, time-specific communication would be difficult running along north-south lines. The time zone system was first created in the U.S. to allow train engineers to safely pass each other on tracks, since the timing for this had to be very precise. If the system of time described in your question were to take place, then train companies would face such issues north to south as well.

My guess is that today's digital watches would have become computerized earlier. A computer can handle this type of complicated time system, whereas a simple counter cannot. Analogue clocks would be more difficult to develop. However, because daylight tends to change at a reasonably fixed rate from dat to day, a clock could still remain mostly accurate by changing it's daylight and nightime speeds from time to time. That is, during the day of the winter, the hand might move twice as far per tick than it would during the day of the summer. These changes might also require a switch to make it change from increasing movement to decreasing movement. Additionally, the stopwatch may not have developed at all. If you can't define what a second is precisely, how can you count them? Such things as timed-trials may have become entirely mistrusted, since participating in them on different days would lead to different results.

I think this kind of variance would create a much less time-focused society. If you couldn't precisely state the time (the mechanical clocks devised above would not always be precisely accurate) you would not be as worried about arriving on time. So this society would not be quite so punctual.

Eventually, if technology progressed like it did on this earth, some system of standard time would be required. My guess is something like the Unix Epoch would be created. A specific type of second (maybe that of a solstice or an equinox) would be picked, and a stopwatch (possibly the first invention of this kind of device) would measure the time since a starting point. All people communicating across distance would need such a clock to tell each other the precise time. So something like "I'll pass you at 20 years, 10 months, 20 days, and 4 hours standard time". This value would be the same for all users of "standard time". Because of the clunkyness of such a system, it would probably be employed only for those requiring precise times. However, it's possible that in a similar way to how all countries now employ time zones, all countries may some day employ a single unit of time (like our second), to help create world unity and enable communication in the digital age. Or a world leader might have their time be christened as "standard" and all others base there calculations of of it (Similar to how Greenwich works with time zones).


I think you are confusing the relation of causality: even when clocks were available, most people did not live by them, but by sunrise and sunset. What changed that was not the availability of clocks/watches or the stablishment of timezones; the big game changer was the spread of artificial light (before that, it did not matter the hour if there was no light enough to walk through the streets/see the work tools/etc.). A world like the one you propose would mean not much of a change

To answer your points:

  • Clocks would become very complicated, to the point that before the electronic age they would not be portable. Think that the day length varies every day, and also with latitude. You would have a few "tower clocks" in town, with people dedicated to adjust the length of hours each day (and night).

  • Work contracts would have to adapt to changing length of hours during the year:

    • be based in production ("I do not worry how many hours do you spend, I will pay you X for unit produced")

    • have different hour lengths during the year (your work day will be 8 light hours between June & July, 8.5 light hours between August & Setember, etc.)

    • have different hourly wages during the year (winter hours would be cheaper than the longer summer hours).

    • In particularly stable (all year round) industries, maybe an average hour rate could be stablished.

  • And of course, timetables would be a lot more complicated. Most likely, instead of given a fixed time, they would give a time range ("the train from London will arrive between 10:30 and 11:20, and will depart 5 minutes after arrival"). Nothing Earth shattering there, neither.


I think it was important to keep accurate time for logistic reasons, and to build practical portable timepieces: the railroad pocket watch.

They would use them, using "linear uniform seconds" to keep track of schedules and when to throw switches. Maybe they would adopt different names to keep it distinct from civil time.

The passengers, meanwhile, have some reason for wanting noon to mean actual civil noon and sunrise/sunset to be symmetrically arranged (aside: perhaps religious observations? Or some sense that makes it obvious when the sun passes the zenith).

So the public clock tower at each station would be set based on the Equation of time, which is the inverse of how linear time is determined at the observatory that notes the astronomical noon. eqn of time

Train schedules would be printed using local civil time. That works just fine for doing transfers and coordinating appointments at the destination.

Now, the civil clock would be different. It would not just be set to noon, but would need to run at a different speed each day. This would be done by adjusting the frequency base: probably the pendulum length for such a clock.

The railway workers would have a schedule that showed both civil time and railway time. They might write it differently as well as use different names, so a value is clear even without labelling. They would see, at each departure, that the public clock matches the expected time on the printed schedule when the railway time is the corresponding value, and fiddle with the public clock to keep its speed right.

Since transit times are naturally linear and unchanging, a repeating daily schedule would easily be done using railway time. But the civil time would drift a bit each day. To allow a schedule to be good for a week or other longer period, slack would be built in with the schedule made to work on the shortest day (first or last depending on the season) of the group, with extra padding appearing in the longer days. People would see that the train leaves a minute or two late and arrives a minute or two early based on the printed schedule, on the longer days in the week over which that schedule is used.

Actually if the the length of a day changes by 6 minutes from one day to the next, and the clock is arranged around noon, then each one hour of transit would have the train arrive early by 30 seconds, since the civil hour is longer but the train is the same physics speed.

People might get a feel for how tight the schedule is based on the day of the week: the postings just changed, so we've got plenty of time to make the train. Or, the postings change tomorrow so the train leaves right on the dot.

By week I mean whatever span of days the same timetable is used over. If that doesn't follow a nice civil unit like a week, it will need its own name and will become a thing known to seasoned travellers.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .