Timekeeping
Aug. 10th, 2011 10:07 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
steorran_woruldeI was thinking about timekeeping in my blind society, and that got me thinking about the basic ideas of timekeeping in general. Here are some thoughts.
Natural cycles
The most basic foundations of timekeeping come from observation of repeating natural cycles. The most significant of these are the day and year - on most parts of the earth, it would be hard not to notice the repeated daily and yearly cycles. Less obvious, but still easy to notice, is the cycle of lunar phases. There are other cycles, but those three are the ones that most commonly play a role in timekeeping around the world. Part of observing a cycle is observing different phases and points within the cycle. For the daily cycle, this includes such things as day vs. night, sunrise and sunset, noon, morning and evening. For the yearly cycle, this includes, on the most basic level of observation, seasons as reflected in weather and temperature and plant life; on a more complex level of observation, it includes such things as solstices and equinoxes. For the lunar cycle, this most basically includes lunar phases. Without observing the differences within the cycle, you don't have observations of a cycle.
Sequences of cycles
Cycles inherently repeat. Day follows day, (lunar) month follows month, year follows year. There are a few things you can do with this. You can count them linearly from some designated starting point - for example, counting years from an era start date, whether that's A.D. or a king's reign, or anything else. You can group a certain amount of them into a larger repeating cycle, as days are grouped into a week; the number of smaller cycles in the larger cycle may be arbitrary. Once you have smaller cycles grouped into a larger cycle, you do a few things. You can define a start point for the larger cycle and count smaller cycles from that start point - for example, counting the first through seventh days of the week, and then back to the first day. You can also name the smaller cycles within the larger cycle, like the names of the days of the week; a sequence of names need not have a starting point (compare how the first day of the week is variably considered to be Sunday or Monday; both are essentially arbitrary, and it would be quite possible to have a week without considering any day to be first).
Subdivisions of cycles
As mentioned above, natural cycles naturally have inherent subparts. Artificial cycles made of groups of natural cycles also have inherent subparts - the natural cycles that form their units. But cycles (natural or artificial) can also be dividied into non-inherent units. The subdivision of days into hours (and minutes and seconds) is such a division. These artificial units within a cycle can be numbered or named in just the same way as natural cycles that form units within longer cycles.
Reconciling cycles
Naturally defined cycles usually have no inherent relationship between their lengths, and this is true for the main cycles that form the basis of Earth timekeeping. Days and (lunar) months and years run independently, and a system of strictly observational timekeeping would have to reckon them largely independently (though you could remain essentially observational but still count or name smaller from a starting point defined in a larger unit - for example, the fifth day after the full moon, or the second month after the summer solstice). But it's natural to try to apply to these cycles the principles of grouping smaller cycles into larger ones / subdividing larger cycles into smaller units (you can look at it from either angle, since in this case both the smaller and larger units are given). If you do this, you run into a problem: the lengths don't match. There aren't an even number of months or days in a year. There aren't an even number of days in a month. There are all sorts of compromises to work with this issue. Consider months in a year: you can let either your year or your month get out of sync with its natural basis. The Julian (and subsequently Gregorian) calendars attempt to keep the year in sync, but define months in a way that no longer corresponds to the lunar month. The Islamic calendar keeps months in sync with lunar months, but lets the year get out of sync. The Chinese calendar keeps months in sync, and keeps years roughly in sync by adding intercalary months when needed. The daily cycle is fundamental enough to human life (at least in non-polar regions) that calendars don't let days get out of sync, so reconciling day lengths with year or month lengths normally involves either desynchronization of the year/month, or intercalation.
Coexisting cycles of different lengths
Some calendars have cycles of different sizes built of the same units that run alongside each other. In the Julian/Gregorian calendar, months and weeks are sort of like this, although it's a bit of a weird example because months don't have constant length. A better example is probably the Mayan calendar (and other Mesoamerican calendars), which has a 20-day cycle that ran alongside a 13-day cycle; these two cycles resynchronize to the same point every 260 days (20*13). This 260 day period is thus a repeating cycle of its own. Similarly, the Akan calendar has a 6-day week and a 7-day week that run alongside each other, which create a longer cycle of 42 days (6*7).
I think those are the core ideas of timekeeping, at least on a calendar scale; I haven't talked or thought much about timekeeping on a scale smaller than days.
Natural cycles
The most basic foundations of timekeeping come from observation of repeating natural cycles. The most significant of these are the day and year - on most parts of the earth, it would be hard not to notice the repeated daily and yearly cycles. Less obvious, but still easy to notice, is the cycle of lunar phases. There are other cycles, but those three are the ones that most commonly play a role in timekeeping around the world. Part of observing a cycle is observing different phases and points within the cycle. For the daily cycle, this includes such things as day vs. night, sunrise and sunset, noon, morning and evening. For the yearly cycle, this includes, on the most basic level of observation, seasons as reflected in weather and temperature and plant life; on a more complex level of observation, it includes such things as solstices and equinoxes. For the lunar cycle, this most basically includes lunar phases. Without observing the differences within the cycle, you don't have observations of a cycle.
Sequences of cycles
Cycles inherently repeat. Day follows day, (lunar) month follows month, year follows year. There are a few things you can do with this. You can count them linearly from some designated starting point - for example, counting years from an era start date, whether that's A.D. or a king's reign, or anything else. You can group a certain amount of them into a larger repeating cycle, as days are grouped into a week; the number of smaller cycles in the larger cycle may be arbitrary. Once you have smaller cycles grouped into a larger cycle, you do a few things. You can define a start point for the larger cycle and count smaller cycles from that start point - for example, counting the first through seventh days of the week, and then back to the first day. You can also name the smaller cycles within the larger cycle, like the names of the days of the week; a sequence of names need not have a starting point (compare how the first day of the week is variably considered to be Sunday or Monday; both are essentially arbitrary, and it would be quite possible to have a week without considering any day to be first).
Subdivisions of cycles
As mentioned above, natural cycles naturally have inherent subparts. Artificial cycles made of groups of natural cycles also have inherent subparts - the natural cycles that form their units. But cycles (natural or artificial) can also be dividied into non-inherent units. The subdivision of days into hours (and minutes and seconds) is such a division. These artificial units within a cycle can be numbered or named in just the same way as natural cycles that form units within longer cycles.
Reconciling cycles
Naturally defined cycles usually have no inherent relationship between their lengths, and this is true for the main cycles that form the basis of Earth timekeeping. Days and (lunar) months and years run independently, and a system of strictly observational timekeeping would have to reckon them largely independently (though you could remain essentially observational but still count or name smaller from a starting point defined in a larger unit - for example, the fifth day after the full moon, or the second month after the summer solstice). But it's natural to try to apply to these cycles the principles of grouping smaller cycles into larger ones / subdividing larger cycles into smaller units (you can look at it from either angle, since in this case both the smaller and larger units are given). If you do this, you run into a problem: the lengths don't match. There aren't an even number of months or days in a year. There aren't an even number of days in a month. There are all sorts of compromises to work with this issue. Consider months in a year: you can let either your year or your month get out of sync with its natural basis. The Julian (and subsequently Gregorian) calendars attempt to keep the year in sync, but define months in a way that no longer corresponds to the lunar month. The Islamic calendar keeps months in sync with lunar months, but lets the year get out of sync. The Chinese calendar keeps months in sync, and keeps years roughly in sync by adding intercalary months when needed. The daily cycle is fundamental enough to human life (at least in non-polar regions) that calendars don't let days get out of sync, so reconciling day lengths with year or month lengths normally involves either desynchronization of the year/month, or intercalation.
Coexisting cycles of different lengths
Some calendars have cycles of different sizes built of the same units that run alongside each other. In the Julian/Gregorian calendar, months and weeks are sort of like this, although it's a bit of a weird example because months don't have constant length. A better example is probably the Mayan calendar (and other Mesoamerican calendars), which has a 20-day cycle that ran alongside a 13-day cycle; these two cycles resynchronize to the same point every 260 days (20*13). This 260 day period is thus a repeating cycle of its own. Similarly, the Akan calendar has a 6-day week and a 7-day week that run alongside each other, which create a longer cycle of 42 days (6*7).
I think those are the core ideas of timekeeping, at least on a calendar scale; I haven't talked or thought much about timekeeping on a scale smaller than days.
no subject
Date: 2011-08-11 03:23 pm (UTC)Similarly with the Hebrew calendar, though that's not as closely lunar, since each month has a fixed number of days (not the same for all months).
I haven't talked or thought much about timekeeping on a scale smaller than days.
Two obvious sub-day scales are dividing the full day (typically reckoned as one of midnight-to-midnight, dawn-to-dawn, dusk-to-dusk, or noon-to-noon) into equal periods (modern European style); or to divide the light and dark periods separately into sub-periods (as the Romans and, I think, the Ancient Greeks did); the latter case means that "hours" will, in general, vary in length with the seasons.
Also, the number of subdivisions does not need to be the same for day and night; I think the Romans had [looks up Wikipedia] twelve hours during the day and "three or four" night watches.
Wikipedia also mentions that the "divide full day" variant can either divide the apparent solar day or the mean solar day; only in the latter are hours always the same length throughout the year.
no subject
Date: 2011-08-11 04:31 pm (UTC)My first draft of this post actually used the Hebrew calendar as its example, but I changed it because the Hebrew calendar has more complications, including adjustments to keep Rosh Hashana from falling on certain days of the week.
Really getting into reckoning of sub-day timescales requires thinking about why that size of unit, and how you measure them.
no subject
Date: 2011-08-11 06:08 pm (UTC)I imagine the "why is it that size" is often "because it's convenient" (though Wikipedia suspects that the number 12 is "magic" and has to do with the number of months in a year).
How you measure them is more various - I suppose the measurement method could influence the size (e.g. how long does a candle of "typical size" burn? kind of like how carob seeds became a weight unit), but that more often, the size comes first and then people figure out how to calibrate a candle/pendulum/clepsydra/clockwork/whatever to measure that unit.