The Death and Life of the 13-Month Calendar
citylab.com
citylab.com
Spoken like someone who's never had to write date handling logic. Having "personality" in your calendar system is a bug, not a feature. I feel confident that we could maintain our overall levels of personality/joie de vivre/je ne sais quoi by spending less time dealing with an incredibly arbitrary calendar system and more time doing things we actually enjoy.
Besides, even as a member of that small group, I like the variety of the current calendar, inefficient as it is.
I'm pointing at you, tempestn.
It imperial units for days; or like trying to do math in a numerical system where each significant digit has a different random radix. Does it have variety? Yes... Is even remotely fun to do anything practical with? No.
But yes, we probably should quash the "variety is fun" argument early in the discussion. The problem is that it's a persistent low cost in the back of everyone's heads. Sure, when you think about it you might like it a lot, but it will then stress you out in hard-to-perceive ways during the rest of the year, with no real benefit to show for it.
The huge error that Eastman made was to keep the same month-names for the other months. This is a bad idea; if two systems are going to exist in parallel they need to be disambiguated in practice.
The problem is, you'll find, that naming things is an incredibly political act. Agreeing on a new name for anything is a terribly complicated process, and when you're talking about something every single person has a stake in... well, it would have never happened. He tried very skillfully to bypass all that, and basically succeeded.
No, his mistake was to underestimate the clout that religious tradition still had on large sectors of the ruling elites. Even now, when we pride ourself in the "scientificness" and rationality of our societies, we're still enslaved by stories written by agrarian priests at the dawn of civilization -- in many ways, it's worse now because nobody would even entertain proposals as bold as modernizing a system we've been using, almost unchanged, for two millennia, no matter how broken it might be.
Fortunately we don't have a very pressing need to answer these sorts of questions; we have tools galore to handle them for us: smartphones, calendars, messaging apps, meetup etc. Topics like these often remind me of one of Einstein's more famous quotes:
[I do not] carry such information in my mind since it is readily available in books. ...The value of a college education is not the learning of many facts but the training of the mind to think.
We used to spend an enormous amount of time teaching our children mathematical tricks which have now all been made completely obsolete by the scientific calculator. Only the most rudimentary of arithmetical strategies have survived, if only for the fact that it's still a little inconvenient to reach into your pocket when scanning and counting objects with your eyes.
Casual calendar coordination is really easy to mess up in the current state.
This is precisely why the metric system, or uniform length of months, are useful.
Imperial made quite a lot of sense when numbers were small and calculations done without the aid of instruments: Using bases divisible by both 3 and 4 makes mental calculation quite quick.
Easier than base 10?
For bigger projects, yards are fairly easy to decompose too, into whole feet and inches.
It's a good system for projects of human or slightly super-human scale using low resolution instruments and mental math. It's rotten for high precision computerized tasks of widely varying scales.
Science would of course need to have a standardised second almost always different from the current second, but that's no different from having a standardised temperature and pressure almost always different from the current temperature and pressure.
Dates are almost impossible to reason about, and yet people are so used to our current broken system they aren't even aware of how badly they manage it. It is routine for date-time packages to produce bizarre results like "a month before a month from now" not being "today" in all cases because no one has any idea what anyone else means by "month", although everyone who has never thought about these problems is certain that everyone else means exactly what they mean in all ambiguous cases.
Anyone who has worked on date-time problems (not even libraries, just trying to correct for the special cases, cultural differences or just plain weirdness that the library of your choice doesn't deal with) will recognize the horrible truth of the joke, "I used to be in favour of space-travel, then I realized what effect it will have on date-time libraries."
Annoys me to no end that most calendar systems still don't understand that.
If I said "the servers are down for maintenance at 2am tomorrow" when would you think they will be unavailable?
If you said "2am tomorrow" I would assume two hours after next midnight. Perhaps this is a result of my usually late schedule because I normally go to bed around 2am, so everything up to then and including midnight is "tonight".
Person to person speech is almost always in logical days; "tomorrow" would not be a minute past midnight in this system. This at least makes some things make sense.
¹There _still_ tons of issues with this, such as people who work at night and sleep during the day, people can be on different logical days, some places don't have a dawn for several months…
...I think you just killed a piece of my soul... I understood that perfectly. :(
Backing your claim, I'd observe this is already a manifested fact. Consider holidays like Easter, which already are set by rules that have little to do with the calendar itself, and there's a number of such holidays floating about without any special problems of their own (above and beyond the problems of the calendar in general). The "personality" of the year is already independent of how we mark the time.
Birthdays being on the same day of the week is a poor objection: your birthday already falls on a weekday much more often then not, and chances are you celebrate it on the weekend before or after. We'd lose variability in holidays that are "last Thursday of the month": they'd just be on a specific date every year.
One nice thing we'd gain is that the phase of the moon would shift much slower with respect to our calendar (the moon doesn't do a Year Day). I suppose some superstitious healthcare workers might object to this. If you think about it, there are many more things naturally tied to 28 days than 28/30/31.
The arabic calendar which is based on Moon movements used to have an extra month every four(?) years to adjust the season changes. It was called Nasi' month. But after a Quran Sura published that prohibited this month Arabic calendar lost it's accuracy for seasons.[1]
The Iranian Calendar that is based on sun movements was safe from this Sura because it didn't have an extra month but an extra day each five years.[2]
It really helps to understand that in a calendar we're trying to measure three distinct (and variable) cycles, none of which fits precisely into the others, and each of which imposes its own rhythms on human life. Attempts to break each from the current 7/12/365 basis have virtually all failed.
The first is the Earth's rotation about its own axis -- different when measured with respect to the stars or the Sun.
The second is that of the Moon about Earth.
The third is of the Earth about the Sun.
We overlay them on each other and pick rough correspondences.
Much of the division has to do with Babylonian time reckoning, based on 360, and its factors: 2 * 2 * 2 * 3 * 3 * 5. From that you find 2, 4, 5, 6, 8, 9, 10, 12, 15, 18, 20, 24, 30, 45, 72, 90, 180, and 360. A seven-day week doesn't fit this directly but is close (between 6 and 8), and a 30-day month also fits well.
Then you come to realize that all timekeeping, _especially_ that which picks a specific starting point, is arbitrary.
The book also details several attempts to change the system, particularly following the French and Russian revolutions (oh, and the reason for that being the October revolution -- and there's the question of what was the last nation to adopt the Gregorian calendar, and why the output of 'cal 1752' is what it is ... and why even _that_ is arbitrary (it depends on where you're specifying the date).
What was the significance of cycles of the moon in the past, anyway? The main thing I can think of is astrology. Can tides be predicted accurately using phases of the moon?
Aside from the female menstrual cycle, the gestational cycle is also in line with the moon: A baby is brought to term in 10 moons or 280 days (9 months or so).
You would also have an easy way for disparate peoples to mark time, since in the past there was no such thing as telegraphs or even pony express.
Birth control pills have a big effect on this now.
> A baby is brought to term in 10 moons or 280 days (9 months or so).
If you can't tell how far along a pregnancy is without reference to the Moon, there's no help for you.
A human pregnancy is around 266 days +- 16. The reason for 280 days or 9 months is because that measures the time span since the last menstruation, not since conception.
(I was personally a bit surprised by this when expecting our first child. It's one of those things you apparently know about but still use "9 months" as a bit of a fixed expression.)
traditional agriculture uses the moon to determine what to plant when.
i'm not saying this actually makes a difference (altho it might, considering the moon does influence sea levels and hence how water flows through the system), just that it's what agriculture is traditionally based on.
Really not my metier, but a few thoughts.
The Moon is highly significant if you're living along a coastline (and most humans do) as the tides are directly related.
In a pretechnological society, the biggest and cheapest source of night-time lighting was moonlight (though you could locally out-shine it with torches, candles, or lanterns). For travel or any sort of nocturnal activity (including ocean navigation), you'd likely take it into consideration.
In arctic cultures, moonlight might be your only light for much of the year, and on snow, full moonlight is amazingly brilliant.
In terms of dividing up the the year, different moons, particularly named moons ("hunters", "harvest", etc.) are significant.
The Pax calendar has another very cool property, which is that the weekdays are always in sync with Gregorian calendar. This is done by carrying over leapdays until you have one full week accumulated, which is then inserted as a leapweek, in the same way we insert leapdays in the Gregorian calendar.
Each day in the Republican Calendar was divided into ten hours, each hour into 100 decimal minutes, and each decimal minute into 100 decimal seconds. Thus an hour was 144 conventional minutes (more than twice as long as a conventional hour), a minute was 86.4 conventional seconds (44% longer than a conventional minute), and a second was 0.864 conventional seconds (13.6% shorter than a conventional second).
I'm personally in favor of a more structured system like this, my wife on the other hand is viscerally opposed. If that schism is widespread I expect the next calender change to occur when the known world is ruled by an emperor :-)
*I realize this is only true in the northern hemisphere, but 90% of the world's population lives in the Northern Hemisphere. The other 10% can start their year in the Fall.
Honestly, having another month to pay bills (and you know that nobody's going to pro-rate everything down to 13 months) isn't really appealing right now, nor would the additional overhead of two more payroll cycles.
Logically, it makes sense, though.
Having your birthday perpetually on a wednesday would also not be so much fun. Perhaps "year day" could shift every year forward by 1 week day - year day could be Monday and then Jan 1 of the next year could be a Tuesday etc.
Right now if you get paid every other week, you get paid 3 times in 2 months and twice in all the other months.
If you get paid twice a month on the 15th and last day of the month, payroll constantly has to take account of holidays and weekends because the day of the week is always fluctuating.
All of those things would have to change, but they would be simpler after the change. It's the same argument as switching to the metric system in the US. Unfortunately the fixed cost of having to switch outweighs the benefits of switching (certainly for the calendar, less so for metric).
I think all but 5 states have laws dictating pay frequency, and in most cases they require at least two times a month pay frequency.
The most popular pay schedules tend to be 24, 26, and 52 times per year.
EDIT: This was supposed to be a reply to redblacktree. State law examples: - http://www.dir.ca.gov/dlse/faq_paydays.htm - http://labor.ny.gov/legal/counsel/pdf/frequency-of-pay-frequ...
Where did you get this? I have a friend who's paid monthly at his salaried software engineering gig.
So yeah, this looks like a nice change, but it's not going to be completely static from year to year in every culture.
Besides, I don't think it will completely eliminate the difficulty of figuring out what date "next Thursday" is. We at HN tend to be mathematically competent, so we can easily calculate multiples of 7 in our heads and subtract a few without breaking a sweat. Most ordinary people, on the other hand, have difficulty figuring out what date next Monday is even when it's Friday and you just need to add 3. They'll just look it up in a paper calendar or their mobile device.
In other fun facts, March is so named because that's when you go to war.
> Easter would fall on April 15 every year on Cotsworth's calendar
I don't understand why this would be. Can someone explain?
This calendar aligns months with weeks. Months are always exactly 4 weeks.
Just checked... you seem to be correct. The lunar calendar isn't aligned with this well enough. While the equinox date is fixed, the first full moon after that should bounce around by several days at least. Even allowing that it's always the Sunday after that, it still seems as if it could fluctuate by a week in either direction.
I think this is on the basis that it's important for Easter to be on a Sunday, and not nearly as important that it be the Sunday following the first full moon following the equinox, or whatever: just that once you say "it's got to be on a Sunday," you need some damn way to figure out which Sunday each year.
If you count by day-of-year when doing the conversion, though, it's the people born new year's eve on a leap year that are impacted.
7 and 28 are just as arbitrary as 12.
Why do we even have a solar year?
If the year was 100 days, we could have a whole year of summer.
The month is based on the lunar orbits, which are a bit more arbitrary now and less important with an urban rather than agricultural economy, but still not completely arbitrary.
Weeks are a bit arbitrary, as far as we can tell, but very widespread and we seem to be stuck with them.
A lunar month is 28 days.
There are few divisors of 28. 2, 4, 7, and 14 are the only ones. 28 days is too long for convenience if it's your shortest grouping of days. 2 days is too short. So, 4x7 or 7x4. Having 7 4 day weeks every month is probably a bit less convenient than 4 7 day weeks. Hence, 7 day weeks. (Or two 14 day weeks, which I think you essentially used to see some of the time: it seems like 17th-19th Century Britain was more aligned on fortnights than weeks).
Not NECESSARY in the same way that days are, but it's not exactly mysterious how they came about.
Edit: Oh, also, in terms of preferring 7 or 14 to 2 or 4, think about the lunar cycle. It easily breaks down into 7 or 14 day periods (7 days = new to half, or half to full, or full to half, or half to new. 14 days = new to full or full to new. 4 days = ?)
Edit2: Of course, a lunation (full moon to full moon) is 29.5 days, as opposed to the sidereal month (same part of sky to same part of sky) of 27.3 days. But I suspect that in antiquity, most people dealt with a conventional 28 days for both.
The seven day standard seems to have worked as a reasonable compromise across a wide variety of ancient economies and plan
The solar year is convenient so that one can speak of "the average historical weather for this month" and such... but it's true that many others have done lunar calendars etc.
We had a single mainframe subprogram that managed the whole thing. It was written in assembler and the few times it broke, the entire company stopped. Financials, ordering, invoicing, everything. Eventually they rewrote it in a more modern language, COBOL. This was in 2010.
There's also "Sun day"(Sunday), "Moon day"(Monday), "Fire day"(Tuesday), "Water day"(Wednesday), "Tree day"(Thursday), "Gold day"(Friday), "Earth day"(Saturday). So your claim is wrong in Japanese for every day except Sunday.
If you see something written like 四日("4 day"), this refers to the day of the month not the day of the week(for Japanese).
I remembered Japanese copied the Chinese at something there, but, whoops, it was the names of the planets that were copied. So the Chinese names of Mars(火星), Mercury(水星), Jupiter(木星), Venus(金星), and Saturn(土星) are copied onto the same days as planets that Romance languages name their days after.
My interpretation has always been that 日 is meant to represent "zero", but I never put any thought into why they didn't just call it 星期零.
I've noticed while living in China that, whereas Westerners like me virtually always remember what day of week it is but often forget the date in the month, Chinese are more likely to remember the day of the month and forget the day of the week.
Confusingly, the "seven luminaries" system in that article aren't just referencing the planets: they're also the names of Sun, Moon, Fire, Water, Wood, Metal, and Earth. I did get that sorted out by looking up Western alchemical ordering for the planets, which is in the same order. Wikipedia helpfully provides a table that includes the old names for the planets as well: