Julius Caesar's Year of Confusion | Hacker News Reader

Julius Caesar's Year of Confusion

bbc.com

123 pointsrntn109 comments

Nobody ever explained in school that the reason the numbers in the names of September, October, November, and December (7, 8, 9, 10) don't match up to their actual month number (9, 10, 11, 12) is because the Roman calendar began in March, not January. If you start from March, they line up perfectly. It also explains why the seemingly arbitrary choice of storing the extra leap day in February makes sense--just tack it onto the end!

Also makes sense to start the year when the weather start to get warmer, rather than the middle of winter

And the name March comes from Mars, the god of war, because it’s the month when military campaigns can restart after the winter break.

March became the new year again throughout much of the West in the Middle Ages as it moved to the feast of the Annunciation. March 25th was New Year's Day in England until the Gregorian calendar was adopted in the 1750s!

What's "weird" is i think we were taught in school that July and August were added by the Caesars (with their names), causing the 2 month shift you mention. Now i wonder where that idea came from, after this neat article.

I was also taught this.

I wonder if it has something to do with this, from the article:

> On Sosigenes' advice, Caesar added another two never-before-seen months to the year 46BC, one of 33 days and one 34, to bring the calendar in line with the Sun. The additions made the year the longest in history at 445 days long, with 15 months.

The two additional months were a one-time-only thing for that year, but I can see wires getting crossed and having that be a thing that was repeated for a while.

July and August were never added, though, simply renamed.

It's weird our calendar is named after actual people. The 1st French Republic briefly reversed this, but their solution didn't have staying power,

https://en.wikipedia.org/wiki/French_Republican_calendar#Mon...

I wish everyone on HN a happy "Windy"! :)

edit: Oh a cool fact in this tangent: the astronomer who discovered Planet 7 attempted to name it after a human—his patron, the British monarch George III. That would be an entertaining alternate history if it had gone that way!

https://en.wikipedia.org/wiki/Uranus#Name

Many languages do have meaningful names for months. E.g. in many Slavic countries: https://en.wikipedia.org/wiki/Slavic_calendar

The best part of course is we use similar month names, but as they are anchored to agricultural seasons, they correspond to different latin names in different countries.

My favorite is April in Polish, kwiecień, which means something like “month of flowers.”

We would insist it is named after George Washington instead, if that name’s stuck. The whole situation would be insufferable I think.

We don't do that for Georgetown, despite being in Washington DC and what George it's named after being somewhat unclear.

It’s a month of Fury, followed by a month of Birch and then it’s a month of flovers in where I come from

Also of interest - the Senate offered to rename November for Tiberius, but he declined, saying "What will you do if you have 13 Caesars?" (Cited from: https://www.cointalk.com/threads/welcome-to-tiberius%E2%80%9...)

At the other end of that scale is Commodus, who wanted every month named after one of his 12 names: Lucius, Aelius, Aurelius, Commodus, Augustus, Herculeus, Romanus, Exsuperatorius, Amazonius, Invictus, Felix, and Pius

Saparmurat Niyazov, President for Life of Turkmenistan, was comparatively modest when he renamed only January after himself and April after his mom.

https://en.wikipedia.org/wiki/2002_renaming_of_Turkmen_month...

Apparently they were renamed (from the no longer accurate Quintilius and Sextilius). Why they didn't get around to renaming the other numerical months, I don't know...

These two months were not added, they were just renamed after these emperors. In the minds of Romans, Julius Cesar and Augustus were gods.

The part that rarely gets mentioned - back then, months were only for the 'active' part of the year. The winter, when everybody stayed home, did not belong to any month at all.

[Deleted something that duplicated part of the article]

Google tells me that Romans did celebrate birthdays. I'm curious what they did when they were on the calendar that omitted the days between December and March. Surely some people were born on off calendar days.

Did they just count the days of the off calendar time? Move the celebration to some time in the non-sucky part of the year?

Yep, that's actually all mentioned in the article :)

How did we end up with the year starting in January? If you've got eight months and they're often named like that - e.g. "oct" for the 8th month - and you decide to add a couple for the time you're not counting during winter, wouldn't you add them at the end and say the year still starts in March? Or is that exactly what they did and we changed the end-of-year point layer (the article doesn't say)? If so, why did we change it?

https://www.britannica.com/story/why-does-the-new-year-start...

Tldr; There were many changes, with the last change introduced by the Gregorian calendar in 1582. The American colonies didn't adopt it until 1752

Thanks, so that does seem to confirm that when the Romans added Jan/Feb, they put them at the beginning of the year right from the start. Unfortunately it still doesn't really have an answer to why they didn't put them at the end, except that it fits with the chosen name (Janus = God of all beginnings. But of course you could name something before or after deciding where to put it).

For the Romans, the year was a government concept. They counted the years by the administration of a certain official. The change to start in January happened simply because they changed the date of inauguration for the council.

There were different standards in Europe during the middle ages and early modern period. Some places retained a date closer to the Roman New Year around the vernal equinox. These places used the Feast of the Annunciation (March 25). Other places (usually in the north) inherited a date closer to the winter solstice and so they used the Feast of the Circumcision (January 1).

As trade between countries grew during the early modern period countries began to standardize on the Feast of the Circumcision. England was a holdout, though and if you look at record from the first half of the 18th century, they will often include both years for dates between Jan 1 and March 25. The British Parliament formally adopted "Circumcision dates" to come into agreement with the continent in 1751.

https://en.wikipedia.org/wiki/Calendar_(New_Style)_Act_1750

On the other hand, I did remember from school that January was named after Janus, the two faced Roman god of beginnings, so explaining the year started in March might have resulted in more questions than answers!

The Roman calendar started in January at least as far back as the second century BC, and my understanding is that our sources are inconsistent as to who changed it, when it was changed, and why it was changed. Also, most of our sources are writing well after the change occurred (generally, they date to first century AD). (This is why the Julian and Gregorian calendars start on January 1, because the calendar they're based on did as well! Although lots of European countries did decide for much of the past 2000 years to start the year on other days like December 25 or March 25 or Easter).

This gives me a vibe that a lot of the tradition of the Roman calendar numbering is based on a "just-so" story that is made up to explain an interesting fact without being based on any actual evidence.

Originally, the year started on the first day of Spring: the Equinox. That was originally arranged to be the 1st of March, but slippage over the years means that the Equinox falls on or around the 21st of March in today's world. The New Year on 1st March also explains why the Leap Day falls after the 'last day of the year' on a day of its own: the 29th February.

I'd like to see a new calendar that reverts to the Equinox as the New Year's day and has the 28th February (and the 29th February if needed) as special 'Non-Week-Days' as part of a New Year Holiday Period. That would allow the New Year to start on the same day of the week as the previous year. In other words, we would have a perpetual calendar that never changes from one year to the next.

I love everything about this idea.

And then I think about every bit of code that I would need to write to support it and. From deep down inside, where my heart and soul are I can hear the screams "fuck no, fuck that".

And then I remember that I only do consulting now and get paid by the hour and think "chaching, carry on with this as you please".

The real question is if it's better or worse than the existing mess of date/time fun that we have to manage.

Good question!

My bet is this:

Existing nonsense: Doesn't change much.

Transition for all the badly set up systems... Y2K v2 baby!

Cognitive overhead for Users, therefor Product and therefor Developers will be huge. I think a lot of the rage will be fendign off stupid and bad ideas more than anything else.

Note: any one who deals with dates and deep history will have a chuckle and think nothing of it.

Here's a great blog post about fun coding around timezones

https://www.zainrizvi.io/blog/falsehoods-programmers-believe...

I wonder on how many days you could celebrate new year in the same year with various cultures. New Year's Eve, Chinese New Year, Roman new year. I think there's an old Russian new year too.

Oh yes, absolutely in favor of starting the year in the equinox (or at solstice). And while we are at it: change the days of a month to match the cycle of the moon - 28 days. So we would have 13 months. That's how they came to be: the earths rotation around itself is a day. The moons rotation around the earth is a month - and the earths rotation around the sun is a year.

But I believe, the natural rhytms are not perfect and constant, so I cannot claim, that this calendar would "never change from one year to the next". But I like the idea of directly refering to the natural cycles, instead of arbitary ones.

But like the other comment said, implementing any change of dates in reality won't be welcome, nor workfree ..

The original Roman calendar was based on lunar-solar calendars, and they don't line up with a solar year, because a moon cycle is 29.5 days, not 28. The Jewish calendar is one such calendar. It's why Jewish holidays vary so much. Some Christian holidays are based around moon cycles, like Easter. All lunar-solar calendars get around this problem by adding an extra month every few years.

Well sort of. Christian holidays were originally based on the same cycle as the Jewish Holidays but then the church fathers didn't like having to go ask the nearest Rabbi when their holidays were and tried to do it algorithmicly with absolutely no consoltation with anyone that knew anything about time keeping or astronomy and invented the ecclesiastical equinox (always March 21), and ecclesiastical full moon, neither of which bares relation to any astronomical event.

Interestingly the beginning of a lunar month is static (the new moon happens at the same moment everywhere on earth) while the beginning of a calendar month is sequential around the earth by timezone. The lunar cycle, earth orbital cycle (year) and earth rotational cycle (day) do not really sync up. There are approximately 12.37 lunar cycles per year (so once every 3ish years there is a 13th moon which roughly correlates to a "blue moon" although more specifically a blue moon is the second full moon in a given quarter. Each lunar cycle is not consistently the same length. That's just the tip of the iceberg. Source: I spent some time trying to mesh a lunisolar calendar model with our modern Gregorian model (https://www.cycalendars.com). I can't recommend lunisolar calendars as an efficient way of communicating consistent dates. Time becomes radial rather than sequential and integer timestamps cease to exist. But it is cool. :-)

You want probably this: https://en.m.wikipedia.org/wiki/International_Fixed_Calendar

Not exactly this, but pretty close .. thanks, did not heard of that one

> the cycle of the moon - 28 days

29.53059 days, so no, that's not super practical, and there's less than 13 of them in a year...

> I'd like to see a new calendar that reverts to the Equinox as the New Year's day and has the 28th February (and the 29th February if needed) as special 'Non-Week-Days' as part of a New Year Holiday Period. That would allow the New Year to start on the same day of the week as the previous year. In other words, we would have a perpetual calendar that never changes from one year to the next.

Alas, that would wreak havoc around the world, because I'm fairly sure that even if you can convince civil authorities, Christians, Muslims and Jews around the world won't be so easily convinced to shift around their God given days of worship (Sunday, Friday and Saturday respectively).

So their rest days would drift every year through the civil week.

> Originally, the year started on the first day of Spring: the Equinox.

Nowruz [0], the Persian New Year, still does today.

[0]: https://en.wikipedia.org/wiki/Nowruz

Perpetual calendars are a bad idea because of skipping special days. How long until the next Thursday? It depends on if it is February or a leap year. Think of how complicated the logic will be to calculate the day of the week.

The point of weeks is to have consistent cycle that can organize life. Also, why is it advantage to have the same day of week for same day of year? I suspect most people would prefer that their birthday wander around instead of always on Thursday. Calendars solved the problem of figuring out the day of week, and computers make it even easier.

There are also religions that require you to rest every 7th day. Having those two leap days would be anathema to them.

You would end up with the "religious week" and the "standard week" moving relative to one another over the years, in a classic case of trying to fix messy details and thereby adding more details (as in https://xkcd.com/927/).

Even without that issue, many Orthodox Christians still use the Julian rather than the modern Gregorian calendar for religious purposes (while still using the Gregorian for daily life). That's why Orthodox Christmas and Easter are on different days than the Catholic/Protestant ones.

Even the Roman Catholic Church that uses the Gregorian Calendar has moveable dates: Easter Sunday always falls on a Sunday and not on fixed day of a fixed month. On the other hand, Christmas day is always on December 25. Go figure.

> In other words, we would have a perpetual calendar that never changes from one year to the next.

Who really wants their birthday to be on a Tuesday for the rest of their life ?

This presupposes that the current socioeconomic system is the fixed norm. I don't think self-sustaining farming communities of the past cared too much which day it was if they had a cause of celebration. Likewise, I don't see why a post-scarcity future society wouldn't party on a Tuesday.

Our "fixed norm" might be temporary in the Grand Scheme of Things(TM), but for the typical human life span, it could still work out to be nothing but... birthdays on Tuesdays :-/

In such situations, I can see the culture adapting so that you have a "real birthday" and a "celebration birthday".

Famously, Queen Elizabeth II was born on 21 April 1926, but her official birthday was also celebrated on the second Saturday in June each year in the UK.

Why that particular day ? Maybe just because June didn't have any holidays yet ?

The British tax year still follows roughly this. A UK tax year runs from 6 April to the following 5 April.

See, e.g., https://www.theguardian.com/notesandqueries/query/0,5753,-23...

One issue is the religious aspects. For example Abrahamic religions base their whole calendar on the unending 7 day week. Having a non-weekday smudged in is very sacrilegious to may adherents.

It's surprising, in a way, but time is very much a political and religious thing.

I also think each year should start at spring equinox. However, I think that the day of week should be how it is rather than being the same day of week each time. Actually, I think Solar Hijri calendar will be good, but using Scientific Anno Mundi (not the usual Anno Mundi, and not Hijrah either).

Btw, March is named for the Roman god Mars. The god of war.

You can perhaps get a glimpse of Roman priorities from there.

If I remember right, part of the reason the Romans shifted to starting the year in January, was so that their consuls, who were elected for a calendar year each, would have two extra months to prepare. This become increasingly necessary as Rome expanded, and military campaigns went way beyond their local backyard.

I don't remember where I read this but I think the Romans got their calendar from Mesopotamia and it was based on a configuration of the sky from a few thousand years BCE. That configuration drifted so that's why we ended up here. I think the original January (or whatever it was called) was the start of spring.

Initially, there were no months after December, and the period of winter was not part of any month. When the calendar changed to have the year starting after December, they named January after the god of transitions.

Do you have sources for that?

The story I am familiar with had something like 12 months of 30 days each (or so), so that there were only about 5 (or 6) days without a month.

What you write sounds like they had almost a sixth of the year (after December to before March) not allocated to any months?

In contrast, January was named after Janus, the god of starting wars.

Another fun aspect here: The Italian growing season is in the winter, not the summer. In July-September, it's really too hot and dry to grow much in the Mediterranean, but it's much cooler and wetter in December-February, hence when a lot of the grain is grown [0]. And that leads into why March is the for Mars, because that is when the campaigning season starts, because that is when a lot of the harvest has been taken in, so now you can pillage your way across the countryside with all the food for your consular armies already packed up, so you don't need huge baggage trains to move quickly across the landscape.

This is likely true for Greece as well. So the old myth about Penelope and the pomegranate seeds is actually talking about July-September, not the winter months. For the ancient Greeks and Romans, the 'dead' season was the hot part of the year, not the cold part.

[0] I'm talking in general here. The specifics of a particular field, region, species, etc are very important to the ancient peoples and vary a lot over the whole region.

Specifics definitely vary. Eg inside Anatolia (which was well within the Greek known world), it can get very cold in Winter and stay cold until March. Mostly because much of the region has a rather high elevation.

Nobody in school ever pointed out that there were months containing Latin words for numbers or what those numbers were.

It's worth noting that the reason the calends had slipped so badly is because it was Julius Caesar's job as pontifex maximus to handle the periodic updates to the calends that kept it in sync with the solar year. But he was too busy fighting in Gaul and having his adventures to bother with it for years, so the calendar slipped into chaos due to his mismanagement. He then went on to fix his screw-up, but he didn't just decide to fix the calendar, he was in a position where he'd screwed up so long he had to fix it.

So a staunch supporter of Julius Caesar would say that he made it so bad that he had to fix it. A true political innovation

Strategically held off automating away his job until he secured a new position

He could have just added a few months to the current year and would have been done with it.

Caesar procrastinated so hard he changed time itself!

(This is funny, but I'd be interested to read a source on how true this is. Presumably there were priests that could take care of such things. He took care of lots of other Roman business while he was in Gaul.)

Can't beat Canadian Prime Minister and Founding Father John MacDonald blacking out while drunk during the Fenian Raids (Irish American civil war vets fighting for Irish freedom by attempting to invade British North America as a bargaining chip) in the 1860s [0]

[0] - https://www.thestar.com/news/canada/sir-john-a-passed-out-dr...

Update: seems like it was already ~2 months out of sync when Caesar went to Gaul, so this is basically incorrect?

And Julius Cesar in fact outsourced his duties! He hired some priests from Egypt to fix the calendar. And it didn't work right away, because the Egyptians told the Romans to add an extra day in the fourth year, but the Romans counted inclusively so they thought it was every three years.

Shameless plug, but for anyone interested in learning more, I do a podcast on the history of astronomy and did an episode on the Roman calendar:

https://songofurania.com/episode/024

One of the interesting things about their calendar is that because the role of weekdays was much weaker in Roman society, they could actually reuse calendars from year to year. Wealthy people would paint a calendar on the wall of their home and there were calendars literally etched into stone in public places.

Fun fact, expanding the Spain/England remark of misaligned calendars from the article:

Cervantes and Shakespeare died on the same date, but not the same day. They died 10 days apart. [1]

[1] https://www.dw.com/en/shakespeare-and-cervantes-two-geniuses...

Cervantes was buried on April 23 but he probably died on April 22. Shakespeare almost certainly knew Cervantes beacuse there is a lost play attributed to Shakespeare based on an episode in Miguel de Cervantes's Don Quixote. Probably Cervantes never read a Shakespeare's work.

https://www.folger.edu/blogs/shakespeare-and-beyond/shakespe...

https://en.wikipedia.org/wiki/The_History_of_Cardenio

Ha, a proof that by watching YouTube videos you learn something. This was covered by one of my favorite series on YouTube. Here is the particular video: https://www.youtube.com/watch?v=fD-R35DSSZY

Historia Civilis is pure gold for history enjoyers.

One of the best (history) creators on the platform in my opinion! His recent video on the history of work was also quite interesting.

I don't think his recent work video was good at all, namely that he falls into the same trap of idolizing days of yore while thinking about the current day realistically, a form of recency bias known as rosy retrospection. In short, serfs' (and other pre-Industrial peoples') "free time" was not free at all, it was spent doing other sorts of manual labor that wasn't their primary work but nevertheless needed to be done, similar to the chores we have today but much more strenuous. I'll add these comments on reddit that serve as much more of a realistic view of work back then [0] [1] [2].

[0] https://old.reddit.com/r/videos/comments/16vgh2l/the_history...

[1] https://old.reddit.com/r/AskHistorians/comments/2b8ovr/ive_s...

[2] https://old.reddit.com/r/AskHistorians/comments/28q7l5/how_m...

Any other history channels you enjoy? I finished every historia civilis video and need a new history channel to watch.

History of everything. But this is more of a fast talking podcast

I haven't found any channels quite like his one. That being said I do find toldinstone to be quite good regarding ancient history. Outside Youtube, https://acoup.blog/ is a great history blog.

I love that one of the first programming exercises you do as a student is to program a leap year calculator. Most students will forget or not know that every 400 years is a leap year, and I think the valuable lesson there is that even though code may seem trivially simple, it may have logical errors based on faulty assumptions that will bite you.

It's a good lesson - and the really proper take-away is to reject NIH tendencies and just use something like tzdatabase or language functions - your 400 year example isn't comprehensive - there are a series of exceptions beyond the 400 year one and, the best bit, they may be revised without warning. Use a library - let someone dedicated to the problem handle it.

>there are a series of exceptions beyond the 400 year one

There are none, calendar repeats every 400 years.

You are incorrect because the solution to leap years is approximate - we'll need to actually remove leap days on years that are multiples of 3200 and that won't even make things perfect.

Time is more complicated than you assume - and if you program this yourself be prepared to also get messed up by leap seconds which are used to account for earth's rotation varying over time.

It's hilarious to me that the ONE TIME the once-in-four-hundred-years exception mattered [well, I guess some parts of the world adopted it before 1600] is during the time period when students routinely implement leap year calculations on computers.

Strongly recommend Historia Civilis on this genre: https://youtu.be/fD-R35DSSZY

For the reference, Ukrainian Orthodox Church got the memo and switched to Revised Julian calendar as recently as last year. Calendars and timezones are still politcs.

Going off on a tangent: the 19th century Crimean War broke out partially because in one specific year Easter fell on the same date in both the Orthodox and Western Christian calendars, leading to kerfuffle at the Church of the Holy Sepulchre.

Or at least that was a good excuse to start it

Yes. That's why I coyly wrote 'broke out partially because', not 'was caused by'.

That's just for holidays not for regular dates tho, right?

BTW I love how Slavic countries' month names show the climate changes between the countries. Like we all have flowers month and linden trees month and scything month but the months are shifted :)

It’s about holidays, yes. So Christmas was on December 25, just their December 25 was everyone’s January 7th. Not sure what else they used the calendar internally for.

I find very appropriate that a Pope, who holds the modern title of "Pontifex Maximus", was the one that finally updated the calendar again after 1500 years.

Reminds me of one of the top voted answers ever by the #1 ranked account on StackOverflow:

https://stackoverflow.com/a/6841479

Calendars and timezones are hard to program for when you're dealing across centuries of weird changes.

Don’t ask cal command what February looks like in 1918 in some locales. Or do and be happy you don’t have to deal with that

Back in the day the Consuls were in charge of the calendar, and some of them paid more attention to that particular job than others.

Precise time really started with the railroads. Before that, well, you got there when you got there.

"Precise time really started with the railroads. Before that, well, you got there when you got there."

Interesting theory, but I believe binding contracts with time critical elements existed before the railroads. Merchants and loans and buildings and dept and interest rates ...

> Precise time really started with the railroads

Yes, counting hours and minutes and synchronizing them between distant locations, not dates.

Also that numbering of years like we understand and do is only started in 500s I think... Before that it was rather different, like naming years after who was in office. Or in other places who was emperor. So exactness was less important...

Democracy is all very well, but it takes a Caesar to fix time itself!

Sounds about right, judging from the fact that the United States still can't seem to adopt the metric system democratically.

Our calendar is epicycles within epicycles within epicycles.

Historia Civilis has a great video explaining how Caeser used the longest year to his advantage in war: https://youtu.be/fD-R35DSSZY

Feels as if only a few generations later they got bored of all that predictability and introduced the weird algorithm of determining the date of Easter to spice things up again.

Every time I read anything related to our calendars, and how we've changed them over time, I wish someone with a lot of influence led a similar effort nowadays to adopt the Kodak, or IFC, calendar[1]

[1] https://en.wikipedia.org/wiki/International_Fixed_Calendar

We should get started on the Y56k problem.

Reply on news.ycombinator.com