How do we know how long ago AD 1 was?

July 20, 2004 9:22 PM

How do we know how long ago AD 1 was? Lots more to this bizarre question inside.

I'm sure most people don't question that AD 1 was approximately 2003 years ago, but I can't find any evidence to make me believe in this 100% (sorry, comes with my faith, never believe blindly!)

Nowadays we have newspapers, TV guides, and scientific organisations like the NIST who keep track of time for us. The first newspaper of modern times is said to have been started in the early 1600's (in Europe), and I am sure there were no long-running dailies through the Dark Ages. I'm also going to assume there were no people who ticked days off on a piece of wood, which they then handed down to their sons, and so on and so forth. So how do can we accurately know (to the number of days) how much time passed between AD 1 and now?

Two different calendar systems were in common use in the last two thousand years, the Julian and the Gregorian. Leap years for the early AD years are supposedly unknown, and were mostly illogical.

One solution I thought of was that items from Roman times (perhaps coins with dates on) have been carbon dated, which will tell us how many years ago the item was made, and at least prove that AD 1 wasn't really 3,000 years ago. But this still doesn't narrow anything down to the day.. and considering some countries actually leapt over several days in the transition to the Gregorian calendar, is there actually any definitive answer for this?

posted by wackybrit to grab bag (25 comments total)

I'm only guessing, but I imagine that astronomical observations are a great tool for figuring this out. Eclipses are highly predictable, and it's easy to run the clock backwards and find out the dates that very obvious eclipses must have happened.

Then you just need to dig up some old records. Fortunately, people tended to take notes of things like that. "Dear Diary, today is <date>, and the sun disappeared for fifteen minutes. Everyone was terrified, but it came back eventually."
posted by xil at 9:33 PM on July 20, 2004

I fear there is no short answer to your question. Reconstructing history is like assembling a large jigsaw puzzle. As xil points out, many old texts refer explicitly to astronomical events such as eclipses which we can calculate. Otherwise we have successions of events to assemble. If text A refers to text B then B must be older.

I believe kings and dynasties play a huge role, These are usually well documented: A gave birth to B who died when he was 43 and was succeeded by C etc. If we can fix one Roman event based on an astronomical occurrence then we can rely on historians of the time and their histories of the emperors to allow us to fill in a larger gap. Still, this suffers imprecisions as you can see in this summary of the problems with Egyptian chronologies.

Also, you might find this mefi post by misteraitch of interest.
posted by vacapinta at 10:27 PM on July 20, 2004

So how do can we accurately know (to the number of days) how much time passed between AD 1 and now?

My uneducated guess is: they can't, the haven't, and at some point in time some pope or something said "I, being infallible and all that crap, decree today to be exactly 666 years after the birth of JC, etc, etc." and that was that.
posted by signal at 10:43 PM on July 20, 2004

I believe kings and dynasties play a huge role

Or, in other words, the relentless tick-tock of bureaucracy.
posted by five fresh fish at 10:50 PM on July 20, 2004

Read Questioning the Milennium by Stephen Jay Gould. He explains all this business of differing calendars. Also, it's Stephen Jay Gould and he's my hero.
posted by falconred at 10:53 PM on July 20, 2004

Way to pay attention to the message right below the text box, signal.
posted by jmd82 at 10:58 PM on July 20, 2004

Astronomy is the most accurate way and how it has always been done.
posted by raaka at 11:04 PM on July 20, 2004

Well, one reason we have an accurate form of dating is because of the way our current dating system about. As mentioned above, the reigns of emperors had quite a bit to do with it, and in the case of the Romans, the dating of emperors began before the birth of Jesus. Thus, AD 1 is in this year of so-and-so's reign. Keeping track of years of emperors and interregnums allowed such people as Bede (who wrote one of the first popular works, the Historia ecclesiastica gentis Anglorum, to use our "anno domini" system) to figure out the year. Also, it was important so that the date of Easter (which changes from year to year) could be found out. Bede was really interested in calendar computation and wrote extensively on dating Easter and calendar systems.

Years and dates make their way into many works, and the fact that modern scholars (with a larger data pool and the ability to correlate more sources) only suggest slightly different years for the birth of Jesus suggests that ancient scholars got it about right, considering calendar computations for history were done based on textual evidence (e.g., Luke 2:2 "This was the first registration and was taken while Quirinius was governor of Syria") and computations for the time of year were done on primitive astronomical calculations.

As a sidenote I realize I concentrate too much on western Europe and specifically the Anglo-Saxons...take that as a caveat. But it was asked concerning "our" dating system. Also, as in the Anglo-Saxon Chronicle, sometimes the year noted began at a different time, in autumn, for example.
posted by Gnatcho at 11:14 PM on July 20, 2004

Problems with the carbon dating approach:

1. You couldn't date the coin using carbon dating. Metal and stone-hewn coins aren't carbon-datable. Dating of a wood coin (or paper money) would give you the age of the tree from which it was made.

2. Carbon dating would likely be no more accurate (due to its inherent resolution) than the consensus this-year-is-2003-years-after 1AD assumption.

3. Say you had a coin that you believed was made at the "real" 1AD. Even if you could establish by some dating method that the coin was made, say, 2000 years ago, how would you ever know that your assumption about the "real 1AD" origin of the coin is true? Wouldn't that be begging the question?

The last is the most fundamental problem - even assuming we had extremely accurate dating technology, what would we use as our reference standard? Dating by astronomy doesn't solve this issue by itself. Say we knew this particular event happened in Rome roughly 2000 years ago, and on that day some astronomical event happened for which we could extrapolate the precise number of days back from today that it happened. How do we match up that date with the contemporary calendar?
posted by shoos at 12:02 AM on July 21, 2004

How do we match up that date with the contemporary calendar?

Do you mean how do we figure out what day a Roman would say it was if you hopped back through a time machine to then?

I'd guess we don't. What would the point be of knowing, now, that something happened 3 days before the ides of junius in AUC 497?

We'd probably just run our current (Gregorian) calendrical reckoning back to say that something happened on 20 July 1969 BC, or whatever, even though there was no month of July in 1969 BC. Maybe with whatever modifications it might need over really long timescales to keep it lined up with the equinoxes and solstices.

People from other cultures might well wind back the Islamic or Hebrew or whatever calendar to get a date that makes sense to them. If we drop the Gregorian system for something else, then we'd run that back through time instead.

Astronomers might well not bother and just refer to things by their Julian day number.
posted by ROU_Xenophobe at 1:43 AM on July 21, 2004

it's made easier by the seasons. they let you be out by a few days - a few months even - and still not lose track of the year.

consider how hard it would be without that - you wouldn't be sure someone had changed the length of a year. on the other hand, it means you have to factor in ad-hoc corrections as various calendars drifted away from seasons (the ancient equivalent of leap years), so i would guess accuracy to the number of days (rather than years) is much more difficult.

finally, of course, 1 ad is defined relative to today's date. so it was exactly that many days ago. it's just that what happened on the first day of 1 ad that may change...
posted by andrew cooke at 4:49 AM on July 21, 2004

it's made easier by the seasons. they let you be out by a few days - a few months even - and still not lose track of the year.

That's true, but even that alone doesn't necessarily mean that everyone would be keeping track of the year.

For example, if you were in the year 848, you might well not have known how long ago anything was, even if you could tell the season. I think the way we are bombarded with the date everywhere these days, in the newspaper, the TV guide, TiVo, and weblogs conditions us to believe everyone has this same grasp on time, but back before these daily reminders, I'd like to imagine that long periods of time would have been rather vague to define.
posted by wackybrit at 5:03 AM on July 21, 2004

finally, of course, 1 ad is defined relative to today's date. so it was exactly that many days ago. it's just that what happened on the first day of 1 ad that may change...

I think it should be defined as relative to the date when Dionysius Exiguus came up with the idea in the 6th century.

The convention is to use the Julian calendar when looking at historical dates during the period it was used. The date offset between it and Gregorian is the same everywhere, even if it was applied at different times (between 1582 and 1920,) so it's trivial for dates before that period.

Leap years being not exactly consistent in the Julian calendar does complicate things a little, but they were regular by the time our current method of counting years was invented. I thought it was known when they occurred. But since 1AD was defined retroactively in a rather arbitrary fashion, how you should account for those leap years depends on whether Exiguus did so. This page says there are "a considerable number of theories" about how he fixed the day.
posted by sfenders at 6:02 AM on July 21, 2004

At what point did people start using the AD system? That is, was it like some time in the 1600s when someone said "Screw this, today isn't the 24th day of the second year of the reign on King Jackington, it's Tuesday, the 2nd of June, 1621" or did it start much earlier, like in the 200s, etc. In either case, did it carry forward uninterrupted or did at some point someone have to reconstruct today's date in the AD system and issue a press release telling everyone it was 1972, not 1841?
posted by RustyBrooks at 6:06 AM on July 21, 2004

I'm sure I remember a MetaFilter link to a site of some European researchers who had a theory that a king/emperor had changed the year (and forced the church to go along with him) in order to make his family's dynasty seem greater.

They used evidence of a lack of new buildings & graves & other things during a certain several-hundred-year period in the dark ages as their proof.

Damned if I can find it now, though.
posted by bcwinters at 6:28 AM on July 21, 2004

Ok, I have to recommend a book here - "The Calendar" by David Ewing Duncan.

Essentially, the calendar has been messed about with considerably over the last 2 millenia, but not quite as much as before (for example, on moving to the Julian calendar, there was a year with 445 days in it).

The Julian calendar began in 45 BC, and the Romans (and hence, the rest of Europe) reckoned from this date (so, AD 1 was year 45 of the roman calendar). You have to go to the venerable Bede in about 570AD to the first usage of 'Anno domini' (in the year of our lord) itself. The 'AD' usage didn't become common usage until much later.

When Pope Gregory reformed the calendar, he removed 10 days from the calendar in October 1582 - so in October 1582, you had the 4th of October followed by the 15th of October the following day. (The change was introduced largely to reconcile Easter with it's proper place in the year).

There is no year AD 0, which complicates things a little (the number zero was invented by the Indians in the 8th century - too late for the calendar).

Other cultures use different calendars - the islamic calendar is purely based on the phases of the moon, the mayan calendar is based on the great cycle of about 6000 years, and I believe that the Ethiopians still use the Julian calendar, and it's also currently the early 1990s there. The French also attempted to decimilase the calendar after the French revolution - weeks were 10 days long, days were divided into 10 hours and 100 minutes to each hour. Shame it never really took off.
posted by BigCalm at 6:32 AM on July 21, 2004

Ah, sorry, it's not Bede. Perhaps an excerpt from the book might help here:

Dionysius complained that earlier Easter tables used a calendar widley followed at the time, which started at AD 284, the year the Emperor Diocletian ascended to the throne. Under this system the year...was designated as 247 anno diocletiani...But Diocleatian was a notorious persecutor of Christians, noted Dionysius, who tells Petronius that he 'preferred to count and denote the years from the incarnation of our Lord'....Dionysius calculated that Christ was born exactly 531 years earlier - which became his base year of AD 1. Where the aboot got this date for Christ's birth is unknown...Dionysius was the first ever to use the system we all now take for granted when he wrote on his Easter tables anni Domini nostri Jesu Christi 532-627.

(The calendar, David Ewing Duncan, p100-101)
posted by BigCalm at 7:01 AM on July 21, 2004

if you were in the year 848, you might well not have known how long ago anything was, even if you could tell the season

Only if you were a basic peasant. The Church and Kings and such had a good idea what year it was, either AD, AUC, or the N'th year of the current Byzantine emperor's reign.

At what point did people start using the AD system?

From reading stuff last night, it seems that it started taking off with Charlemagne and built from then on. IIRC, there were substantial fears and such about Jesus showing up in 999/1000, so it must have become popular by then?
posted by ROU_Xenophobe at 8:10 AM on July 21, 2004

A lot of wacky stuff that must have seemed like a good idea at the time has been done historically with calendars. My favourite warning on impossible dates:

Indeed. Sweden adopted the leap-year rule of the Gregorian calendar in 1700, making it a non-leap year, but without adjusting the calendar otherwise, so that after that Sweden was out of sync with both Julian and Gregorian calendars. After a while they discovered it was not such a great idea, and in 1712 Sweden moved back to Julian calendar by adding an extra day to February, resulting in the unusual date of 30 February 1712.
posted by Mitheral at 8:10 AM on July 21, 2004

Do you mean how do we figure out what day a Roman would say it was if you hopped back through a time machine to then?

No, I meant, for example, that if both we and Dionysius (who I take as having the last word on when AD 1 really was) used this same great astronomical dating method to date some particular event around, say 2000 years ago, how could we be certain that both our and Dionysius's counting back the days to that event would hit the same point on our respective calendars relative to AD 1? Would they both be the exact same distance from AD 1?

In addition to the astronomical method we'd need something else.
posted by shoos at 10:24 AM on July 21, 2004

And if you really want to date things from the birth of the jesus, you're gonna run into some problems as the whole December 25th thing is an arbitrary day (supposely chosen to be near winter Solstice) for celebration, as we don't know when, exactly, he was born.
posted by kaibutsu at 11:51 AM on July 21, 2004

shoos:
I'm still not sure what you're asking. Let's assume there was a solar eclipse that we would place on 1 September AD 10, or Julian day 1724954. What's your concern? That Dionysius might have said that it was 11 September AD 10?

The properties of the Julian and Gregorian calendars are well-known and converting between the two is just math.
posted by ROU_Xenophobe at 12:02 PM on July 21, 2004

It is also important to note that while the identification of astronomical events referred to in historical documents may not answer all of our questions about the date system in use at the time or calendrical continuity since that time, it does at least allow us to establish how long ago the astronomical event occurred. As in "such-and-such an eclipse occurred 900,321 days before today's date, at 1 hour and 34 minutes past local noontime for the observer whose account we are studying." That's pretty specific, and if we choose we can then represent that date and time in whatever calendar we like.

As for how our current calendar is fixed, it seems best by far to base it on the declaration that today (where I sit, that is) is July 22, 2004, rather than the declaration that Christ was born in 1AD. That way, in the rather unlikely event that we discover a missing or spurious year somewhere in the historical record we do not have to redefine the present date. Infuriating as it would be I imagine it would be far easier to adjust dates in the historical record than it would be to fix the present accounting of dates, which would make the Y2K problem look like child's play.
posted by Songdog at 8:41 AM on July 22, 2004

ROU_Xenophobe, in the original post wackbrit asked how we could be absolutely certain of when 1 AD really was, relative to our calendar, and suggested that we might use some hi-tech method to resolve that question (independently of the recorded history of the past 2 millenia, which he seemed to be not absolutely confident in)

My point was just that such a hi-tech method would be no more accurate than referring to our calendar.
posted by shoos at 4:42 PM on July 22, 2004

That should read wackybrit. He's definitely not wack.
posted by shoos at 8:36 PM on July 22, 2004