A central issue in the field of archaeology is that of dating. No, I don’t mean a social interaction with a fellow excavator sipping Arabic coffee over a headlamp-lit dinner of hummus and shawarma. I’m speaking of archaeological dating, which is the act of assigning a date to a historical object or event. There are two types of dating: relative dating, which places objects within a timeline relative to other objects discovered locally, and absolute dating, which seeks to establish a fixed date corresponding to the rest of world history.
Today, the entire world uses a calendar based on a sidereal solar year, that is, one that measures the length of a year based on the sun’s journey through our sky from when it returns to the same position with respect to the distant, fixed stars. This calendar has a breaking point around two millennia ago: dates of the current era are designated as Anno Domini (A.D., “In the year of the Lord”) or Common Era (C.E.), and those before the current era are Before Christ (B.C.) or Before Common Era (B.C.E.). Theoretically we should be able to date everything according to this system by identifying the year in which an event took place. One of the best ways of doing this is by finding a written source mentioning an event in the heavens—a lunar or solar eclipse, the heliacal rising (annually, when a celestial body first becomes visible above the eastern horizon before sunrise) of a “star” (Sothis in Egypt, Venus in Mesopotamia), a new moon, etc. Astronomers can usually calculate exactly when such events took place in the past and where they were visible. The student of the 050ancient world then chooses a suitable event from what the astronomers can offer.
A classic case is that of Alexander the Great (356–323 B.C.E.). His life is directly related to the history of Greece, Asia Minor, Palestine, Egypt, Mesopotamia, and India. He himself and his father minted coins—as did his successors. Given the Greek system of counting in terms of the games at Olympia, which allegedly had taken place every four years since 776 B.C.E., it is possible to date precisely most prominent people in the Greek world. Regardless of whether the games began at a later date, the Greeks counted according to their understanding of the sequence of the Olympic Games and left a reliable means of following their history.
There is even an ancient account of Alexander’s campaigns written by the Roman general Arrian, who lived about five centuries after Alexander. Arrian had access to the writings of the companions of Alexander and recorded that there had been a lunar eclipse just before the Battle of Gaugamela, which brought Alexander victory in Mesopotamia. By piecing all the evidence together, we can establish that this campaign took place in 331 B.C.E. Therefore, the date of that lunar eclipse should be September 20, 331 B.C.E. (which NASA tells us was visible in Mesopotamia).
One occasionally reads that the eclipse recorded by Arrian occurred on 1 October -331, but this is not accurate. This misidentification stems from the fact that modern astronomers use an exceptional calendar because they realize that in the B.C.E./C.E. system, there is no year “0,” and thus there is a year missing. Therefore, they insert a year “0” before 1 C.E., and our 2 B.C.E. becomes -1. Thus, the eclipse of 1 October -331 (332 B.C.E.) was actually a year earlier than Alexander’s campaign in Mesopotamia.
Indeed this case illustrates how easy it is—with proper attention!—to fit astronomical events of the distant past (which can be mathematically calculated) into a historical narrative. However, for earlier ages, one must juggle with several unknowns. Even though there is today no longer any doubt about the length of a year, we must recognize that the ancients had no regular system of recording time, and for most of human history our ancestors did not even know how long a year was. Furthermore, when they dated documents, they could make mistakes. Thus we must struggle to fit the ancient world into our modern, precise system, for they did not bequeath one to us.
In principle, it would be ideal if we could assign exact dates to everything, so that we would be able to say that a specific Celtic burial, for instance, dates to the same time as a military event affecting Egypt, and a transformation of class status in China. In fact, we are able only to suggest that the earliest inscriptions testifying to the emergence of the shuren class in Zhou China date to the sixth century B.C.E., before the Persian conquest of Egypt in 525 B.C.E., while the burial of a Celtic prince at Hochdorf in Germany may have preceded or followed that. We can’t get much closer than simply recognizing that it is not often that an event can be dated precisely, and radiocarbon or dendrochronological dates can rarely get more precise than typologically settled ones.
Even our precise dates are only approximations, since the actual historical events cannot usually be exactly lined up with the material that provides a date. Archaeologists must always make a compromise to provide a date that makes sense, and this frequently leads to dates that are either uncomfortably vague but roughly accurate or too precise and possibly misleading.
051Ultimately, we have three separate problems: (1) fixing accurate dates, (2) determining the probability of the dates’ accuracy, and (3) relating that to human history.
We are typically more confident establishing dates of objects in our own field of study with relative accuracy, as we know the material. However, when scholars assess objects outside their own field, they generally have to rely on the testimony of others. This often results in confusion, because chronological debates and issues of chronological probability are fairly nuanced and often familiar only to those working within their own field of study.
Those of us who are concerned about this conundrum are prepared to struggle with the “certainty of uncertainty,” while still aiming at precision. We basically employ every possible means 052available, combining (1) relative “typological” and “stratigraphic” chronologies with (2) textual sources recording astronomical observations, dates relating to kings and/or buildings, etc., and (3) more precise means such as radiocarbon dating and dendrochronology, along with (4) numismatic evidence linked to archaeological contexts and textual sources.
The very earliest dates in human history can be keyed to geological events, so that skeletons or artifacts (manufactured or reworked by humans) buried beneath or above a specific geological deposit can be dated as being older or younger than that deposit. For the earliest, the dates can be disconcertingly vague, such as “around four and a half million years ago.” In most cases, we have nothing but dates calculated by reckoning atomic transformations, such as energy loss (thermoluminescence) or radioactive decay of carbon atoms (radiocarbon). Thermoluminescence measures the ages of rocks, while radiocarbon measures material with organic content.
In recent years, measuring with the celebrated radioactive carbon (14C) isotopes has improved. First of all, the period of the half-life (the rate at which the atoms gradually degrade) was corrected from the Libby half-life of 5,568 ± 30 years to the currently accepted half-life of 5,730 ± 40 years. Second, smaller and smaller quantities of material can be measured with increasing accuracy. It allows one to roughly estimate the age of the death of a piece of wood (or whatever organic material one has). But the “± 40 years” subtracts from the precision—and this uncertainty is further increased by the gradual recognition that the quantities of radiocarbon in the atmosphere have changed over time. Thus, merely measuring the half-life and projecting a date based on that is insufficient, as these variations must be taken into account.
053For historical events and objects, such calculated dates (relying on calibrated radiocarbon dating) can now be so unnervingly precise that they invite skepticism. For example, one recently proposed radiocarbon date for the volcanic eruption at Thera in the Aegean is 1613 ± 13 B.C.E. This new radiocarbon date means that there is an extremely high probability (98 percent) that the eruption took place between 1600 and 1627 B.C.E. However, many archaeologists would prefer it be decades or even a century later!
For earlier times, the time scale is unsettlingly vague and measured in terms of B.P. (Before Present, which in practice is set at 1950 C.E.) rather than calendar years C.E. or B.C.E. The uncertainty for these dates is high because both radiocarbon and thermoluminescence estimate a measurable atomic decay, and, thus, error margins are inevitable.
However, for the past few thousand years, we can use dendrochronology—counting tree-rings (one by one, lining up trunk after trunk) of long-lived bristlecone pines in the White Mountains of California—in order to improve radiocarbon dating. Measuring the radiocarbon data from any specific ring gives an isotope value hinting at how much radiocarbon has been degraded since the tree-ring formed (and 054stopped absorbing oxygen from the atmosphere). Since the quantities of radiocarbon (14C) in the atmosphere have changed but the number of rings (even when lining up several trees) is certain, it means that one ring has been precisely dated. Using calculations, the real age of the wood can be used to “calibrate” the radiocarbon dates of objects found anywhere else on earth.
Thus the earliest written materials (clay tablets) in Mesopotamia can be dated by the wood associated with them, and, according to the most recent calibrations, this works out to about 3500 B.C.E. (± 40 years or so with a 95 percent probability). Yet the clay tablets were rubbish by the time they were thrown into the context where the wood was, so writing was probably invented sometime well before 3500 B.C.E.
Unfortunately, however, writing was invented in Mesopotamia for accounting purposes and used only to record historical data and poetry centuries later. Thus, the historical period begins around the mid-third millennium B.C.E. (and not with the invention of writing a millennium earlier). Since the Egyptians and the Mesopotamians had already recognized the concept of a year by the beginning of the third millennium B.C.E., we can reference dates, usually defined in terms of the reigns of kings. However, the years were not real solar years; the Egyptians used a year of 360 and/or 365 days, while the Mesopotamians used a year with lunar months where they introduced an intercalary month whenever necessary to bring the months and years back into line.
The problem is that lunar months are either 29 or 30 days long, so a lunar year of 12 lunar months gives a year of some 354 days. Simplifying to 30-day months solves nothing since the resulting year of 360 days corresponds to neither real lunar months nor a full solar year. A sidereal solar year marks the point when the sun returns to the same position with respect to the fixed stars, but it does not correspond to a number of days divided by an integer, 055and thus the methods of the Bronze Age Near East proved inadequate at measuring a year. The Egyptians experimented with 360-day years until they discovered that more than five full days were missing. Adding five days was not enough since the sidereal solar year is 365.25636 days, and thus the Egyptians were missing a bit more than a quarter of a day—and the Mesopotamians around ten full days (depending on whether lunar month years or 360-day years were used). The Mesopotamians adopted the principle of inserting an intercalary month—an occasional corrective period of time added to a year—when the calendar became out of sync with the seasons. The Egyptians simply observed the fixed stars and ceased to try to alter their calendar. Neither method was useful for creating a system with continuously reliable year-lengths.
Ultimately, the system was solved. The most important developments came when the Greek scientist Hipparchus (190–120 B.C.E.) realized the nature and importance of a sidereal year, and the Roman ruler Julius Caesar (100–44 B.C.E.) introduced a calendar compatible with it. But this did not stop experimentation. The Jewish calendar still uses intercalary months, and the Islamic calendar dispenses with them, keeping exclusively lunar months and thus annually “losing” weeks with respect to the solar year. The traditional lunisolar Chinese calendar has also spawned a wealth of other calendars trying to link the sun, moon, and stars. None of these calendars is really practical—but until recently, there was no adequate means of measuring the year.
The Mesopotamians and Egyptians composed lists of kings and dated years according to the reign of the king. Although some of these copies are unreliable or not able to be linked to other sequences, usually, with enough data from inscriptions, we can estimate the lengths of the mentioned reigns. There are also several other relatively important written sources for the history of this period, such as the so-called eponym lists from Mesopotamia that give the names of a long series of officials who are each associated with a single year (a precursor of the Roman system of consuls). This allows a sequence to be established and linked to other lists, kings, and events. Also relevant are the building inscriptions of various Assyrian kings, some of whom declared that they had repaired or rebuilt a temple—and informed us who had done so previously and when. The various lists can be used with caution, but they are not reliable sources from which to copy dates.
Using a combination of all these possibilities, we can estimate roughly the dates—within a range of about a century or so—for the kings of Mesopotamia and Egypt in the third millennium B.C.E.
Fortunately, we have occasional astronomical observations in written records from around the end of the third millennium B.C.E. onward, and thanks to NASA, we now possess relatively reliable data about the visibility of lunar and solar eclipses in the past few thousand years. These various sources allow us to propose some precise dates for certain events, but there is no agreement, let alone a consensus, on the matter. In principle, the king lists, the astronomical observations, the relative chronologies of the pottery, and the dendrochronological and occasional radiocarbon dates should offer some kind of certainty. In fact, at the present moment, the entire chronology for the archaeology of Western Asia, Egypt, and the Eastern Mediterranean for the first half of the second millennium B.C.E. is still a matter of debate. In Egypt, it is a matter of decades (of whether Dynasty 12 began early or late in the 20th century B.C.E.); for the Late Minoan I period in the Aegean and the Old Babylonian period in Mesopotamia, it is a difference of up to a century.
The dust really begins to settle only in the second half of the second millennium—and here it is our luck that we have the letters exchanged between the kings of the Near East—letters found in the capitals of the Hittites in Anatolia and the Egyptians at Amarna. First of all, we have the rough guides of the various dynasties, and we have rough dates concerning certain wars between Elam, Babylon, Assyria, Mittani, and Hatti. And then we have the letters exchanged between the kings; since the Assyrian king Aššur-uballiṭ actually corresponded with Akhenaten in Egypt, the two were contemporaries, and their reigns can be narrowed down to the second half of the 14th century B.C.E.—along with a series of other Egyptian, Mittani, Hittite, Babylonian, and Assyrian kings who lived at roughly the same time. In the same fashion, the letters exchanged between Ramesses II of Egypt 077and the Hittite kings mean that we can date Shalmaneser I of Assyria, as well as Murshili and Muwatalli of the Hittites, and Tukulti-Ninurta I of Assyria. They all lived late in the 13th century.
As we gradually increase the precision of the dates for the earlier part of the second millennium B.C.E., we will probably be able to extend the reach and meaning of what has been called the historical era. With improved understanding of events in the second half of the third millennium, we would be able to take our knowledge of history back to the beginning of the period for which we can practice historiography seriously (in the sense of using written documents that convey useful historical information).
Yet, even in more recent times, this promise of clarity is eclipsed by a dark age of several centuries after the end of the Near Eastern Bronze Age, c. 1200 B.C.E. When the light returns, Israel has fallen (c. 720 B.C.E.), and Jerusalem is threatened by the Assyrians (c. 700 B.C.E.). For Mesopotamian and Egyptian chronology, we have lists of kings and events that allow us to confirm that the Biblical narratives are in agreement with historical events known elsewhere from around 700 B.C.E. onward. A hundred and fifty years later, the Athenians are minting tetradrachmas, and by the fourth century B.C.E., several kingdoms are issuing coins that can be dated to rulers and events in their reigns.
Thus, in the case of a house in Babylonia, we found a cuneiform tablet dating to the reign of Darius I of Persia (c. 550–486 B.C.E.) in a foundation trench, a coin of Alexander the Great (356–323 B.C.E.) on a wall, and coins dating to the decades after his death (as far as 311 B.C.E.) above the rubble of the walls. The tablet was obviously discarded rubbish by the time it ended up in the foundations, and the building itself was built much later. The coins above the building were in good shape and probably new; the building probably ceased to exist around the time of Alexander. This meant that the building was erected well after the end of the sixth century and in ruins by the end of the fourth century.
The Roman system of dating was chaotic (with irregular years, counting by consuls, etc.). Yet despite their chaotic calendars, precision increases enormously with the Romans who issued clearly dated coins for centuries. The range of coins in a given hoard may cover centuries, but the final coin may be a reasonable approximation of the date of the event (as coins will be hidden when a threat is approaching).
Shortly thereafter (with the subversion of the Roman Empire by an oriental sect, a.k.a., Christianity), a calendar was introduced that effectively offered a compatible means of dating ancient and modern documents around the world. The Jewish Anno Mundi system is similar, but more modest, since it relates to the age of the earth rather than a particular dynasty. Before this concept of a universal calendar, years were dated individually according to a reigning king or emperor, and at best, the lists recorded the names of all of the kings in history. But this was always problematic. 078The introduction of the Christian Calendar changed things. Calculations allow our calendar to be used to date Chinese and Islamic kings and dynasties—as well as our own history—and to project our own dates back into early historic and prehistoric times as I have described here.
One minor remaining challenge is that the Gregorian and Julian calendars have both remained in use, so that the Christian Church may have the same year, but there is a gap of some two weeks between the calendar dates used in the calendars (due to different opinions about how to integrate a regular calendar based on days with a sidereal solar year of 365.24 days—and thus incompatible with the specific number of integers, as we require for our daily use). However, in practice, the Chinese, Hindus, Jews, and Muslims (etc.) have all adopted the calendar propagated by Pope Gregory XIII in October 1582. Only the Orthodox Christians—and the few of us dealing with ancient chronology—regularly turn to the Julian calendar.
Those of us studying ancient chronology and dating will continue to fumble with the hands of time, but collectively, one day we may be able to synchronize our ancient chronologies, achieving a precise idea of which events happened when and where. With the means described here, we have already come far closer to the goal of creating a chronological skeleton for history.
Dating is one of the most important aspects of the archaeological world. In the first article of a new series called Biblical Archaeology 101, discover the difference between relative and absolute dating, and learn about the many techniques archaeologists use to date sites, people, objects, and historical events.