Ancient history books are full of dates. One reads, for instance, that the Great Pyramid at Giza was built around 2500 B.C., not 3000 B.C. or 2000 B.C., and that Jerusalem was captured by the Babylonians in 586 B.C., not 566 B.C. But how do we know these dates are accurate? Couldn’t ancient history have been hijacked years ago by a conspiracy 028of chronology whizzes who set in motion an endless succession of wrong dates?
Compared to other inductive sciences, like biology or geology, ancient chronology is triply handicapped. The first problem is that it is non-replicable: We cannot take Egyptian history to a laboratory and make it run all over again to test our current understanding. Second, ancient historians cannot view their material directly; they must rely on reports from ancient sources of varying reliability. The third handicap is the paucity of information. Not only do students of antiquity have to evaluate the reliability of evidence—asking what it says for certain, what it may say, and what it does not say—but they also have to accept the fact that there is so little of it to evaluate. Most evidence from the ancient world has been lost, and so every scrap counts.
In 45 B.C. Julius Caesar instituted the Julian calendar, which was slightly modified in 1582 under Pope Gregory XIII. This Julian-Gregorian calendar provides a continuity of dates from 45 B.C. to the present, meaning that historians of Charlemagne or the Renaissance need not be overly concerned about chronology. In general, chronology demands much more attention and energy from “B.C.” historians than it does from “A.D.” historians.
B.C. chronology begins soon after B.C. itself. The early Christian scholar Jerome (340–420 A.D.) produced an expanded Latin edition of the Greek Chronicle by Eusebius of Caesarea (260–340 A.D.). This work then became the basic textbook of ancient history and chronology in the West for more than a thousand years. But its B.C. chronology is not very accurate.a One reason was the break with the past that accompanied the extinction of hieroglyphic and cuneiform forms of writing—and therefore the loss of any history, especially of ancient Egypt and Mesopotamia, recorded in these scripts—by the second century A.D., not long before the Chronicle was composed. Another reason was the desire to subordinate all chronology to biblical chronology, which is not very precise.
029
For more than a thousand years, B.C. chronology hardly progressed past the Chronicle. Then the decline and fall of Constantinople in the 15th century precipitated a flow of Greek manuscripts to the West, which contained much classical Greek and Hellenistic learning—until then unknown to Western scholars. One of the prominent scholars to work on this material was Joseph Scaliger (1540–1609 A.D.), a French Huguenot refugee who spent the latter part of his career at the University of Leiden in Holland. In De Emendatione Temporum (1583), Scaliger assembled and organized vast amounts of data on ancient chronology.
Among these ancient Greek manuscripts was a chronological tool that became the foundation of B.C. chronology for more than 300 years: a king list.b The list begins on February 26, 747 B.C.,c with the reign of the Babylonian ruler Nabonassar. In some late versions, the list is extended to the 15th-century A.D. Ottoman rulers of Constantinople. What is important for B.C. chronology is that the list includes 44 successive kings of the first millennium B.C.; the 44th king is the Roman ruler Octavian, later called Emperor Augustus (64 B.C.-14 A.D.). The early kings of the B.C. portion of the list are Assyrian, Babylonian and Persian rulers of Babylon. The later kings are rulers of Egypt,d which Octavian/Augustus annexed in 30 B.C.
Although it may not seem that impressive, this list is the most important chronological document that has come down to us from antiquity. (A slightly modernized version of the king list’s first 44 rulers appears opposite.) Each line of text contains three items: a royal name and two numbers. The number to the left is the length of reign in Egyptian civil years of 365 days (12 months of 30 days plus 5 so-called epagomenal days).e The number to the right is the cumulative count of years, starting with February 03026, 747 B.C. This cumulative count is also called the Era of Nabonassar, after the list’s first king.f
The earliest-known version of the king list appears in the works of antiquity’s greatest astronomer, Claudius Ptolemy (no known relation to the Greek-Egyptian kings of that name), who lived in the second century A.D. and worked at the Alexandria Library.g The list is therefore called Ptolemy’s Canon. The Canon was made for astronomers, not historians, but historians have also been its beneficiaries. The Canon’s year of 365 days is wonderfully simple, which comes in handy if you need to count the exact number of days between two events that are hundreds of years apart.
Because the ancient Egyptians did not rectify their calendar with leap years, their 365-day year wandered back through the actual solar calendar at a rate of about one day every four years. In 747 B.C. the Egyptian New Year’s Day fell on February 26, and in 525 B.C. it fell on January 2. By 30 B.C.,the first day of the Egyptian new year had moved back to August 31. In this way, the Egyptian new year moved back through all the seasons, completing a cycle in about 1,460 years (4 x 365).
When the Canon surfaced in Greek manuscripts in the early 17th century A.D., its list of kings quickly became the foundation of B.C. chronology—at least for the period it covers. Little was known of the preceding period, the second and third millennia B.C. Egyptian hieroglyphics and Mesopotamian cuneiform were not deciphered until the 19th century; until then, there was barely any known history of the earlier period, and so there was little need for a chronology to date it.
But is the Canon true? Did the kings listed in the Canon really reign when the Canon says they did? It is a fact that the Canon never has been proved true beyond the shadow of a doubt. Over the last 200 years, however, enough evidence has accumulated to convince all mainstream historians that the Canon is fundamentally sound.
In the late 19th century, a couple of decades after the decipherment of cuneiform script, Babylonian cuneiform texts of astronomical purport surfaced.1 The key texts arrived at the British Museum between 1876 and 1882, mostly from excavations at Babylon. Among these texts are the so-called Diaries, in which Babylonian priest-scholars recorded celestial events day by day for some 10,000 lunar months, or 800 years, from the eighth century B.C. to the first century A.D. (This is the longest uninterrupted research project in 031human history.) It is in the Diaries and related texts that the proof for Ptolemy’s Canon is to be found.2
The astronomical events reported in the Babylonian Diaries include combinations of positions of celestial bodies (the sun, the moon, the planets and the stars) that occur periodically, as well as combinations that have occurred only once in human history. On this basis, modern astronomers, who understand the movements of these bodies, can count back from the present time to when an event occurred. The dates in the Diaries match the dates obtained by modern computation, thus proving the accuracy of the Diaries.
Furthermore, the dates recorded in the Babylonian astronomical texts match the dates of Ptolemy’s Canon. For example, Ptolemy, who obviously used the chronological system of the Canon named after him, recorded a lunar eclipse that took place, according to the Egyptian calendar, on Month 7, Day 17 of Year 7 of the Persian king Cambyses (the Persians conquered Egypt in the sixth century B.C.). A Babylonian astronomical cuneiform tablet (known as Strassmaier Cambyses 400) records a lunar eclipse as taking place, according to the Babylonian civil calendar, on Month 4, Day 14 of Year 7 of Cambyses. These Egyptian and Babylonian dates, according to our modern calendar, are both July 16, 523 B.C. Both texts describe the eclipse as beginning about an hour before midnight; clearly, these two eclipses were the same event recorded independently in equally reliable systems.
Such proof of the Canon’s accuracy based on the Babylonian evidence has only been presented piece-meal so far. A full systematic proof of Ptolemy’s Canon is a task that still lies ahead, though there is no reason to suspect at this point that the Canon is in anyway erroneous.
As the Babylonian astronomical tablets were being studied, another type of chronological evidence came to light: fifth-century B.C. Aramaic papyri from Egypt, many of which were found on the southern Egyptian site of Elephantine Island.3 These papyri contain double dates recorded in both the Babylonian lunar calendar and the Egyptian non-lunar calendar. These dates match; that is, modern scholars’ understanding of the Egyptian dates matches modern scholars’ understanding of the respective Babylonian dates. So our interpretation of both calendars is very likely correct.
By a happy confluence of circumstances, the Persian Achaemenid king Cambyses conquered Egypt around 525 B.C., and the Persian kings then ruled both Egypt and Babylon for more than a century. Asa result, Persian kings are dated by Egyptian sources, Ptolemy’s Canon and Babylonian astronomical texts—all at the same time. This intertwining and overlap allows us to fix Egyptian chronology back to 525 B.C.
Another set of texts takes us back to 664 B.C.: inscriptions in the burial chambers of the sacred Apis bulls in Memphis. The texts list the birth date, death date and life-span of the Apis bulls. By manipulating this information, we can string together the reigns of Egyptian pharaohs with certitude back to 664 B.C.
That year, 664 B.C., may well mark the most important chronological division in all of human history. It is the beginning of day-exact chronology for Egypt. In other words, if we have an ancient date for an event, then it is possible to establish to the day how long ago, counting from today, the event occurred. For example, Psammetichus II is known to have died on 032Month 1, Day 23 of Year 7 of his reign—or on February 9, 589 B.C., in the fictional extension of our calendar back into the past.
Before 664 B.C. only approximate dates are possible. Assyrian year-lists and radiocarbon dating, however, inspire confidence in this chronology.4 The Assyrian year-lists help to establish or confirm Egyptian chronology back to about 1500 B.C., because Anatolian and Mesopotamian kings corresponded with Egyptian pharaohs. That makes it possible to know which Assyrian kings are contemporary with which Egyptian kings. These matches are called synchronies.
For dating the second and third millennia B.C., the current model of ancient Egyptian chronology relies on so-called Sothic dating.
Sothic dating is the most characteristic concept of Egyptian chronology. The term “Sothic” means “pertaining to the rising of the goddess-star Sirius.” “Sothic” derives from Sothis, the Greek form of the Egyptian Spdt, which is the Egyptian name for Sirius. Each year, as the earth revolves around the sun, Sirius becomes obscured by the sun for about two months. The star then reappears one morning in July, just before sunrise. This first visibility of Sirius is called “the rising of Sirius”; the Egyptian term is prt Spdt, or“the coming forth of Sirius.”
Sothis was an important goddess, often associated with Isis, who each night revivified her brother-husband Osiris. Ancient Egyptians took note of the return of the star, presumably happy that Sothis had not left them for good, by recording the date of the rising of Sirius; such a record of arising according to the Egyptian 365-day civil calendar is called a Sothic date. Although not many Sothic dates have been preserved, two stand out. The first is Month 8, Day 16 of Year 7. Although the king is not named, most 033scholars agree that he must be the Middle Kingdom pharaoh Sesostris III (c. 1836–1818 B.C.), because the handwriting of the document containing the date resembles that of other documents dating to the reign of Sesostris III. The second Sothic date is Month 11, Day 9 of Year 9 of the New Kingdom pharaoh Amenhotep I (c. 1514–1493 B.C.), who is named.
Sothic dating is the dating of Egyptian history in relation to our time by extrapolating data from Sothic dates. Sothic dates can be used as anchors for Egyptian chronology. The rising of Sirius occurs once a year, about every 365.25 days, but the Egyptian year is only 365 days long. The rising therefore shifts forward to the next calendar day after about four years. A Sothic cycle is the time that it takes for the rising of Sirius to return to the Egyptian new year—or, to put it the other way, for the Egyptian new year to wander back around the calendar until it returns to the Sothic date. That happens after about 1,460 years (assuming the calendar itself was not altered). If the rising shifts forward by one day every four years, then every set of four years in Egyptian history has its own Egyptian date for the rising. Once we have an Egyptian date of a rising, we can assign that rising to a certain set of four years in Egyptian history. Since the rising shifts forward by one day every four years, we only need to know when it fell at one point in Egyptian history to roll the wheel backward and forward and know when it fell at all other times in Egyptian history. From a report by the Roman author Censorinus, we know that the rising fell on the Egyptian New Year’s Day in 139 A.D.h
Ptolemy’s Canon, Babylonian astronomy, Assyrian year-lists, radiocarbon dating, texts commemorating the deaths of Apis bulls, and Sothic dates—these provide the grounding of the current model of ancient chronology. Back to the seventh century B.C., this model is very strong. (For the prior periods, the strands of evidence are difficult and complex—but that’s another story.) The discipline of ancient history would be stronger if more people understood how this model works and why most scholars are convinced of its reliability.
It is responsible, after all, for the earliest dates in the history books.
Ancient history books are full of dates. One reads, for instance, that the Great Pyramid at Giza was built around 2500 B.C., not 3000 B.C. or 2000 B.C., and that Jerusalem was captured by the Babylonians in 586 B.C., not 566 B.C. But how do we know these dates are accurate? Couldn’t ancient history have been hijacked years ago by a conspiracy 028of chronology whizzes who set in motion an endless succession of wrong dates? Compared to other inductive sciences, like biology or geology, ancient chronology is triply handicapped. The first problem is that it is non-replicable: We cannot […]
You have already read your free article for this month. Please join the BAS Library or become an All Access member of BAS to gain full access to this article and so much more.
Eusebius/Jerome counted years from the first Olympiad, beginning in 776 B.C. They dated events as occurring during the first, second, third or fourth year of a specific Olympiad. The notion of counting years A.D.—that is, Anno Domini, or “in the year of the Lord”—from the birth of Jesus was the invention of Dionysius Exiguus (Dennis the Little), a monk who lived in Rome in the sixth century A.D. (see Leonora Neville, “Fixing the Millennium,” Origins, Archaeology Odyssey, 03:01). The practice of dating the earlier era B.C.—or “Before Christ”—was not used until the 17th century A.D.
2.
Although Joseph Scaliger knew only a defective version of this king list, by the early 17th century a number of correct versions were available to scholars. The three earliest correct manuscripts, written in Byzantine uncial script, date from the eighth to tenth century A.D.; they are kept in Leiden (Leidensis BPG 78), the Vatican (Vaticanus graecus 1291) and Florence (Laurentianus 28–26).
3.
Obviously, on February 26, 747 B.C., there was neither a February 26 nor a 747 B.C. Our modern time-reckoning has been extended backward into the past, according to a convention first applied in the 17th century.
4.
These rulers of Egypt are of Macedonian descent, having come to power after Alexander’s conquest of Egypt in 332 B.C. The last and most famous of the rulers of this Greek/Egyptian dynasty, called the Ptolemaic dynasty, was Cleopatra.
5.
Kings normally don’t begin or end their reign on New Year’s Day. The Egyptians had an elaborate system for rounding off kings’ reigns.
6.
Although the Canon is a product of Alexandria, a city founded in the late fourth century B.C., Greek astronomers, including Ptolemy, relied upon Babylonian astronomical observations dating to before the city’s founding. The Canon thus shifts from rulers of Assyria and Babylonia to rulers of Egypt.
The margin of error of dates obtained from Sothic dating is roughly 10 to 20 years for the second half of the second millennium B.C., 40 to 50 years for the first half of the second millennium B.C., and one to two centuries for the third millennium B.C.
Endnotes
1.
The decipherment of Babylonian astronomy began at the end of the 19th century. A milestone is Franz Xaver Kugler’s book on lunar motion according to the Babylonians (Die Babylonische Mondrechnung [1900]). With such studies as Astronomical Cuneiform Texts (1955) and History of Ancient Mathematical Astronomy (1975), Otto Neugebauer of Brown University played an important role in this development.
2.
Only recently, however, have the Diaries become easily accessible, because of publications by Abraham Sachs and Hermann Hunger; the first three volumes of their Astronomical Diaries and Related Texts from Babylonia, containing most of the Diaries, appeared in 1988, 1989 and 1996.
3.
These texts can be found in Bezalel Porten and Ada Yardeni, Textbook of Aramaic Documents from Ancient Egypt, 3 vols. (Jerusalem: Academon, 1986–1993).
4.
In these lists, each year is named after an official called the limu, whose main function was to lend his name to the year (eponymously). The Assyrian year-lists are not king lists, though sometimes kings act as the limu. These lists are collected in Alan Millard, The Eponyms of the Assyrian Empire, 910–612 BC (Helsinki: Neo-Assyrian Text Corpus Project, 1994).
