Empires of Climate: Some Reflections on Periodization and the Study of Ancient Climate Change

When I began researching the topic of climate change in the historical ancient Mediterranean world (traditionally dated to between roughly 800 BCE and CE 400 --  before lies the prehistoric period and after the medieval period), I proceeded to collect as much scientific data as possible on the climate of the Mediterranean in an effort to try to understand some of the broader trends of the last few millennia. I quickly discovered that one of the most important distinctions in the palaeoclimatology of Europe and the Middle East is between the instrumental data produced in the last 500 or so years (so called because they were obtained using instruments that measure basic meteorological phenomena) and the proxy data (so called because they are derived from indirect indicators of past climate), which scholars must rely on for the most part in the absence of, though which are oftentimes only a poor substitute for, instrumental data before around 1500 CE. It thus seemed at first that all scholars of the pre-Modern European and Middle Eastern worlds interested in climate would be roughly on an equal footing, dealing with evidence derived from the analysis of proxies such as tree rings, the layering of lake sediment, and the growth of stalagmites and stalactites. But over time I came to realize a distinct pattern: detailed palaeoclimatological graphs of proxy data, if they do extend back to the ancient world, almost always extend back 2,000 years and rarely further. This fact would likely be little cause for consternation among climate scientists, but for ancient historians it has major consequences.

The reason for extending back 2,000 years is clear: this marks the BC/AD or BCE/CE divide, which has remained the single strongest chronological division in Western history and a satisfying partition between ages. This chronological scheme centres, of course, on the supposed birth date of Jesus Christ, hence the designation “before Christ” and “anno domini” (“in the year of our lord” in Latin). This system was invented in the early 6th century CE, when a monk known as Dionysius Exiguus decided to fix the date of Jesus’ birth in an effort to facilitate the proper calculation of Easter (Mosshammer 2008). (Strictly speaking, only the designation AD was invented at this point – years BC were not labeled at all until much later; hence the awkward mixture of English and Latin in BC/AD.) Despite its simplicity when compared to other contemporary dating conventions, however, this system was not widely adopted before the 17th century, largely because many scholars disputed Jesus’ actual birthdate and therefore considered it flawed. Indeed, when another much later Dionysius, this one surnamed Petavius, argued for the use of the BC/AD system in his 1627 work Opus de Doctrina Temporum, he did so while noting that the birth of Christ represented merely a useful if arbitrary date around which a consensus had formed, and not the actual date of the historical figure, which most came to place c. 6-4 BC.


While the importance of Christian chronology diminished in the succeeding centuries and this scheme was in the twentieth century changed superficially to the more culturally neutral terms BCE/CE (“before the common era” and “common era”), as scholars have noted, the BC/AD system, along with the much younger recognition of the century and the decade as meaningful analytical categories, has had a major impact on the way we think about time and history (Wilcox 1987; Zerubavel 2003). Most notably, it holds a particular significance in popular thinking as a convenient marker of antiquity: you only need to watch a History Channel program on the ancient world or read about a new archaeological discovery in a newspaper to realize that any date within about a half a millennium on either side of 1 AD is conventionally rounded to “2,000 years ago.”

In scientific fields studying the past, a noble effort was also made to abandon the BCE/CE system altogether, instead relying on dates BP (“before present”) centred on the year 1950 (so that 1 BCE is 1950 BP). Despite this, the BC/AD divide continues to hold sway over palaeoclimatologists. This is in part because in many cases when dealing with data from the last few millennia dates are converted into BCE/CE anyway to make charts more accessible for a general readership. But another apparent influence is the fact that the BCE/CE linchpin also aligns roughly with what is considered to be the beginning of the Roman Empire under Augustus after he became the sole ruler of Rome (his reign as emperor is traditionally dated between 31 BCE and 14 CE). As the largest and most prominent state of the ancient Western world, the Roman Empire has also naturally attracted attention among historians interested in long-term connections between changes in society and climate. The so-called Imperial period of Roman history (most often dated from 31 BCE – 324 CE) was one of exceptional interconnection in Europe, the Mediterranean, and the Middle East which witnessed noticeable large-scale impacts on the natural world, as is becoming increasingly clear. For instance, atmospheric lead pollution in the Northern Hemisphere reached its pre-Industrial Revolution peak in the 1st c. CE largely due to Roman industrial activity (Hong et al 1994). Thus, the importance of this arbitrary chronological division has in turn been affirmed for many palaeoclimatologists interested in the impact of climate on pre-Modern Europe, who then continue to use it as a convenient endpoint for their data.

An arbitrary chronological linchpin, a convenient means devised to calculate the date of Easter some 15 centuries ago, has had a very real effect on the modern study of climate change in the ancient world.


This focus on the Roman Imperial period has had very real consequences. When an historian of ancient Greece interested in the centuries BCE like myself consults palaeoclimatological publications such as the reports of the IPCC (Intergovernmental Panel on Climate Change) or even the massive, broad-sweeping report produced in 2011 by the Bank of Greece entitled “The Environmental, Economic and Social Impacts of Climate Change in Greece,” one finds inevitably that graphs either include high-resolution data extending back up to 2,000 years, or low-resolution data visualized on geological timescales of tens or hundreds of thousands of years. For a historian interested in relatively short-term trends on the annual or decadal scale, the difference in quality is often akin to that between HD video and a slideshow. Thus, for anyone interested in analyzing how climate change impacted, for instance, the Athenian empire of the 5th c. BCE, or the empires of Alexander the Great and his successors in the 4th-1st c. BCE, data of sufficient quality oftentimes simply are not there; all too often they lie, tantalizingly, just beyond the edge of the graph.


Thus, an arbitrary chronological linchpin, a convenient means devised to calculate the date of Easter some 15 centuries ago, has had a very real effect on the modern study of the impact of climate change in the ancient world. This is beginning to change as climate scientists push their investigations back further chronologically – almost always by well-rounded blocks of 500 years – but, for the moment, scholars of the pre-Imperial Roman Mediterranean world cannot help but be frustrated by the persistence of the BC/AD division. This phenomenon is a reminder of the continuing capacity of the lingering traces of the ancient world to affect its own study, even in such a rigorously scientific field as palaeoclimatology.

Select Bibliography

Hong, S. et al. 1994. “Greenland Ice Evidence of Hemispheric Lead Pollution Two Millennia Ago by Greek and Roman Civilizations.” Science 265, 5180 (1994): 1841-3.Mosshammer, A.A. 2008. The Easter Computus and the Origins of the Christian Era. Oxford: Oxford University Press.

No author. 2011. The Environmental, Economic and Social Impacts of Climate Change in Greece. Athens: Bank of Greece.

Stocker, Thomas F. et al. 2013. Climate Change 2013: The Physical Science Basis. Working Group I Contribution to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge: Cambridge University Press, 2013.

Wilcox, D.J. 1987. The Measures of Times Past: Pre-Newtonian Chronologies and the Rhetoric of Relative Time. Chicago: University of Chicago Press.

Zerubavel, E. 2003. Time Maps: Collective Memory and the Social Shape of the Past. Chicago: University of Chicago Press.