Can old history books supply us with scientific data?

This is a weekly post by Nidhal Guessoum (see his earlier posts here). Nidhal is an astrophysicist and Professor of Physics at American University of Sharjah and is the author of Islam's Quantum Question: Reconciling Muslim Tradition and Modern Science.

Just as I was reading the recent paper ‘How useful could Arabic documentary sources be for reconstructing past climate?’, and I was planning to devote my weekly post to it, Salman posted excerpts from a Science article about it. But I decided to still write about it because a few interesting issues are raised by this paper (and its approach), and because it has produced a variety of reactions.

The paper, which came out in the March 2012 issue of the journal Weather, has already been commented on in noteworthy publications and websites, such as Science, Science-Daily, Popular Archaeology, and numerous blogs. The paper itself can only be accessed by those who have institutional subscription to the journal Weather; and since it does not carry an abstract, the first page is made available at the journal’s site, here.

The idea is simple: go back to medieval historical narrations of various aspects of life in those times and regions and see if references to climate can be extracted and aggregated robustly enough to include in the climate data, since real weather measurements (temperature, precipitation, etc.) are not available for those times, having existed only for less than three centuries.

I myself was intrigued about how much solid information one could possibly find in such historical accounts, since I imagined that the writers in those times were not much interested in relating to later generations how the weather or the climate changed but rather what was happening in those times and possibly how life was like. How could that be used to obtain useful climate data?

So I read the paper, and I was actually quite impressed. (I will mention the skeptics’ arguments of doubts about this approach further below).

What the authors (all Spanish, though one Portuguese institution was in the lot) did was collect any weather/climate information mentioned in books by ten historians of the region around Baghdad in the period between the tenth and the fifteenth centuries (CE), specifically about years when the following occurred: drought, flood, rain, hail, unusual coldness, unusual heat, unusual wind, and locust. They realize that in many cases, the narrators are copying from one another, and they try to determine which ones are independent and which ones are replications.

Then, and this is the part that impressed me, they map those references into a plot of “temperature” vs. year, where the “temperature” is a level of coldness ascribed by references such as “cold anomaly”, which would count as “below normal”, “frozen liquids”, which would be lower (somewhere around 0 C), snowfall, which is lower than that, and “frozen river” (Tigris and/or Euphrates), which would imply a very low temperature.

When they do this, they find that the documents present climatic references which line up quite interestingly in specific periods, particularly that between 900 and 945 AD. Incidentally, the time of Al-Ma’mun (who reigned from 813 to 833 AD) only gets one “data” point (just about the time of his death), and it is labeled “cold anomaly”.

Having identified an interesting period when there was apparently a significant climatic dip, the authors go and search the global references for any abnormalities during that period (900 to 945 AD), and they find references to an unusually cold summer in Sweden as determined by tree-ring reconstructions. They also find other bibliographical support for a “relative minimum centered in the first half of the tenth century, immediately before the Mediaeval Warm Period.”

Seeing that their research is apparently being vindicated by other, more scientific methods, they conclude that this line of investigation, of course with rigorous checks and corroborations, can be a fruitful research path that involves historians who are experts in the life and conditions of a particular area and period along with climatologists. They mention a project titled ‘Historical Climatology of the Middle East based on Arabic sources back to 800 AD’, which is funded by the German Research Foundation.

What methodological criticism could be leveled at this kind of research? In ‘Scientists practicing bad history’ and ‘Scientists and bad history’, Darin Hayton insists that historical narrations, which are often written long after the events, are almost always subjective and are influenced by political and religious interpretations of events, hence they are scientifically unreliable. Hayton further insists that just because someone says “the river froze” does not mean it actually did. Sometimes narrations of (hi)stories are metaphorical. Furthermore, translating “not being able to sleep on the house roofs (as people in Baghdad did and still do in the summertime”) into “night temperature must have been less than 18 degrees” (the minimal temperature at which ones sleeps comfortably) is far from scientifically rigorous or acceptable (according to Hayton).

The authors, however, had anticipated this kind of critique and note that just as we do not doubt the reporting of astronomical events (e.g. eclipses, meteor showers, etc.) in historical documents, we have no reason to doubt when a historian (or better yet, several historians) report snow or hail storms in Baghdad. The reported date may be inaccurate, but absolute accuracy is not really needed here, and multiple references for a whole period is what climate studies want to have.

Interestingly, a Saudi astronomer, Dr. Hasan Basurah (a professor at King Abdelaziz University in Jeddah and a friend of mine), who was very briefly mentioned in the introduction of this paper, does similar work pertaining to astronomical/space phenomena, such as aurorae and meteoritic impacts that can be found in the literary records of the Arabian Peninsula, e.g. poems describing such events. Very low-latitude aurorae then tell us about very strong solar activity in the years when the events are reported. The solar cycle could thus be reconstructed many centuries back, though probably with sporadic references/data-points.

Perhaps I’ll check out Dr. Basurah’s work and report on it in the future. But this is quite fascinating stuff, opening new windows for both historical and scientific investigations.