Tuesday, March 06, 2012

Scientists and Bad History

Posted by Darin Hayton on 03/06 at 02:02 PM

In a previous post I worried that a group of scientists were practicing bad history: Scientists Practicing Bad History.1 As I pointed out in the opening paragraph of that post, my concerns were based on summaries of an original journal article in Weather. A person on Twitter wondered how much of my concern should be addressed to reporting on this article and how much was attributable to the original article (F. Domínguez-Castro et al., “How useful could Arabic documentary sources be for reconstructing past climate?Weather 67(2012): 76–82).2

Tweet asking about the source of “anachronisms” (Source: This tweet)

Having now looked at the article, I am now convinced that these scientists are practicing bad history.

The authors have selected 10 “chronicles that narrate the social, political and religious history.”3 Most of the authors recount events that happened well before their own lives, sometimes centuries earlier. Yet no effort has been made to understand how a particular chronicler came to recount a particular event or why relating it might be useful for the chronicler’s larger narrative. These questions don’t even occur to the authors because they assume that the reporting of meteorological phenomena unproblematically reflects reality:

While the narration of human deeds in such works [historical texts] is obviously subject to personal, political and religious interpretations, and indeed bias, nothing of the kind hinders the accurate recording of heavy rains, extreme cold or a solar eclipse.4

Yet all these texts are historical texts, often written long after the events reported. Of the 14 meteorological events reported, only two were probably witnessed by the authors who recounted them. In three cases the author might have been a child when the event occurred. In other words, these documents do not simply mirror reality but reflect the various cultural and social dynamics that engraved certain meteorological reports into cultural memory while perhaps losing others to the sands of time. These texts are, after all, historical texts that were created to narrate human deeds.

There are other problems with this article that arise from the authors’ efforts to read quantitative specificity out of ambiguous historical sources. The authors use the categories drought, flood, rain, hail, cold, hot, wind, and locust to guide their interpretation of the sources. Like some Collingwood-esque automata armed with sharp scissors they cut examples of these terms out of the sources and paste them into a table.5

Table 2 lists the various weather conditions: drought, flood, rain, hail, cold, hot, wind, locust (Source: F. Domínguez-Castro et al., “How Useful Could Arabic Documentary Sources Be For Reconstructing Past Climate?” Weather 67(2012): 76–82, table on 78)

Even if these terms transparently transmit what really happened, they are inherently vague and imprecise. What counted as “cold” or “rain” for one person did not necessarily correspond to another person’s idea of those terms. Despite these limitations, the authors use “cold spells” to make specific claims about the temperature in Baghdad during July 920. Assuming that people sleep in temperatures between 18-22°C, they conclude that the cold spell in July 920 must have had nighttime temperatures below 18°C, or “a negative anomaly of up to 9 degC [sic].”6 Other assumptions about the temperature at which it snows lead them to conclude that “temperatures during historical snowfalls were around 0°C.”7 In both cases, they use modern, quantitative data to extract out of the historical record precise, numerical results. Given the ambiguity of the original sources, such inferences are difficult if not impossible to justify.

There is a more profound problem with this article. The authors assume that meteorological phenomena reported in historical sources transparently record physical reality, i.e., nature. Despite claiming that these documents “narrate the social, political and religious history,” the authors assume that the narration suddenly becomes unbiased when talking about “heavy rains, extreme cold or a solar eclipse.” They take references to “snow” and “frozen liquids” and “frozen rivers” as “accurate recording” of meteorological phenomena that really occurred. Their sources were and remain historical chronicles that put human deeds into a particular context in order to inform how subsequent people interpret those deeds. For example, Hartmann Schedel regularly invokes meteorological phenomena because they are prodigious and were thought to be connected to the political and religious history. In his Nuremberg Chronicle, he often reports that comets appeared to signal the deaths of rulers. Milk and blood fell from the sky, presaging a serious pestilence in Rome that resulted from divine wrath. Another time “Fiery rays were observed in the heavens and blood fell in drops from the clouds. This presaged that human blood would be spilled afterwards.”

A page from Hartmann Schedel’s Nuremberg Chronicle that reports fiery rays and blood drops from the clouds.

Examples of such narrative uses of meteorological phenomena are legion through out the historical record. In other words, the seeming natural, unbiased context cannot be divorced from the obviously biased human content. The entire source is subject to human bias—the personal, political and religious biases of the author of the original text. The authors of “How Useful Could Arabic Documentary Sources Be For Reconstructing Past Climate?” fail to question how and why the particular events came to be reported as they were in the historical sources. They deny these events any role in constructing or shaping the meaning of a text or a passage within that text. The authors of those historical chronicles chose to report certain phenomena. They did not accidentally insert them into the narrative. Failing to ask why these phenomena were chosen or to ask how they shaped the narrative is to practice bad history.

1Here I am concerned only with the practice of history and not with the scientific merits. As logic and inference are powerful tools in both the practice of history and the practice of science, some of my criticisms here will apply to both.
2In general, I am skeptical of historical documents providing useful specificity, especially quantitative specificity. My skepticism is reinforced when the quantity in question could not have been measured because the concepts or the instruments did not exist to measure that quantity. Such is the case with temperature. In this case, whether the term used is “chilly” or “cold” or “hot” or “windy” or even “freezing,” there is too much ambiguity and subjectivity to be of much use. I have other concerns with the interpretation of the historical record, interpretation that had to occur in the original article and could only be repeated in the subsequent reporting. Consequently, I see little difference between the reporting on the original article and the original article.
3F. Domínguez-Castro et al., “How Useful Could Arabic Documentary Sources Be For Reconstructing Past Climate?” Weather 67(2012), 76.
4Ibid, 77.
5R.G. Collingwood, The Idea of History.
6F. Domínguez-Castro et al., “How Useful Could Arabic Documentary Sources Be For Reconstructing Past Climate?” Weather 67(2012), 81.
7Ibid., 80.

Tags: historiography, medieval weather, meteorology, r.g. collingwood


In the Blogs

Contributor Login

Recent Entries

Current Contributors


Recent Tags


  • The views and opinions expressed on this blog are strictly those of their respective authors and do not necessarily reflect the views and opinions of the Philadelphia Area Center for History of Science.