Friday, December 03, 2010

Dawn of the Living Brahe: Retro-Diagnosis that Refuses to Die

Posted by Darin Hayton on 12/03 at 10:25 PM

Last week the NY Times columnist John Tierney jumped on the “Who killed Tycho Brahe” bandwagon. Tierney’s “Murder! Intrigue! Astronomer?” would have been funny if it had been meant as a piece of “comic relief,” but he seems to believe that these recent efforts to discover what killed Brahe and thereby who killed Brahe are worthwhile and interesting. He assumes that the science is unassailable and that the fantastical story offered in Heavenly Intrigue is believable. Moreover, he relegates those who actually know something about the past—both of whom reject the notion that Brahe was intentionally poisoned—to a brief mention toward the end of the article and at the end. Instead, Tierney trots out the claim that Kepler killed Brahe to get his observational data and to make himself famous. Let’s be clear about it: there is no evidence for such a claim, regardless of what the authors of Heavenly Intrigue might write. Further, there is little new in the claim that Brahe might have died of mercury poisoning. And finally, the desire to diagnose Brahe’s illness and to attribute it to mercury is no more believable now than it was two decades ago.

J.R. Christianson pointed out back in 1998 that there was a long tradition of speculation in Scandinavian scholarship about Brahe’s poisoning and possible murder.1 As early as 1955 scholars were retro-diagnosing Brahe’s illness. Based on their reading of the three documents traditionally thought to be the best sources recounting Brahe’s death, the Danish historian of medicine E. Gotfredsen and two urologists concluded that Brahe had suffered from prostatic hypertrophy and subsequent uremia.1a This seems to be yet another example of scholars finding just what they want to find and scientists spotting what they have been trained to see.2

A pair of articles by B. Kaempe et al. in 1993 seems to have launched the most recent wave of speculation. Kaempe and his colleagues reject Gotfredsen et al.’s conclusion about prostatic hypertrophy because Tycho was “a little to young for prostatic hypertrophy” but accept the uremia diagnosis because “the cause of uremia could be poisoning with inorganic compounds especially mercury (from self testing the elixir [one of the plague compounds Brahe was reported to have developed]).”3 Once again, scientists accustomed to seeing the world through certain lenses looked at Brahe through them. They analyzed a small sample of Brahe’s mustache—by dissolving 0.1233 grams, or about 4 cm in nitric acid—and measured the mercury concentrations at 6.25 μg/g of mustache. This, they claimed, concentration is considerably “higher than the normal value for danes [sic].” Presumably based on their assumption that Brahe was poisoned 11-12 days before he died and an assumption about the growth rate of hair, they claim that this concentration of mercury would have been located in the first 0.5 cm of the hair, prompting them to claim that the concentration in Brahe’s mustache hair exceeded 50 μg/g. Clearly “the uremia of Tycho Brahe probably can be traced to a poisoning with mercury by all accounts arised [sic] by Tycho Brahes [sic] own experiments with his elixir 11-12 days before death.”4 This looks suspiciously like an example of Maslow’s Maxim about hammers and nails. At no point did Kaempe and colleagues bother to question the uremia diagnosis. Rather they selectively rejected the urological explanation for it and then re-diagnosed the problem through their own toxicological framework.

Thanks to modern science, mercury was the obvious and unassailable cause of Tycho’s death. The only task that remained was for scientists to refine the analysis. Jan Pallon was the first to take up that task (see my previous post on Brahe for a summary of Pallon’s early work). One report of Pallon’s study reveals what should be the tentative nature of the findings. Of the several hair strands analyzed using the PIXE technique, only one strand “exhibited very high local concentration of mercury.” The researchers claim that the rise in mercury concentration was quick, “maybe 5-10 minutes,” as was the decline. Assuming a constant rate of growth in the hair up to the moment of death, Brahe’s exposure to the mercury must have occurred a day before he died. More precisely, they noticed a spike in mercury levels ca. 13 hours before Tycho died:

A graph showing the relative concentrations of mercury, calcium, iron, and sulfur—note that the unusually tall spike for Hg is an artifact of a different scale (Source: “Planetärt mord uppklarat”Bulletin Från Centrum för Yrkes- och miljömedicin Lund/Malmö 22 (2004), 9)

The advantage of a single datapoint—in this case a single hair—is how incredibly easy it is to draw any curve through it. Likewise, a well crafted graph can convey just about any information—in the graph above, the spike associated with mercury looks unusually large because the mercury scale has been magnified to make the spike visible. Pallon and his colleagues were able to specify not only that this sudden concentration occurred 13 hours before Tycho’s death, but also that the concentration in the hair might be as high as 300 μg/g.5

This was all that Gilder and Gilder needed to fabricate a story about a murderous Johannes Kepler who coveted Brahe’s observational data.6 As a piece of historical fiction, this might be acceptable. Regrettably, it was marketed as nonfiction and has been cited by at least one scientist as actual scholarship.7

Before raising the questions that really motivate this post—questions that center around why bother—it might be interesting to try to assess the scientific claims on their own terms. I confess to know little about the most of the science involved and so can offer only tentative suggestion. But having done a little reading about mercury levels in humans, which is currently a hot topic especially around contaminated spaces and with respect to consumption of certain fish, I am struck by a few seemingly salient points.8 A recent article, “Interpreting Hair Mercury Levels in Individual Patients” in Annals of Clinical & Laboratory Science 36 (2006): 248–61, raises a number of questions about the reliability of using mercury levels in the hair as a marker for levels in the blood. In particular, they note that rates of mercury deposit in the hair depends on duration of exposure and the species of mercury. They also point to the fact that concentrations of mercury found in the hair is about two orders of magnitude greater than the concentrations in the blood. They report cases of concentrations in the hair as high as 328 μg/g and 339 μg/g in patients that show no symptoms. If these numbers are useful, Pallon’s estimate of 300 μg/g don’t look so bad.

A table from a recent article reviewing levels of mercury in hair (Source: “Interpreting Hair Mercury Levels in Individual Patients” in Annals of Clinical & Laboratory Science 36 (2006), 251 here)

Citing another recent article (N. H. Gosselin et al., “Reconstruction of methylmercury intakes in indigenous populations from biomarker data” J. Expo. Anal. Environ. Epidemiol. (2005): 1-11), the authors claim: “Hair can be used to estimate the earlier methylmercury concentrations in blood, although accurate reconstruction is complex.” An older article finds that average ration of concentration in hair to body is about 137:1.9 I have not conducted an exhaustive review of the literature, but the few other articles in reputable science and medical journals don’t seem to contradict what I found here. And what I have found points to both the difficulty in using mercury concentrations in hair alone as a metric for estimating the concentration in the blood or the body and to the typically high concentrations in the hair relative to the concentration in the blood.

But the real question remains: Why does anybody care? What do we, whether historians, scientists, or the 21st-century public, stand to gain? The efforts by scientists to retro-diagnose Brahe’s illness seem to be either an effort to deny expertise to historians or a secular form of hagiography. In the first instance, the scientists are explicit in their rejection of historians’ expertise when they make statements like: “New scientific facts however question the historically believed explanations on his death.” The historical sources no longer have to be interpreted or understood in their own contexts, but rather they have to be forced to answer to the questions modern scientists put to them. Questions that often inform and shape the possible answers—when you can test for mercury, it’s amazing how often you find mercury. Historically believable has been superseded by scientific facts, just as historians will be superseded by scientists.

Another perhaps more generous way to interpret these efforts is see them as akin to the severing, preserving, and ultimate display of Galileo’s middle finger.

Galileo’s middle finger on display in Florence(Source: Museo Galileo, “IV.10 Middle finger of Galileo’s right hand”)

That is to say, scientists seem to want to compile a list of scientific saints. Carefully selecting those who contributed to the narrative of modern science, they construct their hagiographies. A key aspect of hagiographic literature is the story of the life and death of the saint. Whenever possible hagiographies dwell with morbid fascination on the unfortunate and often untimely death of the saint. And whenever possible, bodily parts of the saint are distributed to shrines and places of worship. It should be unsurprising that we even have samples of Brahe’s mustache, that in 1901 when Brahe was exhumed, samples of both Brahe’s mustache and clothes were sent to the national museum in Prague, and finally that in 1991 some of these samples were given to the Danish ambassador who delivered them to the recently opened Planetarium of Tycho Brahe.

In one case efforts to retro-diagnose Brahe’s illness seem mere antiquarian delectation and in the other it threatens to efface the discipline of history. Unfortunately, in neither instance does it amount to good history.

1J.R. Christianson “Tycho Brahe in Scandinavian Scholarship” History of Science 36(1998): 467–84 (available here) See also J.R. Christianson, “The Legacy of Tycho Brahe” Centaurus 44(2002): 228–47 (behind a paywall here).
1aE. Gotfredsen et al., “Tycho Brahes sidste sygdom og død” Fund og Forskning 2(1955): 33–38.
2This was my basic point in the series of posts on secret anatomical structures in the Sistine Chapel: here, here, and here.
3B. Kaempe et al., “The Cause of Death of Tycho Brahe in 1601” in Contributions to Forensic Toxicology: Proceedings Of The 31st Int Meeting Of The Int Ass Of Forensic Toxicologists (Leipzig: MOLINA Press, 1994), 309–15, quotations from 311.
4Ibid., 314.
5“Planetärt mord uppklarat”Bulletin Från Centrum för Yrkes- och miljömedicin Lund/Malmö 22 (2004), 9.
6Gilder and Gilder, Heavenly Intrigue: Johannes Kepler, Tycho Brahe, and the Murder Behind One of History’s Greatest Scientific Discoveries (New York: Doubleday, 2004). Informed opinion considers the book “shit.” See Thony’s concluding comment here.
7Gilder and Gilder are cited as an authority regarding both Brahe’s death and Kepler’s use of Brahe’s date. G. T. van Belle, “Interferometric Astrometry” VLTI Summer School Proceedings (2009), 2 fn. 3.
8Recent studies focus on methylmercury and so may not be applicable to other cases where the form of mercury ingested differs. At least one study claims that of methylmercury, ethylmercury, and inorganic mercury, methylmercury has the highest rates of deposit in the hair.
9H. Al-Shahristani et al. “Mercury in hair as an indicator of total body burden” Bulletin of the World Health Organization 53(1976): 105–12, see table on 110 (available here).

Tags: johannes kepler, john tierney, retro-diagnosis, tycho brahe