DIO responds to Sky & Telescope’s
article by Bradley E. Schaefer (February 2002)


In its February 2002 issue, Sky & Telescope published an article by Bradley E. Schaefer on the controversy over the Ancient Star Catalog. Was the catalog observed by Claudius Ptolemy in about 137 AD, as he claimed, or did Ptolemy plagiarize from the earlier catalog of Hipparchos, as most astronomers believe? DIO has had a longstanding interest in this controversy.

Not only is Schaefer's article rife with errors and omissions on the controversy itself, but the article also made a number of false and defamatory statements regarding DIO and its publisher, Dennis Rawlins. Although Sky & Telescope editor Richard Fienberg initially agreed to run our response in its letters column, a last-minute, lawyer-inspired impulse to protect his publication (from what, we wonder?) resulted in our letter being replaced by one from S&T's own former cartoonist (whose connection to S&T's editorship was strangely unacknowleged, a journalistic faux pas somehow not repeated for DIO's E. Myles Standish).

Here is the letter that Sky & Telescope refused to publish.

February 5, 2002

Dr. Richard Tresch Fienberg, Editor
Sky & Telescope
49 Bay State Road
Cambridge, MA 02138-1200

Dear Sir,

Before writing about the longstanding dispute between DIO: The International Journal of Scientific History and the Journal for the History of Astronomy (JHA) over Ptolemy, it would have been helpful if Brad Schaefer had consulted the historical record. Much of the Dennis Rawlins-Michael Hoskin correspondence has been published in DIO 1.2 and DIO 9.1. (This, plus a more extensive reply to Schaefer’s article, are available at http://dioi.org/.)

Schaefer’s statements that Rawlins had a paper “rejected” by JHA and “then started sending abusive letters to Hoskin” are outright fabrications. Rawlins’ paper was accepted for publication; and it was Hoskin who began throwing stones, after Rawlins revealed that a completely different JHA paper was miscomputed. Hoskin accused Rawlins of a “damned lie” and cut correspondence – before learning that Rawlins’ criticism was valid, leading the paper to be recomputed by its author (JHA, June 1984). Despite our overtures (DIO 9.1 pp. 2, 4), the publishers of the two leading journals of astronomical history have not communicated in nearly two decades. We encourage everyone to read both journals; JHA does not.

It’s likely no coincidence that Schaefer has slammed DIO immediately ahead of our upcoming refutation of Schaefer’s JHA analysis of the Ancient Star Catalog. His characterization of JHA as a “premier” journal while dismissing DIO as “very small” just betrays his bias; the circulations of the two journals are of the same order of magnitude. Although Schaefer seems to believe that DIO is a one-man operation, our board contains world-renowned astronomers such as Charles Kowal (Johns Hopkins APL), discoverer of Chiron, and E. Myles Standish (JPL), whose planetary ephemeris guides NASA’s deep space probes. Finally, DIO discoveries have been recognized by display in the British Musuem, and twice on page 1 of the New York Times; that’s a record few other premier journals can match.

Keith Pickering

Editor, DIO: The International Journal of Scientific History


In addition to the attached letter, I thought I would also send you privately a list of errors (of commission and omission) in Schaefer’s article that appeared in Sky & Telescope. You may find it worthwhile to correct some of these (or not; it’s your magazine). The misrepresentations of others’ work in numbers 3 and 17 are particularly egregious, in my opinion. If you feel that this issue is deserving of an extra hundred words or so, feel free to add any of these to my published letter.

  1. Schaefer credits Hipparchos with the discovery of precession of the equinoxes. As Schaefer knows (and both of JHA’s 1982 referees agreed), there is credible evidence that precession was known to Aristarchos, more than a century earlier (see DIO 9.1). The reason for Schaefer’s omission may have been to avoid admitting that Dennis Rawlins has made significant contributions to the field.
  2. Schaefer states that Hipparchos “recorded positional data for more than 300 stars.” There are actually about 400 stars mentioned in Hipparchos’s work Commentary on Aratus and Eudoxus.
  3. John Britton’s 1967 doctoral thesis (p. 44) does not show, as Schaefer asserts, that “ordinary atmospheric refraction would cause just such a one-day error [in equinox observations] about half the time.” On the contrary, after considering refraction effects carefully, Britton concluded that refraction effects cannot account for the equinox errors in any consistent way, and that Ptolemy’s equinox “observations” must have been derived from computation. See John Phillips Britton, Models and precision: the quality of Ptolemy's observations and parameters (Garland 1992), p. 37.
  4. Equinoxes are not (and were not anciently) determined “near sunrise and at sunset” as Schaefer asserts; and that cannot be true for the equinox of 132 AD, which was allegedly observed at 2 PM. Ptolemy says his equinoxes were determined by noon observations using a “meridian” transit instrument (Almagest 3.1’s reference to I.12; Toomer p 61) which only can observe on the meridian.
  5. Therefore, given #3 and #4 above, it is also incorrect that “we have both a guilty and an innocent explanation” for the equinox error, as Schaefer asserts.
  6. Schaefer claims that R. R. Newton attempted to “oversell his case” in The Crime of Claudius Ptolemy. In fact, after a quarter century, not a single one of the book’s detailed exposures of Ptolemy’s faked “observations” has ever been shown to have been miscomputed. Time and time again, Ptolemy’s data do not agree with the sky, but do agree with his models (see our summary at DIO 8.1).
  7. Schaefer claims that it is a “blunder” for historians to apply modern standards to ancient actions; but that is only true in cases where standards of behavior have changed over time – as in the case of slavery. In the case of plagiarism, standards of behavior have not changed since at least the 5th century BC. See also DIO 2.1 p. 18
  8. Dennis Rawlins does not argue, as Schaefer asserts, that “errors in ecliptic latitudes of the stars vary sinusoidally around the ecliptic in a manner that could have arisen if Ptolemy stole Hipparchus’s star catalog”. In fact, the reverse is true: Rawlins’ simple argument (published in a refereed astronomical journal) is that errors in ecliptic latitudes do not vary around the ecliptic in a way that must be true if Ptolemy’s huge longitude errors had been caused by innocent mistakes (such as that proposed by Laplace).
  9. It is not true that the paper Rawlins submitted to JHA was rejected for publication, as Schaefer claims (see JHA advertisement of this paper in Isis 1982 March); in fact, the paper passed [double] peer review and was accepted for publication. Hoskin simply “delayed” publishing it (for 18 years. [So, instead of in the JHA, the paper finally appeared as DIO 9.1 [1999] ‡3 [pp.30f].])
  10. As discussed in the letter for publication, it is not true that Rawlins “then started sending abusive letters to Hoskin”.
  11. It is not true that DIO is “devoted to bashing the establishment.” We are devoted to high scholarship in the field of scientific history. We apply high critical standards to work from all quarters: kooks, establishment institutions, and our own.
  12. As discussed in the letter for publication, it is not true that DIO has a “very small circulation.” In the field (which is small), our circulation is wide.
  13. It is not true that DIO “is written mostly by Rawlins himself.” In fairness, this was true during the early years of DIO, but during the past six years, Rawlins has been the principal author of only a single DIO volume (vol. 10, co-published with the University of Cambridge).
  14. Although Schaefer (p. 41) cites Laplace’s hypothesis on longitudes to illustrate a “fundamental ambiguity” in the evidence, he fails to note that Laplace’s hypothesis was falsified long ago – by Rawlins, in the very PASP paper Schaefer cites elsewhere in his article (see error 9 above). In fact, the 1° longitude error no longer has any credible explanation that is favorable to Ptolemy. (Here one is reminded of the way some creationists cite Kelvin’s 19th century computation of the Earth’s age to refute evolution, while failing to note that Kelvin’s computation itself was long ago refuted.)
  15. While wasting space noting that proper motion studies are useless (p. 42) – an argument peripheral to the issue – Schaefer ignores the most useful line of evidence that is crucial to the controversy: the fractional endings argument of R. R. Newton (DIO 8.1 p. 10-11). Apparently the only reason for leaving this decisive evidence out of the article is that it is fatal to the Ptolemy case, it has no pro-Ptolemy spin, and it has never been refuted.
  16. You should be aware that Schaefer and I had an extensive exchange of e-mails last July, during which I informed him of many errors I had found in his JHA paper – some of which are utterly devastating to his entire analysis. At the time he had no compelling response, and he ended our correspondence by pleading a heavy workload. So when Schaefer claims his JHA paper is “strong” and “unambiguous” (p. 42), he’s pulling your leg. Instead of dealing with these serious problems in an honest and forthright manner, Schaefer’s reaction was to run to S&T and libel DIO as a crank journal, just before our refutation is published. (And Schaefer wonders why this issue runs on emotion instead of evidence? He should look in the mirror.)
  17. Schaefer states (p.44) that Dennis Duke’s analysis of correlated errors between the Almagest and the Commentary imply “that perhaps the majority of the star positions in the Almagest are indeed derived from Hipparchus.” In fact, Duke concluded that the fraction stolen was well over 90%, and that 99% could not be excluded. This is rather more than “perhaps the majority.” Duke’s paper is still in review, but I can send you a copy if you wish; or you can contact him directly at dduke@scri.fsu.edu. Schafer ignores the crux of Duke’s devastating new point: the southern stars have the largest errors, and are therefore the easiest to show were observed by Hipparchos. But these are exactly the stars that Schaefer claims (to a high statistical probability) could not have been seen by Hipparchos.
  18. The graphs on page 44 are misleading. If, as Schaefer implies, the only difference in the “Ptolemy’s latitude” line and the “Hipparchus’s latitude” line are different latitudes, then the two curves should not cross; they should simply be displaced left and right from each other. [DR is in the DIO minority, in questioning whether this (single) comment here (curves crossing) is a valid criticism.] Any latitude can be made to fit the data under differing assumptions of atmospheric extinction. All known pre-industrial data [from Tycho, Hevelius, Hipparchos, Eudoxus, Apollodorus, Hippocrates, even Ptolemy] indicate that extinction was quite low on at least some nights, which supports Hipparchos.)

Dennis Rawlins' letter to S&T editor Richard Fienberg