Science in the Archives reveals affinities and continuities among the sciences of the archives, across many disciplines and centuries, in order to present a better picture of essential archival practices and, thereby, the meaning of science. For in both the natural and human sciences, archives of the most diverse forms make cumulative, collective knowledge possible. Yet in contrast to laboratories, observatories, or the field, archives have yet to be studied across the board as central sites of science. The volume covers episodes in the history of astronomy, geology, genetics, classical philology, climatology, history, medicine, and ancient natural philosophy, as well as fundamental practices such as collecting, retrieval strategies, and data mining. The time frame spans doxology in Greco-Roman antiquity to NSA surveillance techniques and the quantified-self movement. Each chapter explores the practices, politics, economics, and open-ended potential of the sciences of the archives, making this the first book devoted to the role of archives in the natural and human sciences.
|Publisher:||University of Chicago Press|
|Product dimensions:||6.00(w) x 8.90(h) x 1.10(d)|
About the Author
Lorraine Daston is director of the Max Planck Institute for the History of Science in Berlin and is visiting professor in the Committee on Social Thought at the University of Chicago.
Read an Excerpt
Science in the Archives
Pasts, Presents, Futures
By Lorraine Daston
The University of Chicago PressCopyright © 2017 The University of Chicago
All rights reserved.
Astronomy after the Deluge
Of all practitioners of the sciences of deep time, astronomers may well be bottom-trawlers of the highest order. Astronomy's empirical basis consists of signals (radiation and high-energy particles) that travel long and far to reach earth-bound observers. The possible meanings of such distant celestial messages are often obscure to those who record them, their decryption and pattern dependent on the hindsight of some future generation of sky watchers. Whether one has in view the orbital motion of our nearest celestial neighbor or the beginnings of the universe itself, the longue durée of astronomical processes seems to demand an archive to match, one that ideally comprehends all observable celestial phenomena.
Plumbing astronomy's archival depths, however, has posed challenges. A 1998 progress report on the inclusion of the "historical literature of astronomy" in NASA's Astrophysics Data System characterized "old" items as those dating from before 1940. Few astronomers today venture into what one of its leading practitioners calls "applied historical astronomy" to identify and extract material from distant disciplinary, linguistic, and cultural substrates. Consider the fragmentary character of cuneiform testimony to the Babylonian astronomical legacy; frustratingly laconic remarks on eclipses and comets buried deep in classical Greek and Roman sources as well as medieval European chronicles; works in Persian, Arabic, Syriac, and Sanskrit composed across the Islamic world; and the possible manipulation of Chinese celestial observations during the imperial period for their political significance: such factors make contemporary data mining difficult, to say the least. The relief experienced by contemporary applied historical astronomers in handling early modern European records is palpable with the inception of the telescopic period, which brought "high quality observations" timed with enough accuracy and described with sufficient specificity to make them "usable."
There surely was a sea change in early modern European astronomical practices that made the resulting material more familiar to modern eyes. Starting in the sixteenth century, reiterated observation of daily as well as long-term periodic phenomena gradually became the norm. Expectations for precision measurement also rose dramatically, a shift accelerated less by the telescope itself than by other innovations: micrometers that ran parallel lines across a telescope's optical field to allow for fine-grained measurements of angular separation; the application of telescopic sights to angle measuring instruments such as quadrants and sextants; and the pendulum clock for timing celestial events. And yet the total astronomical archive with its exhaustive compilation of all observed celestial phenomena is a relatively recent desideratum, one whose logic is hardly inevitable. This essay studies its emergence from the approach to empirical research that dominated astronomical practice from classical Greece through the modern era: the canon of data curated for their probative value. The family resemblances between different modes of presenting astronomical observations hide a multitude of motivations that are often at odds with the modern archival ideal. Teasing them apart will help us better understand the complexities in the truism that "all astronomers use historical data."
The Canonical Archive
You noticed, one doesn't write, luminously, on a darkened field. Only the alphabet of stars reveals itself that way, sketched or interrupted: man pursues black on white.
— Stéphane Mallarmé, "L'Action restreinte" (1897)
The Roman historian Flavius Josephus set the origins of celestial science in the depths of antediluvian time. Claiming in his Judean Antiquities that God had given the patriarchs such exceedingly long lives in part so they could make "discoveries" in astronomy and geometry, Josephus recounted how Seth's sons tried to safeguard their hard-won knowledge of the heavenly bodies from Adam's prediction of world-destroying fire and flood by inscribing their discoveries on two stelae, one brick, the other stone. Should the brick pillar be lost to the waves, the stone one would "remain and offer an opportunity to teach men what had been written on it and to reveal that also one of brick had been set up by them. And it remains until today in the land of Seiris." Whether the stone stele told of knowledge forever lost with its brick counterpart in the receding waters, Josephus did not say, nor whether anyone could read its inscriptions following the confusion of tongues. But the threefold moral burden of postdiluvian astronomy for Josephus was unmistakable: to record, preserve, and pass one's findings on to future generations.
Amongst the earliest memory-traces of astronomy's archival dreams are the wedge-shaped marks that Mesopotamian scribes pressed into clay tablets, a form of inscription thought to parallel the "heavenly writing" the gods constellated upon the night sky. Greek and Roman scholars testified wonderingly to the cumulative weight of such ancient star watching. Diodorus Siculus noted in his massive universal history that the Chaldaeans of Babylon had "observed the stars over a long period of time" (and included their claim — scarcely believable, he remarked — to 473,000 years of such study), while Pliny the Elder relayed reports by some that the Babylonians possessed 720,000, others 490,000, years' worth of celestial observations. Rumors of such Babylonian treasures enjoyed a long life in a wide range of textual traditions. But where was this famed archive of 'heavenly writings'? Aristotle referred to the Egyptians and Babylonians, "who have long made observations over a very great number of years and from whom we have many reports about each of the heavenly bodies" to support his characterizations of planetary motion, and asked his nephew Callisthenes, traveling in Alexander the Great's company, to send such records from Babylon where, according to Porphyry, they had been preserved for 31,000 years. Yet the material seems never to have arrived, a point Simplicius deployed centuries later in his commentary on Aristotle's On the Heavens to excuse faulty efforts to "save the phenomena." Another millennium passed and scholars still waited for tidings from Babylon. When the French astronomer Jean-Sylvain Bailly repeated claims for the antiquity of the Chaldaean astronomical tradition in his 1775 history of ancient astronomy "from its origin," he summarized learned opinion in giving them little credence, pointedly remarking, "We have few details on the nature of the Chaldaean observations."
It was not until the late nineteenth century that the long-buried celestial riches Pliny described as "inscribed on baked bricks" began to be painstakingly deciphered from cuneiform tablets. These include daily systematic observations of planetary phenomena, extant though far from continuous for the period from 652 to 61 BCE in a corpus of about 1,200 texts now known as the "astronomical diaries," a rendering of the term "regular watching" written on the edges and ends of the tablets. The astronomical diaries, in turn, seem to be the general empirical foundation for other kinds of cuneiform texts, some drawn from the diaries and largely observational in character; others predictive in anticipating future celestial events on the basis of observed periodic phenomena; and still others comprising computed ephemerides of planetary phenomena in tabular form, together with related procedure texts indicating rules and parameters for their construction. Specialists in this area are keenly aware of the partial nature of their object of study. Most of the known texts survive as physical fragments long made inaccessible to scholars by the British Museum. A murmur of this difficult history is audible in an 1974 article by the preeminent scholar of the diaries for most of the twentieth century, Abraham Sachs, in which he published photographs of "two pieces rejoined for a few hours" thanks to the "cooperation of museum officials" from the British Museum and the Wellcome Historical Medical Museum and Library, institutions but "5 minutes' walk" from each other. The laborious reconstruction of physical tablets and textual corpora comprises a significant aspect of scholarship on Babylonian celestial science, notably Otto Neugebauer's and Sachs' editions of nondivinatory cuneiform material in 1955 and Sachs' projected publication of all the known astronomical diaries and related texts, an endeavor continued by Hermann Hunger. Together with editions of divinatory material, such compilations now constitute the archival landscape for Babylonian celestial science.
Just how isomorphic these modern archival formations are with the record-making and record-keeping practices of scribes at work in the complex political and social circumstances of ancient Mesopotamia is unclear. The late nineteenth- and early twentieth-century excavations that uncovered most of the relevant material were largely unsystematic, as were the processes by which the tablets arrived at their present institutional homes; neither their representativeness nor most details of their excavation sites are known. Absent archaeological context, reconstitution of both tablets and texts has depended heavily on physical evidence, philological analysis of linguistic and generic characteristics, mathematical and technical consistencies, and correlations with modern retrodictions of astronomical phenomena. Yet these sundry approaches to reuniting material sundered by historical circumstance have been bound together by an archival conception far broader than the historically contingent and culturally specific bounds of ancient Mesopotamia. Neugebauer announced this ambitious research agenda with the 1962 publication of a curious ephemeris for planets visible to the naked eye that gave their positions from 601 BCE to 1 CE as seen at 7 P.M. "Babylon civil time" — "chosen because of the evening epoch of the Babylonian lunar calendar" — and in the Julian calendar, "which is standard for historical work." Computed by the mathematician Bryant Tuckerman at the Institute for Advanced Study and the IBM Research Center, the ephemeris drew on modern planetary theories and methods of handling selected periodic perturbations to compute positions with a precision "somewhat — but not excessively — better than that to be expected from the material to be studied, which is based upon naked-eye observations." Identifying records of ancient celestial events was but the first step; the ultimate goal was nothing less than the restoration of "a huge archive of well over a thousand texts to its full usefulness for the astronomer as well as for the historian of astronomy and the historian of the Hellenistic age ... in all its aspects, philologically, historically, and astronomically."
While the first installment of the Tuckerman tables opened the door to dating political, military, and economic changes and establishing connections with other textual corpora and genres of Mesopotamian celestial science, the larger relevance of Babylonian observational material was soon made clear. The tables' second volume gave planetary positions through the period in which astronomical work took place at the "important meridians" of Toledo, Greenwich, Hven, Prague, Constantinople, Alexandria, Damascus, Samarkand, and Ujjain; its seventeenth-century terminus "was chosen in order to include the observations of Brahe and Kepler which form the foundation of [Kepler's] 'Astronomia nova.'" Allied efforts further extended the Tuckerman tables' coverage to nearly three millenia, thanks to the collaborative labor of astronomers, computer scientists, and historians. By setting in motion these twinned projects — critical editions of ancient texts on the one hand, and long-term ephemerides based on modern planetary theories and computing technologies on the other — Sachs and Neugebauer put Babylonian astronomical observations on the same footing as those made anywhere or anytime, making an immense corpus of observational data visible to the contemporary astronomer.
As Neugebauer pointed out, the problem of secular accelerations — long-term variations in observed planetary positions resistant to control by gravitational theory — offered a key example of how historical records could advance astronomical knowledge. In 1695 Edmond Halley published his suspicion that the moon was moving faster in its path than it had in the past, based in part on his evaluation of observations made by the medieval Muslim astronomer al-Battani. Half a century later Richard Dunthorne, Tobias Mayer, and Joseph-Jérôme Lalande drew on ancient and medieval eclipse observations to empirically determine values for the moon's secular acceleration. These early efforts indicated just how modest the acceleration at issue might be, perhaps 9 or 10 arcseconds of celestial longitude per century squared (even the modern value of about 26 arcseconds is a minute fraction of the moon's apparent diameter of about 1,800 arcseconds as seen from earth). Yet they also suggested just how valuable a set of observations made over a long timescale could be, all the more so when efforts to predict the quantity in terms of gravitational theory proved controversial. While attempts through the nineteenth century and into the twentieth to run secular acceleration to ground revealed significant disagreement over the probative value of old observational material, the Tuckerman tables and its successors held open the possibility of its eventual resolution through "the recovery of great masses of detailed eclipse records, accurate data for phenomena like occultations, etc." that would "eventually give reliable early elements for the testing of empirical constants related to the problem of secular acceleration." The Tuckerman tables themselves have long been superseded by ephemerides computed with techniques able to address the accuracy of planetary positions obtained by radar and laser ranging as well as optical measurements. Yet the integration of historical and contemporary astronomical work projected through their gridded columns has developed rapidly. Today's practitioners of applied historical astronomy expect to draw from a totalizing archive of serialized celestial observations spanning centuries, continents, and cultural contexts, the long-term historical analogue to contemporary synoptic sky surveys that sweep the entire sky every few days to create a "panoramic cosmic cinematography" of celestial events.
The successful configuration of Babylonian astronomical material as "the first and longest-running data collection project in world history" capitalized on a form of archival desire concretized in the modern genre of the eclipse canon. Meant for scholars intent on eclipse records as a means of securely dating historical events, Theodor Oppolzer's Canon of eclipses (1887) provided instructions for using its tabulated data to calculate the visibilities of 8,000 solar and 5,200 lunar eclipses between 1207 BCE and 2161 CE with precision. In contrast with earlier such aids, Oppolzer chose the term "canon" to signal the work's status as "a standard for all future investigation": "the Canon, to justify its title, ought contain all possible eclipses taking place within its time span, even though the greater part of them have no relevance." What counted as "relevance" was, to be sure, a product of Oppolzer's skewed sense of both the past and future of celestial observation. In mapping the approximate paths of solar eclipses, Oppolzer declared that "it would be meaningless to graph the eclipses that take place in the southern hemisphere," dismissing the possibility of observers at Santiago, Sydney, the Cape of Good Hope, and other points south. The implied cultural canonicity in Oppolzer's remark was explicit in the work of his close collaborator Friedrich Karl Ginzel, whose 1899 "special canon" paired a table of revised eclipse elements and visibilities at Babylon, Memphis, Athens, and Rome —"the principal towns of the civilized world" between 900 BCE and 600 CE — with a collection of a hundred or so eclipse records drawn from cuneiform tablets and classical texts, each matched to an Oppolzer eclipse number. Yet the notion of a comprehensive astronomical archive has underwritten the ongoing production of eclipse canons engaged directly with Oppolzer's work. Of these, NASA's Five Millennium canons are the most spectacular example, listing a total of 11,898 solar and 12,064 lunar eclipses potentially visible to earth-bound observers between 2000 BCE and 3000 CE. The canonical archive that such eclipse canons project is slowly being realized through the patient work of scholars working to produce the historical counterpart to the federated datasets of today's "virtual observatory."
Excerpted from Science in the Archives by Lorraine Daston. Copyright © 2017 The University of Chicago. Excerpted by permission of The University of Chicago Press.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
Table of Contents
Introduction: Third Nature Lorraine Daston 1
I Nature's Own Canon: Archives of the Historical Sciences
1 Astronomy after the Deluge Florence Hsia 17
2 The Earth as Archive: Contingency, Narrative, and the History of Life David Sepkoski 53
3 Empiricism in the Library: Medicine's Case Histories J. Andrew Mendelsohn 85
II Spanning the Centuries: Archives from Ancient to Modern
4 Archiving Scientific Ideas in Greco-Roman Antiquity Liba Taub 113
5 Ancient History in the Age of Archival Research Suzanne Marchand 137
6 The Immortal Archive: Nineteenth-Century Science Imagines the Future Lorraine Daston 159
III Problems and Politics: Controversies in the Global Archive
7 The "Data Deluge": Turning Private Data into Public Archives Bruno J. Strasser 185
8 Evolutionary Genetics and the Politics of the Human Archive Cathy Gere 203
9 Montage and Metamorphosis: Climatological Data Archiving and the U.S. National Climate Program Vladimir Jankovic 223
IV The Future of Data: Archives of the New Millennium
10 Archives-of-Self: The Vicissitudes of Time and Self in a Technologically Determinist Future Rebecca Lemov 247
11 An Archive of Words Daniel Rosenberg 271
12 Querying the Archive: Data Mining from Apriori to PageRank Matthew L. Jones 311
Epilogue: The Time of the Archive Lorraine Daston 329