Variations in the Earth’s Orbit: Pacemaker of the Ice Ages.
[Palisades, New York: Pre-publication typescript for <i>Science</i>, 1976]. The pre-publication typescript of the paper that proved the astronomical theory of the ice ages — headed ‘Manuscript accepted for publication by Science (as submitted Oct. 12, 1976)’ — and the working copy of Stephen Jay Gould, who read it with a sceptic’s pencil and did not believe it. Hays, Imbrie and Shackleton demonstrated here, from two deep-sea sediment cores, that the rhythm of the Pleistocene glaciations over the past half-million years matches the slow variations in the geometry of the Earth’s orbit — the eccentricity, obliquity and precession cycles calculated by Milutin Milanković — and so revived and confirmed a theory that most geologists had abandoned. Gould, then a Harvard palaeontologist whose own work on Pleistocene land snails tracked exactly these glacial cycles, worked through the typescript in dense pencil annotation on its title page, dissecting the argument into its own four steps and concluding that the orbital signal was ‘just an effect for Milank[ovitch]’s, not claims’ — that what looks good in direction is, in his own summary phrase, ‘a major problem in magnitude (my point).’ The copy is unsigned but carries his filing mark and his hand; his verdict lands, as it happens, precisely on the weakness the authors themselves confessed. The provenance rests on the boxed filing mark ‘H5’ at the head of the title page, which matches Gould’s practice elsewhere, together with the pencil hand of the annotations; the document is not signed. What the hand records is a working analysis rather than a formal review: the typescript is a copy of the manuscript already accepted by Science, reproduced and circulated after the editorial decision was taken, and Gould has taken it apart for his own purposes, as an outside reader thinking on paper. The distinction matters, because it fixes what the object is — not a referee’s verdict solicited by the journal, but the private reckoning of a working scientist with a result he found important and could not accept. Under a heading he set down as ‘My approach’ he numbers the paper’s logic in four movements. First, the data: the two cores, the oxygen-isotope record in the foraminifer Globigerina bulloides, the summer sea-surface temperatures reconstructed from radiolarian assemblages, and the puzzling abundance of the radiolarian Cycladophora davisiana, beside which he pauses to work out for himself what the species signifies — ‘not clear what [it] means, but lots today’ where the summer surface is of low salinity and cold water lies beneath, near the Antarctic and its summer melt. Second, the prediction: the twenty-one-thousand-year precession and forty-one-thousand-year tilt cycles, and the sharp question ‘why just ecc[entricity]’, with a note on the tenth-of-a-percent insolation difference and a reference to ‘Broecker’s curve’. Third, the raw curves and their calibration, where the assumption of constant accumulation rates and the radiocarbon anchoring draw the flat marginal verdict ‘120,000 correlation (problem)’. And fourth, the results, where he sets the paper’s own variance figures against one another — twenty-three-thousand years at ten percent, forty-two-thousand at twenty-five, ‘but 50% ≈ 100,000’ — and writes ‘OK but not match prediction of 1 to 1 response.’ The objection is exact. The dominant hundred-thousand-year cycle of climate corresponds to eccentricity, which is by far the feeblest of the three orbital forcings, and the effect is far larger than any linear, one-to-one response to that forcing could produce. It is a problem of magnitude, not of direction, and it is the one weakness Hays, Imbrie and Shackleton had themselves confessed, in a section of their paper headed ‘The 100,000-year cycle.’ A lesser reader might have missed it under the weight of the result; Gould went straight to it. Gould had good contemporary reason to care, and it was not the reason a later reader might assume. Through the late 1960s and 1970s his empirical research was on the Pleistocene land snails of Bermuda and the Bahamas — Poecilozonites, the subject of his doctoral work, and later the West Indian Cerion — whose shell form shifted and then held still across the glacial and interglacial phases this paper was for the first time dating with precision. A rigorous chronology of those phases bore directly on the tempo of the evolutionary change he was documenting, and on the theory of punctuated equilibrium he and Niles Eldredge had proposed in 1972, which held that species change rapidly and then persist unchanged for long stretches rather than transforming gradually. A climate that jumped between states on an orbital beat, rather than drifting smoothly, was the kind of external pattern that bore on whether the stasis he saw in the snails was imposed from without or held from within. The deeper cast of mind that the annotations reveal — a resistance to strong quantitative causal claims, a preference for interrogating an apparent large-scale periodicity rather than accepting it — would find its fullest expression only later, in The Mismeasure of Man (1981), in his part in the fierce debate over the claimed twenty-six-million-year periodicity of mass extinctions, and in Time’s Arrow, Time’s Cycle (1987), his book-length meditation on the tension between cyclical and directional conceptions of geological time. A paper arguing that a repeating astronomical cycle paces the directional history of climate sits exactly on the fault line Gould found most interesting, and his marginal ‘just an effect … not claims’ is the compressed germ of the later scepticism. No response from him was ever published, and these annotations are his only known record of engaging the result. The problem the three authors solved was more than a century old. Joseph Adhémar had first tied the ice ages to the precession of the equinoxes in his Révolutions de la mer (1842), reasoning wrongly from the length of the seasons alone; James Croll, in Climate and Time in their Geological Relations (1875), built the first serious theory linking glaciation to the changing eccentricity of the Earth’s orbit and the way it modulates the precessional cycle. It was the Serbian engineer Milutin Milanković (1879–1958) who completed the mathematics. Trained in Vienna as a civil engineer and appointed to the University of Belgrade in 1909, he took up the insolation problem in 1912 and spent some twenty-five years computing, by hand, the solar radiation received at each latitude and season across the eccentricity, obliquity and precession cycles. He fixed on summer insolation at sixty-five degrees north as the critical quantity, on the reasoning that ice sheets are made or unmade not by cold winters but by cool summers, when a reduction in high-latitude sunshine lets some of the winter snow survive the melt season and accumulate year upon year. His Kanon der Erdbestrahlung und seine Anwendung auf das Eiszeitenproblem (Belgrade, 1941) appeared as the German army invaded Serbia; the printing house was destroyed in the bombing of Belgrade, though the printed edition itself, already stored, survived — and the book nonetheless reached the wider world only after the war, and in English only in 1969. His insolation curve at first seemed to fit the four-fold Alpine glacial chronology of Penck and Brückner, and the theory gained ground. But when radiocarbon dating in the 1950s produced Pleistocene ages that appeared to conflict with his calculated timings, and as meteorologists objected that the insolation changes were simply too small to move the climate, most geologists abandoned the astronomical theory altogether. It was, by the early 1970s, a discarded idea with a distinguished pedigree. Its revival came from the deep sea. Wallace Broecker and his colleagues, dating raised coral terraces by the uranium-thorium method in 1968, found high sea stands at the ages Milanković predicted, bypassing the ceiling of radiocarbon — the work behind the ‘Broecker’s curve’ that Gould names in the margin. Cesare Emiliani had already shown that the oxygen-isotope ratio in the calcite of fossil foraminifera preserves a record of past ice volume, and Nicholas Shackleton and Neil Opdyke had extended that record, in their study of core V28-238 in 1973, into a continuous stratigraphy running back through twenty-two isotope stages. What Hays, Imbrie and Shackleton added was the decisive test. From several hundred cores raised by the CLIMAP project they chose two from the southern Indian Ocean, RC11-120 and E49-18, sited near the boundary of the Subtropical Convergence and the Antarctic polar front, long enough and fast-accumulating enough to resolve continuously the past four hundred and fifty thousand years. In each they measured, at ten-centimetre intervals, three independent signals: the oxygen-isotope ratio of Globigerina bulloides, a proxy for the volume of ice locked up on the continents; a summer sea-surface temperature reconstructed from radiolarian assemblages through the statistical transfer functions Imbrie had pioneered; and the changing abundance of Cycladophora davisiana. Then, treating the down-core series as records in time, they subjected them to spectral analysis — the frequency-domain technique that decomposes a fluctuating signal into the periodic components that compose it. The variance of climate concentrated into three sharp, statistically significant peaks: at roughly a hundred thousand, forty-two thousand, and twenty-three and nineteen thousand years, carrying about fifty, twenty-five and ten percent of the variance respectively. These are the precise periods of the eccentricity, obliquity, and precession of the Earth’s orbit. The astronomical theory had made a prediction in the frequency domain, and the sediment record met it, peak for peak. Half a million years of climate had marched, as the authors put it, to the beat of the orbit. The authors were careful about what they had shown, and their care is the hinge of Gould’s objection. They called the changes in the Earth’s orbital geometry the fundamental cause of the succession of the Quaternary ice ages — but they insisted that the response is not linear: the changes in insolation are small, and the climate system must amplify them into the great advances and retreats of the ice. The word in their title, pacemaker, carried that reservation, and later readers would sharpen it into the doctrine that the orbit paces the ice ages rather than simply causing them. The hundred-thousand-year cycle was the hardest case, because it dominates the recent record yet corresponds to eccentricity, whose direct effect on insolation is about a tenth of a percent — too weak, by any linear reckoning, to drive the largest climatic swings. In a section headed ‘The 100,000-year cycle’ they conceded exactly this, concluding that the cycle must be driven by eccentricity through some nonlinear mechanism — a modulation of the precession signal, an internal feedback in the growth and collapse of the ice sheets — that they could not specify. That admission is the ‘major problem in magnitude’ Gould marked on the title page. It is now known as the hundred-thousand-year problem, and it remains unresolved and actively debated half a century later: some argue that the pacing of the recent deglaciations follows obliquity rather than eccentricity, others that it combines several cycles, and the amplifying role of atmospheric carbon dioxide, invisible in 1976, has since been drawn into the account. What Gould could not have known is that his marginal objection would outlive the certainty of the result it questioned. The sceptic’s pencil had found the real fault line, not a spurious one; he doubted the paper for the right reason, in the place its own authors had marked as unfinished. The three authors were the complementary specialists of a single project, and the paper could not have been written by any one of them alone. James Hays, a radiolarian micropalaeontologist at the Lamont-Doherty observatory and Columbia, was a leading figure of CLIMAP — the Climate: Long-Range Investigation, Mapping and Prediction project, the international effort to reconstruct the ice-age oceans that brought the collaboration together and supplied its cores. John Imbrie, Henry L. Doherty Professor of Oceanography at Brown, was the master of time-series and spectral analysis who extracted the periodicities from the curves and tested their significance; the story of the discovery is told in his popular account, written with his daughter, Ice Ages: Solving the Mystery (1979), and he would receive the Milutin Milanković Medal in 2003, the theory’s own honour returning to one of the men who proved it. Nicholas Shackleton, of the Sub-department of Quaternary Research at Cambridge — a great-nephew of the polar explorer, later Fellow of the Royal Society and knighted for his services to the Earth sciences — had built the high-precision oxygen-isotope stratigraphy on which the whole edifice stood, extending in his study of core V28-238 with Neil Opdyke in 1973 the record Cesare Emiliani had opened in 1955. The paper’s method — tuning the ages of a marine record to the orbital clock — became one of the most powerful tools in stratigraphy, carried forward by the same authors into the SPECMAP timescale of the 1980s and used since to date climate archives back several million years. Marking its fortieth anniversary in Nature, Mark Maslin called it ‘one of the most influential papers in the study of Earth’s past climate’; it is routinely described as the empirical proof of Milanković’s theory and the founding document of quantitative palaeoclimatology, and it has been cited in the thousands. Its finding — that changes in our climate can be induced by processes outside the Earth itself, and read in the sediment as a clock — reordered a whole science around the search for orbital signals in the rock record, and they have since been found in deposits more than a billion years old. A pre-publication typescript belongs to a small and distinct class of object: the working document of a landmark paper, caught between manuscript and print. Where a printed offprint records what was published, a typescript such as this records the paper at the moment of transmission, before the type was set — and this copy carries, besides, the critical response of a reader who mattered. That combination is uncommon in modern science, and scarcer still because of whose hand it is. Gould’s own library and papers passed intact to Stanford University after his death in 2002, catalogued there and effectively removed from commerce, so that material bearing his working annotation almost never reaches the market at all; unlike an author whose library is dispersed at auction and whose association copies recur, Gould leaves behind no such supply. No second copy of this typescript has been traced. The founding texts of the astronomical theory stand around it on the shelf — Croll’s Climate and Time in their Geological Relations of 1875, Milanković’s Canon of Insolation, Emiliani’s and Shackleton’s isotope stratigraphies, Imbrie’s own Ice Ages: Solving the Mystery — but this is the only one of them written in two hands at once, the proof and the doubt on a single leaf. What Gould set down in the margin was not a misreading but a diagnosis. The orbital theory looks right in direction and remains a problem in magnitude — the hundred-thousand-year problem he marked in 1976 is still argued over today, in journals that were not yet founded when he wrote — and the palaeontologist who would go on, in Time’s Arrow, Time’s Cycle (1987), to make the tension between recurrence and irreversible history his great subject had already put his finger, on this title page, on the place where the pacemaker of the ice ages does not quite keep time. The proof of Milanković and its most searching early doubt lie here on the same leaf, in the same pencil — the discovery and the dissent inseparable, as they so often were in the century’s science, and rarely so cleanly preserved. References: Hays, Imbrie & Shackleton, ‘Variations in the Earth’s Orbit: Pacemaker of the Ice Ages’, Science 194 (10 December 1976), pp. 1121–1132 — Maslin, ‘In Retrospect: Forty Years of Linking Orbits to Ice Ages’, Nature 540 (2016), pp. 208–210 — Imbrie & Imbrie, Ice Ages: Solving the Mystery (Enslow, 1979) — Milanković, Kanon der Erdbestrahlung und seine Anwendung auf das Eiszeitenproblem (Belgrade, 1941; English translation 1969) — Broecker, Thurber, Goddard, Ku, Matthews & Mesolella, ‘Milankovitch Hypothesis Supported by Precise Dating of Coral Reefs and Deep-Sea Sediments’, Science 159 (1968), pp. 297–300 — Shackleton & Opdyke, ‘Oxygen Isotope and Palaeomagnetic Stratigraphy of Equatorial Pacific Core V28-238’, Quaternary Research 3 (1973), pp. 39–55 — Emiliani, ‘Pleistocene Temperatures’, Journal of Geology 63 (1955), pp. 538–578 — Croll, Climate and Time in their Geological Relations (London, 1875) — Gould, ‘The Paradox of the First Tier’, Paleobiology 11 (1985), pp. 2–12 — Gould, Time’s Arrow, Time’s Cycle (Harvard, 1987) — Bennett, ‘Milankovitch Cycles and their Effects on Species in Ecological and Evolutionary Time’, Paleobiology 16 (1990), pp. 11–21 — New Dictionary of Scientific Biography (Milanković) — Stephen Jay Gould Papers, Stanford University Libraries, M1437. Reproduced typescript (280 × 217 mm), 56 leaves, printed on rectos only, with tables and diagrams in the text; stapled at the upper left corner. Caption-title, the header reading ‘Manuscript accepted for publication by Science (as submitted Oct. 12, 1976)’, with the running head ‘Hays’ and pagination throughout. The title page annotated in pencil in a rapid cursive hand: a boxed filing mark ‘H5’ at the upper right; extensive marginal notes under the heading ‘My approach’, in four numbered sections analysing the paper’s argument and concluding ‘Funny / what looks good in direction, / a major problem / in magnitude / (my point)’; and a later pencil note recording the place of publication. Leaves lightly toned at the edges, a few short closed edge-tears, else clean and sound; the staple a little rusted.
Item #6419
Price: $4,500.00
