Gould, Dinosaurs in the Haystack

In this case, the Alvarez theory suggested little more than hard work.

So why wasn't the effort expended before? Paleontologists are an industrious lot; we have faults aplenty, but laziness in the field is not among them. We do love to find fossils; this is why most of us entered the profession in the first place. We didn't scrutinize every inch of sediment for the most basic of all scientific reasons. Life is short and the world is immense; you can't spend your career on a single cliff face. The essence of science is intelligent sampling, not sitting in a single place and trying to get every last one. Under Darwinian gradualism, intelligent sampling followed the usual handful-from-the-haystack method. The results obtained matched the expectations of theory, and conceptual satisfaction (in retrospect, one might say "sloth") set in. No impetus existed for the much more laborious dismember-the-entire-haystack method, a quite unusual approach in science. We could have worked this way, but we didn't because we had no reason to do so. The Alvarez theory made this unusual approach necessary. It forced us to look in a different way. "All observation must be for or against some view if it is to be of any service!"

Consider two premier examples--the best-known marine terrestrial groups to disappear in the Cretaceous-Tertiary extinction: ammonites and dinosaurs. Both had been prominently cited as support for gradual extinction toward the boundary. In each case, the Alvarez hypothesis inspired a closer look via the dismember-the-haystack method, and in each case, this greater scrutiny yielded evidence of persistence to the boundary and potentially catastrophic death.

Ammonites are cephalopods (mollusks classified in the same group as squids and octopuses) with coiled external shells closely resembling those of their nearest living relative, the chambered nautilus. They were a prominent, and often dominant, group of marine predators, and their beautiful fossil shells have always been prized by collectors. They arose in mid-Paleozoic times and had nearly become extinct twice before--in two other mass dyings at the end of the Permian and of the Triassic periods. But a lineage or two had scraped by each time. At the Cretaceous-Tertiary boundary, however, all lineages succumbed, and to cite Wordsworth from another context, there "passed away a glory from the earth."

My friend and colleague Peter Ward, paleontologist from the University of Washington, is one of the world's experts on ammonite extinction; a vigorous, committed man who adores fieldwork and could never be accused of laziness on the outcrop. Peter didn't care much for Alvarez at first, largely because his ammonites seemed to peter out and disappear entirely some thirty feet below the boundary at his favorite site, the cliffs of Zumaya on the Bay of Biscay in Spain. In 1983, Peter wrote an article for Scientific American entitled "The Extinction of the Ammonites." He stated his opposition to the Alvarez theory, then so new and controversial, at least as an explanation for the death of ammonites:

The fossil record suggests, however, that the extinction of the ammonites was a consequence not of this catastrophe but of sweeping changes in the late Cretaceous marine ecosystem. . . . Studies of the fossils from the stratigraphic sections at Zumaya in Spain suggest they became extinct long before the proposed impact of the meteoritic body.

But Peter, as one of the smartest and most honorable men I know, also acknowledged the limits of such "negative evidence." A conclusion based on not finding something has the great virtue of unambiguous potential refutation. Peter wrote: "This evidence is negative and could be overturned by the finding of a single new ammonite specimen."

Without the impact hypothesis, Peter would have had no impetus to search these upper thirty feet of section with any more care. Extinctions were supposed to be gradual, and thirty feet of missing ammonites made perfect sense, so why look any further. But the impact hypothesis, with its clear prediction of ammonite survival right up to the boundary itself, demanded more intense scrutiny of the thirty-foot haystack. In 1986, Peter was still touting sequential disappearance: "Ammonites. . .appear to have become extinct in this basin well before the K/T [Cretaceous-Teritary] boundary, supporting a more gradualistic view of the K/T extinctions" (Palaios, vol. 1, pp. 87-92).

But Peter and his field partners, inspired by Alvarez (if only by a hope of disproving the impact hypothesis), worked on through the haystack:"The remaining part of the Cretaceous section was well exposed and vigorously searched and quarried." Finally, later in 1986, they found a single specimen just three feet below the boundary. It was crushed, and they couldn't tell for certain whether it was an ammonite or a nautiloid, but this specimen did proclaim a need for even more careful search. (Since nautiloids obviously survived the extinction--the chambered nautilus still lives today--such a fossil right at the boundary would occasion no surprise.)

Peter started a much more intense search in 1987, and the ammonites began to turn up--mostly lousy specimens and very rare, but clearly present right up to

397