Adrian Currie’s monograph Rock, Bone, and Ruin: An Optimist’s Guide to the Historical Sciences defends the claim that we should be optimistic about the ability of historical scientists to access the past. Here readers will find a rich and insightful discussion of historical scientific research and the variety of techniques used in this research, a discussion that touches on important philosophical issues in science more broadly. As such, Rock, Bone, and Ruin is most valuable to those interested in the epistemic challenges that face historical scientists, the methods historical scientists use to overcome these challenges, or scientific methodology in general. Philosophers engaged in discussions of realism, idealization, or value pluralism in the sciences will also find significant portions of the book that they can sink their teeth into.

Currie argues that we should be optimistic even when historical scientists find themselves in unlucky circumstances, contexts where trace evidence is limited and midrange theories are ambiguous and uncertain. Currie defends this optimism by undermining a common brand of pessimism about historical science, the pessimistic thesis that much of what we would like to know about the past will remain unknown. According to the pessimist, although we can know some things about deep time, many of the traces needed for discovery have all been lost or destroyed by the ravages of time. In challenging this pessimistic thesis, Currie highlights the features and methods of historical science that allow for epistemic progress even in such circumstances.

The book is divided roughly into three parts: an elaboration of “unlucky circumstances,” an argument against pessimism, and an articulation of optimism and its implications.

The first portion of the book grounds an account of unlucky circumstances in two case studies: the gigantism of sauropods like Brontosaurus and glaciation in the Neoproterozoic tropics (i.e., a “Snowball Earth”). For the former, paleontologists are currently unable to account for the size of sauropods with existing models of maximal terrestrial animal size. Regarding the latter, there is evidence that the Earth experienced global or near-global glacial conditions during the Neoproterozoic. However, simulation models cannot reproduce the conditions required for Snowball Earth to melt without stipulating implausible amounts of greenhouse gases. These cases thus provide a kind of challenge in the present epistemic context because they are unlucky with respect to their present traces. Following from Currie’s ripple model of trace-based evidence (chap. 5), the traces required to resolve these two challenges are of low dispersal, gappy, and faint. These events are expected to have few homogenous causal descendants (low dispersal) and incomplete potential trace sets (gappy). Locating and linking any traces to the past event is also expected to be difficult (faint). In sum, Currie argues that the relevant traces of both sauropod soft tissue and the character of Neoproterozoic glaciation are expected to be gappy and faint while having low dispersal.

Unlucky circumstances motivate a pessimistic argument for historical science. Namely, if (1) our available evidence is limited to past traces, and (2) historical scientists cannot generate novel evidence, then we should be pessimistic about our ability to know about the past in situations where (3) we are unlikely to uncover more pertinent traces (i.e., in unlucky circumstances).

The next portion of the book works to undermine this argument for pessimism. Currie challenges each of the three claims that would, together, establish pessimism about historical science in unlucky circumstances.

First, Currie argues that evidence available to historical science is not limited to traces. In lucky circumstances, historical scientists can take advantage of common-cause explanations and smoking-gun traces (Cleland Reference Cleland2011) or the unification of traces (Tucker Reference Tucker2004). However, these methods are not so useful in unlucky circumstances, where traces are few and far between. Currie argues that historical scientists can generate evidential smoking guns that are not traces themselves. Indeed, there is no essential historical method: historical scientists are methodological omnivores. Historical scientists take advantage of analogies (chaps. 7 and 8) and surrogate experiments performed with models and simulations (chap. 9). As such, historical scientists are not limited to traces.

According to Currie, the second pessimistic claim is also false: historical scientists can generate evidence. Appropriately constructed models and simulations can generate smoking guns that are not based on traces found through fieldwork. Such smoking guns can discriminate between competing hypotheses. Currie discusses how scientists studying the development of Paleozoic echinoderms (e.g., sea urchins and sand dollars) use simulations to discriminate among competing hypotheses (Zachos and Sprinkle Reference Zachos, Sprinkle and Ashraf2011). A simulation with plausible yet simple assumptions that replicates the intricate pattern of Paleozoic echinoderms could be the deciding evidence in favor of that hypothesis. As such, historical scientists can generate evidence, even evidence that functions as a smoking gun.

Thirdly, even in unlucky circumstances, we shouldn’t be pessimistic about the future discovery of traces. As midrange theories continue to develop, novel epistemic avenues will open up. Scientists continue to discover and refine novel techniques for getting at interesting features of the past. I take the color of the dinosaurs to provide an illustrative example of when such novel techniques develop. Derek Turner (Reference Turner2007) argued that scientists would never come to know the color of the dinosaurs, a clear instance of something forever inaccessible to historical scientists. Philosophers like Cleland (Reference Cleland2011) note that not long afterward, paleontologists discovered preserved melanin granules in a fossilized, color-banded bird feather from the Cretaceous. The discovery that many dinosaurs were feathered suggests that a sufficiently preserved dinosaur feather might provide similarly preserved melanin, providing ready access to the color of such dinosaurs. By 2010, scientists were performing melanosome comparisons to identify the color of dinosaurs. Turner (Reference Turner2016), reflecting on the surprising scientific discovery, is right to maintain that pre-2008 background theory did provide some reason for thinking that novel evidence regarding the colors of the dinosaurs would not be forthcoming. Furthermore, Turner is right that a small number of successful discoveries is not sufficient to motivate a more general optimism. Still, the case of dinosaur coloration shows precisely how novel epistemic avenues can become available even when the present background theory suggests that it is unlikely. One can never rule out the possibility that novel methods will emerge for reading existing traces.

The investigation of the color and the dinosaurs serves to highlight another of Currie’s claims: the investigation of the past is scaffolded. It is often impossible to know, before performing empirical investigations, whether available evidence will at some point discriminate between competing hypotheses. We need to know some things about the past before we can reliably assess our ability to know other things about the past. It is otherwise easy to underestimate how much evidence will accrue over time for a given hypothesis, leading to an unwarranted bias toward pessimism. It is easy to presume that we will never discover a preserved dinosaur feather or to presume that preserved dinosaur feathers will lack traces of color that can ever be analyzed. Once we investigate and happen upon preserved dinosaur feathers, we can then look to the traces that are preserved and how we might read those traces. Thus, even though unlucky circumstances provide the worst-case scenario for finding traces, investigative scaffolding and the development of midrange theories provide avenues for the discovery of novel trace evidence.

In sum, Currie defends the claims that historical scientists are not limited to trace evidence, that historical scientists can use models and simulations to generate evidence, and that the development of midrange theory in historical science is likely to generate novel methods for reading traces. This defense successfully deflates the pessimist’s argument.

The final section of the book considers the sorts of optimism that follow and the implications such optimisms have for our understanding of scientific progress, scientific realism, and the promotion of historical science. Currie considers six different forms of optimism and their relative plausibility, distinguishing overly ambitious forms (I’d be willing to bet that we will learn every past good) from those that will depend on the particular epistemic context (For any important past good, I’d be willing to bet that historical science will learn that good—unless it has particular features that make its recovery unlikely) and from more modest forms (Historical scientists will generally increase their pool of knowledge about the past, discovering many important facts). There is something disappointing about leaving the discussion of optimism with a plurality of optimisms (like Socrates to Meno, “While I am looking for one virtue, I have found you to have a whole swarm of them!”). The more precise forms of optimism are contextually specific, leaving only the most general kind of optimism for historical science in general. However, this is not an objection to Currie. It is implausible that anything much more specific could be defended across all historical scientific contexts. As Currie notes, most of the epistemic action is local, a function of the specific situation in which the scientist finds herself. We should expect the most defensible optimistic thesis about general historical science to be the modest thesis: historical scientists will generally increase their pool of knowledge about the past, discovering many important facts.

Currie concludes the book with an argument that 20th-century philosophy of science would have developed very differently had 21st-century historical science provided the case studies. This fantastic counterfactual highlights how 20th-century debates about scientific realism and scientific progress would have developed very differently if guided by the insights of contemporary historical science. Historical scientists are pluralistic in aim and opportunistic in method (methodological omnivores), and progress occurs when exploratory hypotheses are pursued with methods that are useful for the specific, local epistemic context. Philosophers of the 20th century would not have thought science to be the gradual, linear march toward truth (or some analogue of truth) if there had been less emphasis on developing scientific theories as large-scale structures. Furthermore, plurality in scientific method and aim might have been recognized earlier had contemporary historical science been able to provide the paradigm for 20th-century philosophers of science. Indeed, the analysis of historical science throughout the book seems to highlight that historical science exhibits many features that 21st-century philosophers of science are recognizing across the sciences more broadly: value pluralism and methodological diversity across contexts.

In Rock, Bone, and Ruin, Currie highlights the ingenuity of historical scientists in a clear and engaging analysis of the puzzles they face, the techniques they use, and the discoveries they make. By the end of the book readers are left with a modest optimism about historical science in general and with a framework for understanding in which contexts we can be more specifically optimistic. In developing this framework, Currie makes way for future philosophical and scientific research to apply the framework in the exploration and assessment of local scientific contexts.