IF NOT FOR HIS EUGENIC COMMITMENTS, Francis Galton would never have developed the statistical concepts that he did. … Rosalind Franklin would eventually have solved the three-dimensional structure of DNA. … Counterfactuals like these come easily to historians of science. Whenever they attempt to do more than chronicle the scientific past—when they seek to explain its shape, judge its significance, clarify its challenges to understanding—they make claims about what might or might not have happened. Even when these are not explicit, they are, if you know where to look, implicit. (As Steve Fuller says in his contribution to this Focus section, “historians who oppose counterfactuals are not hardheaded about facts but softheaded about causation.”)1

Yet there has been, to date, remarkably little systematic reflection about them. How do they work? What makes for greater or lesser plausibility in conjecturing about alternative scientific pasts? Or must they all suffer equally from, in Steven Shapin's phrase, a “credibility handicap”? Supposing that there are ways of reducing that handicap, can the more credible counterfactuals teach us new things about the scientific past and present? Or are even these doomed merely to serve as colorful restatements of the convictions about science that we start with? Although by no means answering these and related questions in one voice, the contributors to this set of essays all urge the meta-counterfactualist cause.2 Each does so, moreover, from a different perspective: John Henry as a historian of the Scientific Revolution; Peter Bowler as a historian of Darwinism; Steven French as a historically and sociologically engaged philosopher of modern physics; and Fuller as a historically and philosophically engaged sociologist of science. Between them they range over the sciences, the centuries, and the globe. Like the best counterfactual histories, their essays combine rigor and imagination to expose complacencies and inspire their overcoming.

In the recent “science wars,” it was a commonplace that contingency mongering is rife among academic historians of science, who, it was said, seek to undermine scientists' authority by portraying their most cherished ideas about nature and its study as the upshot of social and material contexts that might well have been otherwise.3 Unsurprisingly, the essays that follow have much to say about history of science counterfactuals from the more provocative end of the recent historiography. Henry, for instance, considers suggestions from feminist historians of science about the possibility of a less male-oriented science than the one we inherited and why those suggestions have been little developed; while French summarizes the case for doubting proposals about the contingency of what textbooks teach of quarks and quantum mechanics.

It is well to emphasize at the outset, however, that there is nothing new about history of science counterfactuals. When William Whewell, in the first volume of his History of the Inductive Sciences (1847), considered the history of alchemy and its relationship to chemistry, conventional historiographic wisdom had it that without alchemy chemistry would never have arisen. Far from rejecting the claim as pointlessly speculative, Whewell treated it as demonstrably false:

So, according to Whewell, with no alchemy fogging men's minds and misdirecting their ambition, chemistry would have emerged just as it actually did, only sooner. Note that he did not merely assert that this was so; he appealed to the evidence. Never mind that his evidence—from the history of alchemy/chemistry, the history of astrology/astronomy, and his own observations about what people are like generally when it comes to disinterested truth—is not very impressive. What bears underscoring is that he had some and did not begrudge the marshaling of it to help his readers assess a history of science counterfactual.

Contrary to reputation, counterfactuals do not necessarily inhabit an evidence-free zone. In general terms, we can ask: How should evidence from the actual scientific past and present be used in constructing and evaluating conjectures about alternative scientific pasts? French's essay addresses the problem of evidence most directly. Restricting his attention to theories claimed as could-have-been-contenders, he posits a continuum between theories so rough grained that they cannot be connected up with evidence sufficiently to be judged genuine possibilities or not and theories so finely grained and empirically well supported as not to count as counterfactual anymore. Historians eager to plumb the unrealized potential for success in failed theories, and to do it in an empirically disciplined way, need, according to French's analysis, to draw on whatever documentary clues remain to edge those theories as close as possible to the latter pole. That said, as Fuller reminds us, there are understandings of the historian's duties that make the documents less constraining. Under Lakatosian rules for “rational reconstruction,” what matters is not whether a scientific thinker actually got around to formulating a key idea, but only whether he or she could have done so.5

Instructively, Henry and Bowler take opposite sides on what it means, in testing claims about possible scientific pasts against evidence from the actual past, to take seriously science's social setting. For Henry, counterfactuals that suppose only small-scale changes—the premature death of a discoverer, say—must be intellectually inert, since, according to the anti–great man, social constructionist view as he interprets it, science was bound to develop as it did, one way or another. The mathematization of nature and other legacies of the Scientific Revolution were inevitable, given the wide distribution of the relevant background factors. For Bowler, however, a commitment to social explanation does not require so full blown a denial of the influence of the small. In the case he examines, Charles Darwin turns out to have been the only individual with the combination of (socially explicable) attitudes and interests required to formulate the theory of natural selection; so that, had Darwin, for whatever reason, not produced the theory, it would not have been around to bias discussion in the ways it did. Darwin's theory made a difference to what followed, Bowler suggests, not so much in what was believed but in how it was thought about—what was considered primary and what secondary, which elements were presented to the Zeitgeist and which not.

One aim of these essays, then, is to help clarify the epistemology of history of science counterfactuals: what we can know by the light of the best ones, what makes the best ones so good, and how to tell the best from the rest.6 Another, related aim is to help clarify the associated politics. Returning to the passage from Whewell's History of the Inductive Sciences, we shouldn't fail to notice that in arguing for the inevitability of chemistry Whewell meant to knock alchemy down, as part of a wider indictment of the medieval period. Henry's essay brings into focus the complex politics in our own time of backing inevitability in science. He teases out the way it binds social constructionists and their historiographic nemesis, the positivists—the former supposedly insisting that individual scientists do not matter anywhere near as much as their social contexts do, the latter that nothing disturbs science's onward march to the truth. Bowler, with his more liberal understanding of the constructionist commitment, offers a reading of Darwinian politics that is neither inevitabilist nor positivist. Something like social Darwinism and something like eugenics would have emerged even without the influence of natural selection theory, Bowler suggests; but they would likely have taken less virulent forms. His essay points to the potential of an empirically disciplined and conceptually subtle counterfactualism to ease historians out of stale science-war positioning and onto the path of genuine inquiry, in pursuit of answers that may surprise them as much as the rest of us.7

Nevertheless, and as mentioned above, there is an understandable concern that counterfactual scenarios can only ever reflect back what their authors already believe. French warns that ideas about how open science should ideally be have sometimes led to exaggerated descriptions of how open it has really been. In the tradition of his teacher Heinz Post, and in dissent from any romantic notion of scientific discovery as beyond analysis, French stresses the role of heuristics in science and the need to take these seriously whenever we consider whether a particular investigation or debate might have gone off in other directions than it did. The textbook quark theory, quantum chromodynamics, looks contingently successful only, French argues, if we ignore the established preference among particle physicists for gauge-invariant theories. Once quantum chromodynamics was around, its gauge invariance made its success over its rival, parastatistics, a foregone conclusion. On the side of keeping science open, however, Fuller in his essay here advances what might be termed an activist approach to history of science counterfactuals.8 Let us, he urges, not be content merely to conjecture about how different the scientific present might or might not have been, given this, that, or the other change or changes in the scientific past. Let us in addition bring the scientific past and present into critical dialogue, by letting past science hold present science to account and vice versa—counterfactually bringing the investigators of the past, with their standards, forward in time in order to criticize our science and sending our scientists, standards and all, back in time to criticize the science of their predecessors. Fuller knows, of course, that historians nowadays are trained to quarantine present-day conceptions when approaching the scientific past—indeed, they take pride in that training and the work it produces. He means to hold us to account too.

In building to this bracing vision of counterfactually and critically entangled history, philosophy, and science, Fuller observes that what seems most to bother historians vexed by counterfactuals is that they are not-the-facts—that they are fiction. Undeniably, counterfactual history of science offers scope for historical reasoning at its most imaginative; and that is, for those who find the prospect appealing, part of the appeal. But Fuller helpfully draws attention to how generally counterfactual reasoning features in human cognition. Whenever analytically minded people concern themselves with causation, in science and law no less than in history, they go counterfactual.9 Scientists seem to live comfortably enough with their fictionalizing efforts as aids to factual understanding. They know that the differences between the fictionalizing of nature in a thought experiment, a computer simulation, and a laboratory are matters of degree. But scientists have centuries of methodological reflection behind them. The theory and practice of counterfactual history of science are just beginning. Circumstances permitting (or maybe it is inevitable), historians of science too may one day be able to look back on a richly creative history of efforts to find out more about what never was.10