On an unseasonably warm April day in Chicago 2003, at the annual meeting of the American Educational Research Association (AERA), psychologist Russ Whitehurst stood up to deliver a talk entitled “New Wine, New Bottles.” Whitehurst was the director of the newly created Institute of Education Sciences (IES), and he set out to describe the reasons that the Education Sciences Reform Act of 2002 and IES—which the act created—were necessary, and what he hoped these would do for educational research in general.

Citing long-standing complaints about education research's proclivity for theory development and advocacy, Whitehurst argued that research in the field ought to focus instead on workable strategies that are “relevant to practitioners and policy makers.”Footnote ¹ He outlined four areas of emphasis: effective instructional practices, standards and assessments, education finance, and closing achievement gaps. And he specified that the randomized controlled trial provided the best evidence on which practitioners can rely, though quasi-experimental and correlational designs were also fine, and qualitative methods had a supporting role to play in certain situations.Footnote ² This research agenda represented the “new wine” in his title; the federal funding agency represented his “new bottles.” The basic project Whitehurst outlined in 2003 continues to drive the agenda and methods of influential educational research to this day.

It was difficult for me not to think of Whitehurst's talk—the denigration of theoretical and advocacy work, the emphasis on creating knowledge for practical use, and the veneration of the randomized trial—as I devoured Solovey's magnificent Social Science for What? Although Solovey focuses on the mission and the tribulations of the social sciences at the National Science Foundation (NSF) in particular, the titular battles that he chronicles resonate widely, largely because the historical contexts that shaped the NSF—the Cold War, the social movements of the 1960s, and the rise of Reagan conservatism—obviously shaped other organizations and fields as well. Education specialists will find Solovey's book quite relevant to their own interests. As I did, they might also find it meticulously and exhaustively researched without, somehow, reading ponderously—a neat trick!

The book's exhaustiveness owes primarily to its scope. In order to give readers a sense of the battles and the seesawing political demands that caused the social sciences’ star to rise and fall cyclically at the NSF, Solovey cannot afford to leave anything out. He begins with Vannevar Bush's 1945 Science—the Endless Frontier, the report that proposed something like the NSF, and he carries the narrative all the way into the Trump administration. As he does so, he sets himself the unenviable task of accounting for several synchronic phenomena simultaneously: the NSF's internal organization and mission, the policy-making environment surrounding the NSF, the evolving place and prestige of the social sciences within the agency, and the agency's direct and indirect influences on the practice of social science beyond its walls. In light of this ambitious and complex undertaking, the book's readability stands out all the more starkly as a major accomplishment.

Amid this richly layered historical narrative, two clear through lines emerge. The first is a simmering and unresolved struggle over whether the social sciences are really scientific in the mold of the natural sciences. In this respect, the story of the social sciences at the NSF is indistinguishable from the global story of the social sciences in the latter half of the twentieth century as a whole. Assumed to be less “mature” than the natural sciences, the rigor of social science methods and the quality of their theoretical questions were constant sources of anxiety for both social scientists and their patrons. Solovey terms the view that the social sciences are indeed participating in the same project as the natural sciences “scientism,” which he notes has a “specific sense” in his usage. On this view, there is a “unified scientific enterprise” of which the social and natural sciences both partake (pp. 12–14). Over and over again in the twentieth century, regardless of the personal views of NSF leaders and social scientists themselves, the agency finds itself compelled to justify its funding and its projects by appealing to scientism in Solovey's sense.

To pick one nit with Solovey's book, here, the distinction between his “specific sense” of scientism and the more common sense seems to obscure rather than to clarify the internal relation between them. It can come to look like all Solovey is discussing under the flag of scientism is this theoretical belief in the unity of the sciences, thus unhooking that belief from its practical ramifications—which is where the more common understanding of scientism becomes necessary. This more common sense of the term holds that science is the sole source of trustworthy knowledge (p. 1). The very reason that the NSF must constantly publicly appeal to scientism in Solovey's sense is therefore rooted in scientism in the ordinary sense. Even though these two senses really are different from one another, as their definitions are sufficient to show, they rarely appear in isolation in the real world of policy and practice. The use of each tends to imply the other. Given the amount of time Solovey devotes to the entanglement between social science practices and the politics of social science funding and oversight, I'm not sure it makes sense to put the more ordinary understanding out of play. Is it less clunky to regularly have to say “a unity-of-sciences view” than it is to regularly have to talk about a “long-standing commitment to first-rate scientific inquiry and scholarship at the hard-core end of the social research continuum” (p. 273), which is more or less what the ordinary sense of scientism involves?

The second through line wending through the book derives from the anxiety over the status of the social sciences vis-à-vis the natural sciences to which scientism in both senses responds. The ambiguous value of the social sciences for producing trustworthy knowledge about the world that can guide public action exposes the social sciences to political attacks, both for their perceived weakness in this regard and for their potential power. Solovey notes throughout the book that NSF funding for social science programs consistently pales in comparison to funding for the natural sciences, which reflects the relatively low opinion that public representatives hold of their trustworthiness and utility. At the same time, however, public funding for social sciences is also under constant threat because of those representatives’ fears of its effectiveness, its potential, lying right there in the name itself, to bring about “socialism” or engage in “social engineering.” In this lose-lose setup, social science is unworthy of public expenditure not only because of what it cannot do but also because of what it can. Politicians of all stripes can at least agree on this: we shouldn't fund the social sciences. From the point of view of an education scholar, this is doubtless the most depressing aspect of the story.

Solovey's account of the NSF's role in the Man: A Course of Study (MACOS) affair—which he covers over the course of two central chapters—will be the most familiar to educational historians and also the most revealing of the book's main arguments. The high hopes for MACOS, representing a turn toward applied science at the NSF, were rooted in the social sciences’ potential for describing social reality rigorously and faithfully. And the conservative backlash against MACOS, based on the idea that social science methods allow ideological indoctrination to masquerade as value-neutral education, reveals the fear that they do not describe reality rigorously and faithfully at all.

The MACOS debacle has weighed heavily on every subsequent proposal to federally support educational research. Think again of Whitehurst's 2003 address at AERA: it plays all the hits. Educational research should methodologically approximate the natural sciences to the greatest possible extent; doing otherwise risks undermining the unity of the sciences and exposing our findings to claims of ideological bias. Educational research should also defer to the practical needs of school leaders and policymakers when it comes to deciding which questions to pursue. In light of Solovey's work, we can see that this move offloads the risk involved in formulating a research agenda, and that Whitehurst's shortlist of addressable needs represents only the most nonthreatening, instrumental avenues researchers might pursue. In order to win monetary support, then, from IES and many private funders alike, educational researchers have to take pains to assuage the same kind of anxieties that have dogged the NSF's social science programs since the agency's founding. They have to claim to be no different from the natural sciences in either their ability to represent reality or their ideological neutrality. Solovey's excellent book reminds us that educational research is not alone in this. For social scientists, this has always, and literally, been the price of the ticket.