Peter Katzenstein’s edited book, Uncertainty and Its Discontents, seeks to expose and explore the worldviews of both IR scholars and their subjects. Katzenstein argues that most IR approaches cannot analyze uncertainty because they are embedded in a Newtonian-Humanist worldview that assumes a world of human-controllable risks. He calls for waking IR from its “Newtonian slumber” (p. 339) to grapple with the unexpected events and planetary crises that this worldview is unable to grasp.

For Katzenstein, worldviews are “unexamined, pre-theoretical foundations of the approaches with which we understand and navigate the world” (p. i) that “offer global overviews evident in relatively constant, repetitive habits of beliefs and emotions that mediate the relations between an individual or group and the world” and “create narratives about what is possible, what is worth doing, and what needs to be done, as well as what is impossible, what is shameful, and what needs to be avoided” (p. 9). They are thus much more than traditional IR paradigms, which for the most part are trapped in the same Newtonian-Humanist iron cage of reason. Under Newtonian approaches, innovation is limited to remixing existing elements to respond to calculable risks. By contrast, in Post-Newtonian approaches, there is room for protean power: improvisation as a response to uncertainty. Humanism, similarly, is limited by endowing agency only to people, placing them on a pedestal as the anthropic center of the world. Hyper-Humanism, by contrast, treats everything as a potential agent.

On this account, the Newtonian worldview embeds substantialism, which “takes pregiven entities as the starting point and imbues them with properties and agency” (p. 19). Importantly, logics of appropriateness, as well as consequences, are substantialist because they differ on causal mechanisms but ultimately are grounded in individualism. By contrast, the Post-Newtonian worldview incorporates relationalism, in which relations are ontologically prior to entities and agency operates through processes, boundary setting, or other ways of engaging with the world.

The book proceeds in four parts: examples from each of the possible combinations of Newtonian/Post-Newtonian and Humanism/Hyper-Humanism (Mark Haas and Henry Nau, Milja Kurki, Jairus Grove, and Michael Barnett); defenses of Newtonian-Humanism (Henry Nau) and Post-Newtonian Hyper-Humanism (Prasenjit Duara); meditations on two specific types of worldviews (Bentley Allan on science, Timothy Byrnes on religion); and a conclusion by Katzenstein navigating a path between Nau’s Scylla of totalitarianism-enabling relationalism and Duara’s Charybdis of planet-destroying individualism.

Nau’s full-throated defense of Newtonian-Humanism warns against a relativist relationalism that cannot defend against the Scylla of potential totalitarianisms. He assumes a particularly extreme version of relationalism in which choice and freedom are eliminated. For example, in this book Grove demotes President Kennedy from a quarterback to a mere mascot, arguing that ExComm was the decision-making collectivity during the Cuban Missile Crisis. This erases Kennedy’s innovative recombining of existing relations to first create and then orchestrate the deliberations of ExComm: the president was a skilled conductor operating under conditions of extreme uncertainty. Here I side with Nau: if there is nuclear use by the United States, it is still the moral responsibility of the US president, however constrained that choice may be by the historical, social, and physical structures that Grove enumerates.

But as Kurki points out, there are many relationalisms; the one Nau objects to is far from the only option. For example, Barnett studies how relational structures and historical events constrain and enable the agency of his subjects to choose between different Jewish worldviews. Nau’s critique treats relationalism as if it operates at the holistic level of analysis. But relationalism largely operates at a relational level of analysis between individualism and holism. Treating relations and processes as ontological primitives does not result in a uniform holistic structure; instead, it creates variegated sets of structures wherein agency lies in the creative rewiring of relations and manipulation of processes to produce innovative solutions.

Duara’s warning, like Nau’s, is concerned with catastrophic outcomes: without a relationalist approach, we will be unable to escape the Charybdis of an individualistic “runaway global technosphere with cascading consequences” (p. 207) fueling global catastrophes. Both warnings bring out the ethical and moral values of different worldviews, with Nau’s Newtonian-Humanism arguing strongly for Enlightenment notions of universal rationality and individual responsibility as a bulwark against the potential evils of an extreme relativism that cannot combat malignant worldviews, whereas a Post-Newtonian-Hyper-Humanism instead locates the evils of the world in “efficiency-driven, resource-exploiting, nature-controlling, and competing nation-states” (p. 287).

Regardless of the path sailed, thinking about worldviews allows for a conceptual framework that can tackle both science (Allan) and religion (Byrnes), as well as the ways in which they intertwine historically and contemporarily. At the level of worldviews, these are simply two different sets of beliefs about how the world works. Yet neither can be satisfactorily understood within the Newtonian worldview.

Consequently, in the conclusion Katzenstein takes a stance against Haas’s and Nau’s suggestion that Newtonian and Post-Newtonian worldviews are rivals, arguing for two possible ways of having them coexist: complexity and subjective probability. The latter includes both classical and quantum Bayesian approaches, both of which are still tied to an individualist ontology and so, in my view, steer too close to Charybdis.

By contrast, complexity approaches, which focus on the emergent properties of systems from relations, reject individualism while still avoiding the Scylla of extreme relationalism. By eliminating the Newtonian closed-system assumption, complexity approaches can analyze “adaptive characteristics of open systems, their emergent properties, and their uncertainties” (p. 308). This move incorporates insights from quantum approaches without importing some of the more mind-boggling ones that lack social analogies (e.g., nonlocality). Importantly, it acknowledges that some phenomena cannot be predicted or explained in advance but may be understood after they occur. These include emergent properties of human behavior, changes in social processes that foil previous models, unanticipated interactions in a complex system, and unknown unknowns.

Even better, complexity provides an answer to the conditions under which a Newtonian worldview can be used: “The determinist or probability-inflected Newtonian world can be thought of as a special case that reveals itself when the quantum world of infinite possibilities and radical uncertainty collapses” (p. 308). It is difficult, however, to set out conditions for when collapse occurs, because, unlike the physical world, the social world does not have a convenient micro–macro transition. Pragmatically, because all models are wrong (but some are useful), I would argue that a Newtonian perspective can be useful for analyzing relatively closed, rational systems in which the controllable risk assumption is plausible; it identifies why certain equilibria exist and persist or even explains endogenous evolution of a system through simple recombinations of existing strategies or scripts.

It is thus unnecessary to resort to extreme relationalism to reap the benefits of adopting a Post-Newtonian social-scientific worldview. Similarly, to consider agents other than humans, it is not necessary to grant things other than humans completely equal standing. But escaping the Newtonian-Humanist straitjacket is necessary to understand systemic changes and to cope with the uncertainty and threat posed by global catastrophic risks. For only through creativity, innovation, and perhaps some luck can we escape the Charybdis that we have created through our runaway global technosphere.