We propose a critique of normativism, defined as the idea that human thinking reflects a normative system against which it should be measured and judged. We analyze the methodological problems associated with normativism, proposing that it invites the controversial “is-ought” inference, much contested in the philosophical literature. This problem is triggered when there are competing normative accounts (the arbitration problem), as empirical evidence can help arbitrate between descriptive theories, but not between normative systems. Drawing on linguistics as a model, we propose that a clear distinction between normative systems and competence theories is essential, arguing that equating them invites an “is-ought” inference: to wit, supporting normative “ought” theories with empirical “is” evidence. We analyze in detail two research programmes with normativist features – Oaksford and Chater's rational analysis and Stanovich and West's individual differences approach – demonstrating how, in each case, equating norm and competence leads to an is-ought inference. Normativism triggers a host of research biases in the psychology of reasoning and decision making: focusing on untrained participants and novel problems, analyzing psychological processes in terms of their normative correlates, and neglecting philosophically significant paradigms when they do not supply clear standards for normative judgement. For example, in a dual-process framework, normativism can lead to a fallacious “ought-is” inference, in which normative responses are taken as diagnostic of analytic reasoning. We propose that little can be gained from normativism that cannot be achieved by descriptivist computational-level analysis, illustrating our position with Hypothetical Thinking Theory and the theory of the suppositional conditional. We conclude that descriptivism is a viable option, and that theories of higher mental processing would be better off freed from normative considerations.

It is neither desirable nor possible to eliminate normative concerns from the psychology of reasoning. Norms define the most fundamental psychological questions: What are people trying to do, and how? Even if no one system of reasoning can be the norm, pure descriptivism is as undesirable and unobtainable in the psychology of reasoning as elsewhere in science.

Reasoning research has traditionally focused on the derivation of beliefs from beliefs, but it is increasingly turning to reasoning about decisions. In the absence of a single, entrenched normative model, the drive toward normativism is weaker in this new field than in its parent fields. The current balance between normativism and descriptivism is illustrated by three approaches to reasoning about decisions.

Three correctives can get researchers out of the trap of constructing unitary theories of “thinking”: (1) Strong inference methods largely avoid problems associated with universal prescriptive normativism; (2) theories must recognize that significant modularity of cognitive processes is antithetical to general accounts of thinking; and (3) consideration of the domain-specificity of rationality render many of the present article's issues moot.

A critique of inferences from “is” to “ought” plays a central role in Elqayam & Evans' (E&E's) defense of descriptivism. However, the reflective equilibrium strategy described by Goodman and embraced by Rawls, Cohen, and many others poses an important challenge to that critique. Dual-system theories may help respond to that challenge.

Elqayam & Evans (E&E) argue against prescriptive normativism and in favor of descriptivism. I challenge the assumption, implicit in their article, that there is a choice to be made between the two approaches. While descriptivism may be the right approach for some questions, others call for a normativist approach. To illustrate the point, I briefly discuss two questions of the latter sort.

Elqayam & Evans' (E&E's) critique of normativism is related to an inherently philosophical question: Is thinking a normative affair? Should thinking be held accountable towards certain norms? I present the historical and philosophical origins of the view that thinking belongs to the realm of normativity and has a tight connection with logic, stressing the pivotal role of Kant in these developments.

Elqayam & Evans' (E&E's) laudable program to keep the scientific investigation of human reasoning norm-free and focused on the facts alone is an essential part of a long tradition in the philosophy of science – but it faces deeper difficulties than the authors seem to realize, since reasoning is a competence, and the very concept of competence is normative.

Though we are in broad agreement with much of Elqayam & Evans' (E&E's) position, we criticize two aspects of their argument. First, rejecting normativism is unlikely to yield the benefits that E&E seek. Second, their conception of rational norms is overly restrictive and, as a consequence, their arguments at most challenge a relatively restrictive version of normativism.

This commentary questions Elqayam & Evans' (E&E's) claims that thinking tasks are doomed to have multiple normative readings and that only applied research allows normative evaluations. In fact, some tasks have just one undisputed normative reading, and not only pathological gamblers but also normal individuals sometimes need normative guidance. To conclude, normative evaluations are inevitable in the investigation of human thinking.

Normative theories can be useful in developing descriptive theories, as when normative subjective expected utility theory is used to develop descriptive rational choice theory and behavioral game theory. “Ought” questions are also the essence of theories of moral reasoning, a domain of higher mental processing that could not survive without normative considerations.

Normative theories provide essential tools for understanding behaviour, not just for reasoning, judgement, and decision-making, but many other areas of cognition as well; and their utility extends to the development of process theories. Furthermore, the way these tools are used has nothing to do with the is-ought fallacy. There therefore seems no basis for the claim that research would be better off without them.

Elqayam & Evans (E&E) argue that evaluative normativism leads to unacceptable research biases, and should be avoided. Though it is stipulated that the particular biases they discuss are cause for concern, this argument should not be generalized. The boundary between evaluative and goal-directed “directive” norms is difficult to define, and normative assumptions are an integral part of academic progress; moreover, the biases that result may have beneficial potential.

Tests of a universal theory often find significant variability and individual differences between cultures. We propose that descriptivism research should focus more on cultural and individual differences, especially those based on motivational factors. Explaining human thinking by focusing on individual difference factors across cultures could provide a parsimonious paradigm, by uncovering the true causal mechanisms of psychological processes.

We are neither as pessimistic nor as optimistic as Elqayam & Evans (E&E). The consequences of normativism have not been uniformly disastrous, even among the examples they consider. However, normativism won't be going away any time soon and in the literature on causal Bayes nets new debates about normativism are emerging. Finally, we suggest that to concentrate on expert reasoners as an antidote to normativism may limit the contribution of research on thinking to basic psychological science.

Elqayam & Evans (E&E) argue that the major objective of research on human thinking should be the development of descriptive theories, and they challenge normativism – “the belief that people ought to conform to a normative standard” (target article, sect. 1, para. 10). I contend that although their argument for the importance of developing descriptive theories is compelling, normative theories are also important, not only for improving thinking but for investigating and understanding it as well.

Mere facts about how the world is cannot determine how we ought to think or behave. Elqayam & Evans (E&E) argue that this “is-ought fallacy” undercuts the use of rational analysis in explaining how people reason, by ourselves and with others. But this presumed application of the “is-ought” fallacy is itself fallacious. Rational analysis seeks to explain how people do reason, for example in laboratory experiments, not how they ought to reason. Thus, no ought is derived from an is; and rational analysis is unchallenged by E&E's arguments.

Normative theories like probability logic provide roadmaps for psychological investigations. They make theorizing precise. Therefore, normative considerations should not be subtracted from psychological research. I explain why conditional elimination inferences involve at least two norm paradigms; why reporting agreement with rationality norms is informative; why alleged asymmetric relations between formal and psychological theories are symmetric; and I discuss the arbitration problem.

Understanding cognitive processes with a formal framework necessitates some limited, internal prescriptive normativism. This is because it is not possible to endorse the psychological relevance of some axioms in a formal framework, but reject that of others. The empirical challenge then becomes identifying the remit of different formal frameworks, an objective consistent with the descriptivism Elqayam & Evans (E&E) advocate.

Elqayam & Evans (E&E) are focused on the normative judgments used by theorists to characterize subjects' performances (e.g. in terms of logic or probability theory). They ignore the fact, however, that subjects themselves have an independent ability to evaluate their own reasoning performance, and that this ability plays a major role in controlling their first-order reasoning tasks.