Document Type


Publication Date


Publication Title

Academy of Management Review



First Page


Last Page



We are grateful to Professor Arend (2020) for his engagement with our work on uncertainty and the choice of strategic logics (Packard & Clark, in press). We easily acknowledge that there are reasons to disagree with our conclusions as they imply a minimization, if not outright rejection, of much of modern behavioral research. It was not surprising, then, to receive Professor Arend’s (2020) criticisms, which appear to be based in what we might call the “epistemic camp” (contra our own “aleatory camp”) of behavioral research.1 The epistemic camp holds all

uncertainty to be what we, in our article, describe as “epistemicuncertainty”andhas elsewhere beencalled “ambiguity” (Packard, Clark, & Klein, 2017)—that is, there is always a probability distribution that can be applied to a decision, even if it is unknown and/or subjectively generated in the mind of the decisionmaker. This assumption is quite seductive, as it allows probabilistic models to be applied to literally any choice situation and lends the appearance of scientific rigor. Professor Arend elaborated this position to conclude that any uncertainty that falls outside of it is essentially chaotic and cannot be managed by anything beyond luck. Our contrary position in the aleatory camp is that this epistemic uncertainty or ambiguity should not be confused with aleatory uncertainty—they are different in nature. Thus, there is no valid way to “convert” aleatory uncertainty into ambiguity by imposing a probability distribution onto something that cannot have one. Most business uncertainty involves such aleatory uncertainty due, in Knight’s (1921: 311) words, to “the inherent, absolute unpredictability of things, out of the sheer brute fact that the results of human activity cannot be anticipated and then only in so far as even a probability calculation in regard to them is impossible and meaningless.” Thus, while the probabilistic approach to decision-making preferred by the epistemic camp has a place, we, like most managers (Harrison, 1977), reject it as unrealistic for the majority of real-world choice scenarios. In the language of set theory, we hold the typical set of options available to an actor to be “open” or, more precisely, “infinite” (Packard et al., 2017), and not “closed,” as is required by probability theory. This renders the probability-based logic employed in the epistemic camp’s behavioral research, and in Professor Arend’s critique, “impossible and meaningless.”