Why affect has to be turned into a measurable quantity first
Formal decision theory describes choice as a function of payoffs, probabilities, and the animal's own state. Affect — fear, boldness, hunger-related arousal — is not a payoff or a probability, so it cannot enter the model directly. It has to be replaced by something observable that can be compared across animals and situations. The foraging-herbivore work points to one candidate: personality traits such as boldness and shyness, which are measured from behavior and then used to predict how individuals make decisions alone and in groups [3]. That is not affect in the emotional sense, but it is an operational stand-in — a measurable behavioral disposition that plays the same modeling role.
Economic choice experiments with laboratory animals offer a second template. In that tradition, animals choose between options under uncertainty, and the resulting choice patterns are used to compare economic models against psychological and biological ones [5]. The useful lesson is methodological: if a state like affect is going to matter, it must show up as a shift in choice behavior that a model can predict in advance.
Where affect could enter a decision model without breaking it
One natural entry point is the threshold at which an animal switches strategy. A quantitative model of animal conflict derives specific switch points — for example, a size ratio of about 62% separating exploitation from cooperation by a larger player, and a 38% ratio separating cooperation from cheating by a smaller player — and tests them against body-mass data from co-cultured fishes [2]. Those thresholds are exactly the kind of parameter affect could plausibly move: a frightened animal might switch to cooperation or surrender earlier than a calm one. The model itself does not include affect, but it shows what a testable affective parameter would look like.
A second entry point is the rule an individual follows inside a collective. The herbivore model links individual behavioral rules to group-level decisions, which means an individual's boldness can be traced from its own choices to the outcome for the group [3]. If affect is defined as a shift in that individual rule, its consequences become measurable at both levels rather than remaining a private feeling.
What a comparative test would actually require
A comparative test needs three things: a formal model with a free parameter, an independent behavioral measure that stands in for affect, and a prediction about how the parameter should change when that measure changes. The personality-and-foraging work supplies the middle piece by treating boldness and shyness as measurable individual differences that feed into decision rules [3]. The economic choice work supplies the logic of comparing competing models against the same choice data [5]. Neither supplies the affective interpretation itself.
Evolutionary game theory adds a caution about scope. When behavior is analyzed as strategies in contests, what looks like an emotional reaction can often be redescribed as a payoff-driven strategy, and the formal analysis is designed to capture exactly that [4]. So a comparative test of affect has to show that the affective parameter adds predictive power beyond strategy alone — otherwise the affect label is doing no work.
The definition itself is still missing
Reviews applying dual-process and Bayesian decision theories to wild primates examine how cognitive load and environment shape decisions, but they do not define affect or give it a measurable role [1]. That is the central gap: the modeling machinery exists, and behavioral proxies such as boldness exist, but no study here joins them into an operational definition that other researchers could apply and compare across species. Building one would mean specifying which observable behavior counts as the affective signal, which model parameter it is predicted to shift, and how large a shift would count as support.
Sources used in this answer
How can we apply decision‐making theories to wild animal behavior? Predictions arising from dual process theory and Bayesian decision theory
Reviews dual-process and Bayesian decision theories for wild animal behavior but does not address affect or offer an operational definition for comparative testing.
A quantitative decision theory of animal conflict
Derives quantitative strategy-switch thresholds for animal conflict and tests them against body-mass data from co-cultured fishes, without incorporating affect.
Personality and collective decision-making in foraging herbivores
Models collective foraging decisions in herbivores from individual behavioral rules, linking personality traits such as boldness and shyness to individual and group decisions.
Game theory and the evolution of behaviour
Argues that evolutionary game theory is a valuable formal tool for analyzing animal behavior and contests, without discussing affect.
Economic choice theory: An experimental analysis of animal behavior
Uses laboratory animal choice experiments under uncertainty to compare economic choice models with psychological and biological models.