Sanja Sreckovic
Affiliation: Ruhr-University Bochum
Category: Philosophy
Keywords: deliberation, reflection, metacognition, nonhuman animals, affects, values, representation, preference, chimpanzees, orangutans
Date: Thursday 4th of September
Time: 18:30
Location: Room 232 (232)
View the full session: Embodied Cognition
There seems to be a widespread, but vague and implicit connection between deliberation and concepts such as ‘voluntary’, ‘controlled’, ‘effortful’, ‘reflective’, ‘aware’, ‘conscious’, ‘attentive’ or ‘intentional’. These concepts are often grouped together as hallmarks of human rational reasoning (Kahneman, 2012; Broome, 2013; Boghossian, 2014; Valaris, 2017; McHugh and Way, 2018). Nonhuman animals are in this view not capable of deliberation, and are only limited to unconscious automatic processes. Buckner (2019) challenges this picture by proposing a deflationary account of deliberation, which involves minimal executive control or cognitive effort, but not necessarily consciousness and reflection. Deliberation, in this minimal sense, can be attributed to nonhuman animals in cases such as uncertainty monitoring (opting out of difficult tasks, in a way that suggests monitoring their own uncertainty, see Beran, 2012; Foote & Crystal, 2007) or inhibition (the Stroop-like tasks, which indicate effortful inhibition of the default response; see Beran, Washburn, & Rumbaugh, 2007). In these cases, decisions are guided or inhibited by how a situation is subjectively represented or ‘taken to be’ - how it seems to the subject. Even if the particular features of the situation that influenced the decision-making may not be directly available for reflection, these cases still show evidence of at least minimal deliberation based on a subjective take in animals (Buckner, 2019). Taken together, the traditional and the deflationary views paint a picture where humans are capable of full-blown deliberation, while nonhuman animals are capable of (at least) a minimal version of it. Here I extend Buckner’s argument, and argue that nonhuman animals are capable of ‘beyond minimal’ deliberation. More specifically, I argue for a nonhuman ability for deliberation that involves reflection and explicit representing of the options. However, while the traditional understanding of human deliberation is cashed out in terms of abstract, intellectual, language-like entities and processes, I show that the ‘meta’ processes involved in deliberation can also be cashed out in affective terms. My argument is based on the evidence from Osvath & Osvath (2008). In this study, chimpanzees and orangutans were forced to choose between an immediate reward (a grape) and a tool (a straw) that enables future access to a larger reward (a bottle of fruit soup only reachable through a straw). Critically, the larger reward was absent and uncued. The experiment was designed to test the capacity of apes to override the drive toward the immediate reward in favor of future needs. While the analysis and discussion of this case focus on planning and future-oriented cognition (Boyle & Brown, 2024), I want to highlight another aspect of the results not addressed in the literature. The choice between two options was based on their values. The apes’ decision involved weighing a directly perceived value (grape) against a decoupled, represented value (fruit soup). Such motivational trade-offs, akin to those relevant for sentience attributions, require integrating information into an evaluative ‘common currency’ (Birch et al., 2021). Deciding between two options, however, where one option is perceived (and directly tempting) during the choice, while the other is absent and only represented, I argue, suggests that the decision-making process involves explicitly representing values (of the options) and deliberating over them. The decoupled value (fruit soup) must be actively represented and compared against an immediately tempting stimulus. This suggests that choosing the decoupled value implies a process akin to the thought ‘value_2>value_1’, where value_2 is not perceptually available. I argue that, first, this case indicates relying on explicit representation, reflection and deliberation, and, second, that these processes are affective in nature. While most observed motivational trade-offs involve directly perceived values, this study shows deliberation in nonhuman animals over decoupled values. The crucial aspect of the Osvath & Osvath study is that it merges metacognition (commonly understood as a ‘higher order’ intellectual capacity) with affective evaluation. Comparing the values of the options need not be intellectual, symbolic computation but could operate through a ‘felt preference’ - a parsimonious explanation for how values are monitored and compared. Consequently, I argue that nonhuman animals are capable of deliberation that involves reflection, and that higher or meta-levels of monitoring can also be affective in nature. I call this process ‘affective metacognition.’ How does this relate to the research on sentience in other species? Although showing a capacity for deliberation might be an argument for consciousness or sentience, I am not interested in the distribution question, i.e., questioning which animals are conscious or sentient (Allen, 2000; Andrews, 2024). Due to their evolutionary proximity to us, great apes are the least problematic candidates for attributing sentience, even for the biggest skeptics of nonhuman sentience. Instead, I adopt the approach suggested by Andrews (2024) to accept the hypothesis that all animals are conscious and question how they are conscious or sentient, or in general what their sentient minds are like. Since consciousness is conceptually tangled up with deliberation, and the leading views of deliberation are overly intellectualized, not recognizing the crucial role of affective processes involved in deliberation also reflects on how we approach researching consciousness in other species. Through this analysis, I want to open the space for reconceptualizing deliberation and metacognition as possible to be implemented by affective processes. This should open up new avenues for understanding consciousness and sentience in other species.