The role of attentional biases in addictive decision-making 

The role of attentional biases in addictive decision-making 
Federico Burdman 

Addiction is of interest to philosophers, among other things, because it raises a puzzle about rationality in practical agency. A common way to lay out the puzzle takes its cue from the standard definition of addiction as persistent drug use despite harmful consequences. Some people with addiction experience insight deficits and are unaware that they have an addiction or of its serious negative consequences (for a review of the evidence, see Raftery et al. (2020)). However, many people with addiction have a relatively clear understanding of their condition and are aware of its harmful effects. Add that rational agents are often pictured as, by and large, doing things because they see them under the guise of the good. Therein lies the puzzle, on a first approach: Why would someone persistently choose to do something with an awareness that it leads to consequences that they would rather avoid? (Pickard, 2018) 

The standard view in the scientific literature is that addictive behavior is the product of compulsion rather than choice (Everitt & Robbins, 2016; Hyman, 2007; Lubman et al., 2004; Volkow et al., 2016). The picture of compulsive drug-seeking and use is sometimes glossed over by the notion of a loss of control (e.g., Lyvers, 2000). This suggests an easy solution to the puzzle: Addictive behavior is persistently unresponsive to negative consequences because it is out of the agent’s control, so it is not influenced by how the agent sees things. There is, of course, something to be said for this picture. A well-known fact about addiction is that quitting active regular use can be exceedingly difficult even for people sincerely committed to doing so (Dennis et al., 2005), which suggests some impairment of the ability to stop. However, any viable account of compulsion or impaired control must reckon with another crucial fact about addiction, namely that the bulk of instances in which a person with addiction uses drugs appear to be aptly characterized as intentional actions.  Habitual and automatic (uncontrolled) actions account for some instances of addictive behavior (Tiffany, 1990). However, it is far more common for people with addiction to use drugs by deciding to do so, fully aware of their actions. This blocks the easy solution to the puzzle of addiction. In other words, solving the puzzle requires viewing addiction as co-opting rather than annulling the mechanisms that support intentional agency. 

My proposed account is an attentional capture model of addictive decision-making. The model explains reduced responsiveness to reasons in addictive decision-making as the result of relatively tenuous but persistent attentional biases. It builds on the notion that an analogue of stimulus-driven attention exists for internal attention: cognitive, non-perceptual representations whose features or contents are particularly prone to capture attention (Chun & Johnson, 2011; van Ede et al., 2020). The core claim of the model is that when a person with addiction deliberates about what to do, considerations that favor drug use tend to capture her attention: they come to mind more easily and are preferentially elaborated once evoked. These attentional biases towards considerations favoring drug use make it harder for considerations against drug use to be kept in mind. The tendency for use-favoring considerations to win out in the retrieval competition and be sustained in working memory explains why particular instances of deliberation about whether to use drugs are often less responsive to reasons against using than expected: considerations that are not attended in the process of weighing pros and cons become deliberatively inert.  

The model is not intended as a complete theory of addiction—addiction is too heterogeneous for any single theory to capture its essence (Flanagan, 2025). While not incompatible with a role for direct motivational factors in explaining addictive behavior, the attentional capture model contributes to the relatively undertheorized understanding of how addiction influences the cognitive processes involved in making choices. And it sheds light on key aspects of the puzzle of addiction. It portrays addictive drug use as the product of choice, even of deliberative decision-making, and as responsive to positive reasons for using. At the same time, the model accounts for the fact that people with addiction often choose to use drugs despite being in some sense aware of sufficient reasons to do otherwise. Attention allocation plays a crucial role in guiding agents’ behavior at particular decisional junctures (Sripada, 2025), including the decision about which option is best in the context of practical deliberation. An agent’s considered judgment about the reason-giving force of considerations against using may fail to influence deliberation and decision-making if those considerations remain in the background at the critical point in time. Preferentially attending to certain considerations guides the process of weighing the pros and cons of drug-using options in a direction consistent with the said biases. The model thus explains reduced responsiveness to reasons against using without positing a generalized deficit in reasons-responsiveness and also avoiding the need to posit drastic short-term shifts in agents’ evaluative judgments.  

Importantly, the model is motivated by empirical findings. There is ample evidence that perceptually available drug-related stimuli can strongly drive the allocation of perceptual attention in people with addiction (Cox et al., 2015, 2016; Spada et al., 2015). There is also an extensive body of research on involuntary attentional capture effects in internal attention (van Ede et al., 2020; Olivers et al., 2006; Sayette et al., 2010; May et al., 2015; Klein, 2007). Preferential elaboration of drug-related items in working memory in people with addiction, another key assumption of the model, is also empirically supported (Caselli & Spada, 2010; Spada et al., 2015; May et al., 2010). The model is also consistent with anecdotal evidence from first-person narratives of addiction. Although these distinct lines of evidence are indirect, together they constitute a strong case for attentional capture effects in the context of practical deliberation in addiction.