Neil Levy on Addiction

In a fascinating paper presented at the St Cross Ethics Seminar in Oxford, on 27 March 2014, Professor Neil Levy (Oxford and Melbourne) sought to solve the following puzzle about addicts: on the one hand, addicts are thought to lack control, but on the other they appear to engage in the kind of reason-responsive behaviour typical of rational agents (for example, many addicts for a small financial incentive will avoid the objects to which they are addicted).

Levy’s central claim was that addicts do lack control, but that this lack of control consists in a lack of control over belief-formation, leading to a change of mind – or ‘judgement-shift’. So addicts are rational in so far as they are acting on the basis of their current beliefs about what is best for them. 

In elucidating the mechanism of addiction, Levy used the notion of the brain as a ‘predictive mechanism’ which seeks to explain input at the sub-personal level. Any failure to explain is then referred upwards, until eventually it reaches the personal or conscious level. If the error is sufficiently large, the agent’s mental system may then shift to another causal model, which may well be in an unstable state (so addicts’ minds may change back again, for example, after they’ve indulged). 

Levy also explained the role of the mesolimbic dopamine system in addiction. This system establishes a link in our minds between pleasurable objects and signals of their availability. So the system will become active, for example, when the addict sees a needle, or her drug-dealer. According to Levy, the system is concerned not with reward-prediction, but with error-prediction, and in addicts it is highly dysfunctional because addictive drugs hijack it, leading to beliefs about the value of drugs which are in effect delusional. The addict in a sense ‘forgets’ her earlier beliefs about the bad effects of drugs on her health and family life, and so on. Things are exacerbated by effects similar to those found in ‘ego-depletion’, which also involves a form of judgement-shift. This has the result that once one has taken the drug a few times, one is more likely to take it next time the opportunity arises. On this model, then, temptation is a prediction-error, ego-depletion weakens ‘top-down’ or personal influence on behaviour, and this leads to a judgement shift in valuation of the drug. The addict changes her mind, but not in response to new information, and it is in this that the loss of control primarily consists. 

As Levy accepts, his model is heavily Socratic (or, if you prefer, Platonic or Aristotelian). The addict is not doing what she straightforwardly believes to be worse for her overall. One obvious problem for the Socratic model of irrational behaviour is that people acting irrationally – from weakness of will, for example – will often say, as they are acting, ‘I know that this is bad for me’. So how can there be a judgement-shift in such cases? 

Here, Levy pointed out how bad we often are at introspection, and in particular at working out what we believe. Self-deception is as common a phenomenon as weakness of will. I am inclined to agree that the Socratic model is not as obviously mistaken as philosophers often assume. Consider someone who smokes (an addict?), and who is often heard to say that she knows smoking is overall bad for her. Now imagine that this person is taken on a tour of the bronchial ward at her local hospital. It seems to me not entirely unlikely that, after that visit, she might say: ‘I thought I knew what it was for smoking to be bad for me. But I didn’t. I didn’t fully understand what I was really doing’. Likewise, it would not seem to me clearly absurd to suggest that an addict, whose relationship with her young child is being destroyed by her addiction, should, when faced with temptation, watch a movie of that child on her mobile ‘phone. I am not saying this strategy will work (though I’d be interested to know if it would). But the fact that it’s not absurd itself suggests that the judgement-shift model of addiction isn’t absurd either.

  • Facebook
  • Twitter
  • Reddit

2 Responses to Neil Levy on Addiction

  • Harald K says:

    “So addicts are rational in so far as they are acting on the basis of their current beliefs about what is best for them. ”

    Are you familiar with Ole-Jørgen Skog’s 199 paper “Hyperbolic discounting, willpower, and addiction”? It has a lot of interesting reflection on time preferences in relation to addiction, and he argued convincingly that you can’t argue rationally for having a longer time perspective, unless you view patience as a good in itself, independent of any good things will result from having it.

  • Roger Crisp says:

    Thx Harald. This does sound interesting — I will forward the reference to Neil.


Subscribe Via Email