The lecturer moved his laser-pointer quickly over the loop of neural circuitry. He explained the role of Mu receptors in activating the circuit, which sent a signal round and round and came out as the behavior pattern we call addiction. It was all very neat.
It was so neat the he could have simplified his diagram by replacing the pretty brain graphic with a switch. Off would be synonymous with no addictive behavior. On would equal addictive behavior. If you took the theory, “Addiction = Brain Circuitry” at face value, anything that flips the switch would cause addiction. Yet we know that that situation does not obtain. Heroin flips the switch, but not everyone who takes heroin manifests addictive behavior.
For the advocate of “Addiction = Brain Circuitry”, there are two ways out of this dilemma. First, he can posit a multiplicity of switches. In other words, he can claim that there is an intervening network of necessary, but not sufficient, switches on either side of the Big Switch, mediating the input and output of the addiction circuit. But then in principle, all those switches could also be replaced with a single switch, and you are right back where you started. No limited set of if/then statements will be completely determinative.
The second way out of the non-correspondence dilemma is to simply abandon a complete and transparent explanation, in favor of reliable facts. Neurons are necessary to behaviors, and we know that because, if we zap certain neurons, we can reliably alter corresponding behaviors. That doesn’t exactly explain the behavior, but it lets us move on to knowledge of neural circuits and the experiences which correspond with changing the configurations of those circuits.
One might denigrate the second solution as an abandonment of truth-seeking. Perhaps, but that is not so bad, on a proper notion of truth. In solution #2, you get a theory, which is a set of reliable facts. To get to the truth what you need is an explanatory reduction. In other words, all the switches and their positions for a specific moment of behavior, across the cosmic board. Such an array is purely didactic. It refers to no knowledge, for it cannot reliably correspond with anything. You may think you know something about it, but you don’t – not until you begin to formulate a theory regarding it.
Johnnie shoots a dose of heroin because he has inherited a susceptible set of receptors, because he contains the dendritic representations of certain permissive life-lessons, because he lacks certain inhibitory representations, because he lives in a society which has heroin, because he anticipates certain effects from heroin injections. And on, and on, and on…
At the end of such an exposition (if there even is an end) what we have is just a snap-shot which we have pre-labeled, “Johnnie’s Addiction”. To make any sense of it – to know anything at all about it – we must delve in to the insufficient necessities, and be satisfied with their mere reliability. When we give Johnnie a medicine for his Addiction, we should expect that it will, to some extent, extinguish the behavior. We should expect that if we take away his heroin, his behavior will, to some extent, change. And in fact, our theory does correspond with the facts which it predicts, and upon which rests.
Like the addiction lecturer, we all frequently feel dissatisfied with reliability. We would like some non-provisional knowledge. Give us some truth, please. Aspiring to truth gets us nowhere, though. Truth is too hefty. To riff on Gettier’s classic thought experiment, Smith has the truth when he observes that a person with 10 coins in his pocket will get the job, once Smith sees that a person with 10 coins in his pocket gets the job. Yet he has no knowledge thereby. He cannot be (provisionally) right or wrong in such a statement, any more than a snapshot can be right or wrong (though our subsequent interpretations – theories – of the snapshot may be).
If Smith says, at his next interview, that the person with 10 coins in his pocket will get the job, and he takes care to put 10 coins in his own pocket in hopes of getting the job, then he may know something. He is making a knowledge claim regarding his experience with coins and interviews, and his claim may or may not correspond with his theory’s fact-conditions. Reliability is what he will get, and he will be happy with it, or not, as will we all.
Dedicated to the Game 
But is the heroin user trying to turn the switch on or off? 😁
As someone who’s used many mind altering substances (so my opinion is based on purely personal and anecdotal evidence) the “addict” is just treading water. The drug is a survival tool.
That’s the really interesting question.
For some, it seems that they are just easily subsumed by the feedback loop. For others, it seems like living in the loop avoids living somewhere less pleasant.
And many cross over as their receptors downregulate.
On and off both, I’d say. though the behavioral output looks superficially similar, thus the common label.