The Neuroscience of Justice, or, the Tumour as Agent

 

 

This post was actually written in the summer of 2011, so please pardon its dust.

I’ve just finished reading David Eagleman’s recent article “The Brain on Trial” (The Atlantic,July/August 2011, available online here), and feel… threatened, I suppose. I strongly suggest reading it – which is why it threatens, because its logic is fair and its intent agreeable; certainly the effort has my sympathies. But looked at closely, this picture of justice and culpability has many faultlines, and building a new judicial system on its terrain is, at best, unwise and, at worst, a guarantee of the unacceptable triumph of the who over the what in sentencing. (Of course, neither is ideal – but replacing one inadequate extreme with its equally or more problematic opposite is preposterous.)

I won’t synopsize, as we’ve all got lives to tend to, but in brief (and with excerpts):

The dense interaction of neural networks, the lack of an “uncaused cause” (a la God) within the brain, the lack of a truly “free” will, which follows, and the supposed progress made in identifying the biological bases of aberrant behaviour all lead him to the conclusion that the presumed subject of judicial sanction – the deliberate “criminal” – is a convenient fiction. That is, the coherent, deliberate and rational decision-maker who has chosen to break the law or delve into perversity is a construction that maps rarely, if ever, onto reality. (I agree with this, though with some caveats that lie outside my present purpose.) As “possession,” reckless abandon and the sin of drunkenness have been demoralized and medicalized as, for example, schizophrenia, bipolar disorder and substance-abuse disorders, so too will pedophilia, shoplifting and premeditated murder ultimately be reduced to the plane of neurological disease or dysfunction.

As we wait for technology to catch up with this intuition, and provide the means of identifying the concrete brain mechanisms of criminality, Eagleman and his colleagues have developed a system meant to maximize reflection and self-control, and so to minimize recidivism (not changing the criminal instincts, whatever they are, but toughening the muscles of restraint). From his article:

We may be on the cusp of finding new rehabilitative strategies as well, affording people better control of their behavior, even in the absence of external authority. To help a citizen reintegrate into society, the ethical goal is to change him as little as possible while bringing his behavior into line with society’s needs. My colleagues and I are proposing a new approach, one that grows from the understanding that the brain operates like a team of rivals, with different neural populations competing to control the single output channel of behavior. Because it’s a competition, the outcome can be tipped. I call the approach “the prefrontal workout.”

 

And on the increased ability to control impulse:

This prefrontal workout is designed to better balance the debate between the long- and short-term parties of the brain, giving the option of reflection before action to those who lack it. And really, that’s all maturation is… The frontal lobes are sometimes called the organ of socialization, because becoming socialized largely involves developing the circuitry to squelch our first impulses.

 

So, where’s the problem? There are several, and for the sake of length I’ll stick to a couple, and for the sake of clarity I’ll divide them into the descriptive and prescriptive. Hell, for the sake of all of our afternoons, I’ll even use bullet points.

Descriptive

• The increased and increasing ability to identify neurological bases for behaviour. A comparison is drawn between the supposed bases of criminality and those of, e.g., depression. Indeed, the comparison may be worthwhile, but not in the way Eagleman thinks. While he’s right that specific drugs have concrete effects for some mental illnesses – reduced symptoms, increased functioning – it doesn’t follow that these illnesses themselves have a discrete etiological base (“etiology” refers to the cause or set of causes of a medical condition). Quite unlike, say, a broken leg, mental illnesses are diagnosed via a set of symptom checklists, as provided by the near-biblical Diagnostic and Statistical Manual (DSM-IV-TR, soon to be in its 5th edition), rather than by any sort of clear identification of a “cause” (i.e. no faulty wiring, no neural cluster, no leaky cortical piping). Diseases of this sort refer to clusters of symptoms, and not single physical causes – indeed, the DSM and its lists of symptoms strive for reliability (consistency of application and diagnosis), notvalidity (a demonstrable connection with neural reality). While drugs do treat patients effectively, they are not rationally designed to correct known and specific neural problems, but are instead tested and re-tested for their ability to mollify symptoms, and then applied in cases where they have the most effect – hence their changing application and use in treatment (see: Risperidone), off-label uses, etc. In fact, the diagnostic criteria in the DSM have evolved to better match the drugs, rather than the reverse. The Manual’s third edition, the first that could be called “modern” and biomedical in its models, sought to provide reliable ways to gather together populations of patients for drugs testing; the testing results would change the working definitions of “persons with syndrome X,” the diagnostic criteria would change (“be refined”) as a result, thus affecting later patient testing populations, and so later diagnostic revisions, etc., in an ongoing feedback loop. So yes, perhaps criminality has neurobiological bases, but in this respect their similarity with mental illnesses is telling: they would always lie a step beyond our efforts for direct identification. For more on certain aspects of the above, I strongly recommend Andrew Lakoff’s book Pharmaceutical Reason: Knowledge and Value in Global Psychiatry, available here. (I realize some may argue that neural imaging allows for identification of specific regions of the brain that may serve as the “root” of criminality or brain dysfunction. I counter with the suggestion that the dense interconnection of neural networks, which is taken as a given in Eagleman’s case against free will, means that the more precise and localized the region identified, the less it will be determining of any given set of criminal behaviours or mental symptoms, as such – dense interconnection means that any such behaviour, at least as currently known and recognized, can’t be reduced to any one region. Any meaningful etiological claims – that is, identifying elements that are relevant to and determining of criminal behaviour – most likely must be multifactorial and span several registers: the neuroanatomical, the psychological, the social, etc.)
• The loss of difference. I’ll exploit the comparison with mental illness once again. Lakoff’s book explores how diagnosis of the DSM sort is able to bring incredibly disparate experiences together under the label, e.g., “depression.” He has a useful example: a suicidal teenaged woman in Buenos Aires is said to have the same mental illness as a recently-bankrupted 54-year-old male in Texas, despite the extreme difference of illness onset, prognosis, treatment options, social context, expression of symptoms, etc. These two cases are brought together by the assumption that a shared cluster of symptoms is indicative of a shared “disease” – that is, a shared chemical imbalance of some sort. While this may be true, virtually all relevant aspects of their suffering, treatment and prognosis differ substantially (with the exception perhaps of one of the several drugs each would likely be treated with). While useful in some contexts (notably, the global sale of pharmaceuticals), this diagnostic convergence masks deep differences. The same would be true of reducing criminal behaviour – which takes a multitude of shapes across a staggering range of social and personal contexts – to a common neural substrate. Premeditation, lack of remorse, denial; these may serve as markers for recidivism in a parole hearing, but it’d be problematic to use them as diagnostic criteria for Criminal Condition X, because it’d reduce or exclude some of the most important dimensions of social crime: class, colour, creed, etc. This would be a quite a loss in our discourses on crime and justice. (To be fair, Eagleman actually accounts for much variation; his understanding of causality is as something extensive and diffuse, and so gives due credit to the relevance of the personal and idiosyncratic. But in practice, reducing that which occurs at the level of society to specific, recurrent neural problems risks a staggering loss of comprehension.)
• The equally arbitrary drawing of boundaries. Where draw the line of agency? We can contract the boundaries of the responsible, ethical agent to circle tightly around a given neural region, but we can with as much validity extend the boundaries to include the immensity of all social contact and influence. Neither is a given, more ontologically true than the other. Our judicial system at the moment finds an expedient, though flawed, middle ground – assuming agency begins and ends with the individual, but that this agency is more or less mature, more or less compromised, and so “going easy” on those with mental illness, who aren’t of age of consent, who were engulfed by the passions. Examples of either pole – of the contracted and neural, of the extended and diffuse – needn’t be drawn solely from other cultures, marked by supposed exoticism and so inapplicable to our world. Of the former, Eagleman provides examples, and indeed there are many more: a father, stricken with a brain tumour, who without precedent savagely murders his son; a mother, suffering a post-partum chemical imbalance, who drowns her children in a car; and on, and on. Of the latter, I ask you to picture a parent taking responsibility for the actions of his or her drunken son, or an apartment owner paying damages to an injured pedestrian, deemed liable for the accidental fall of his air conditioner from the sill to the sidewalk below. What constitutes a “culpable agent,” in our own society as in all others, is very broad and context-dependent. In effect, then, I agree with Eagleman insofar as this flexible view of responsibility calls for a judicial reform of some sort (I can make no claim as to what, exactly); but I take issue with the implication that the neural offers a more solid bedrock for attributions or negations of responsibility in all or most cases.

Prescriptive

• See A Clockwork Orange.

P.S. Please, hold the comments about how I obviously don’t understand the purpose of bullet points. We’re all doing our best over here at… what’s this blog called? Whatever, finish your drink and shuffle on out of here.