When are we justified in calling mental disorders "brain disorders"?
Anneli Jefferson’s philosophical explication of whether and when we can call mental disorders “brain disorders” is one of the best that I’ve come across. Her argument is presented in detail in her book, Are Mental Disorders Brain Disorders? (Routledge, 2022) and she has previously discussed the argument in two articles, What does it take to be a brain disorder? (2020) and On Mental Illness and Broken Brains (2021). Philosophical Psychology is hosting a book symposium on Are Mental Disorders Brain Disorders? and I am contributing a commentary to it as well. Jefferson will subsequently respond to the commentaries, so this symposium is something to look forward to later in 2023 and will highlight current philosophical thinking and disagreements on this issue. In this blog post I want to present an overview of Jefferson’s central argument and share some preliminary thoughts.
Jefferson begins by examining two opposing views of brain disorders: a “narrow view,” which takes conditions such as brain cancer and neurosyphilis to be paradigmatic and requires the presence of brain pathology that can be identified and characterized as disordered independent of the presence of psychological alterations. According to this viewpoint, most classic mental disorders are not brain disorders.
The other polar position is a broad ‘brain-disorder-by-metaphysical-fiat’ view. Since all mental states are realized by the brain, if there is something wrong in the mind, it automatically follows that there must also be something wrong with the brain. “Whatever is going on in the brain that is producing dysfunctional psychological processes must itself be dysfunctional, so the thinking goes. Led by this kind of reasoning, we even get psychiatrists asserting that to deny that mental disorders are brain disorders implies a form of dualism, positing minds and brains as separate entities and allowing spooky souls back into the picture.” And “… no matter whether we find a systematic difference that is shared across people suffering from a disorder, it just follows that if what the brain is doing manifests itself in pathological psychological states, then it is itself doing something wrong.” (2021)
Jefferson offers two main objections to this broad view. One is that “there may be different standards for what counts as pathological at the level of the brain and at the level of the mind. It is theoretically possible to say that the brain is functioning as it should, but things have gone wrong at the level of the mind.” (2022, p31)
That is, what is special about mental disorders as disorders is lost in translation from the mental to the neurological, and what is left is a mere commitment to the physicalist worldview.
The second is that multiple realizability has practical consequences. Multiple realizability refers to the idea that one and the same psychological state can be realized differently by different brain states. In extreme cases of multiple realizability, we can't tell the difference between the brains of people with a certain psychological problem and the brains of people who don't have that problem.
“In other words, if psychological dysfunction is very variably realized in different people’s brains (or indeed, in the brain of one person over time), then calling the associated brain states or processes disordered is of no practical use whatsoever.” (2022, p35)
“Unless we have an identifiable brain difference which is sufficient for brain dysfunction, we lack a scientifically useful feature that would justify speaking of brain disorders. We would instead be in the situation where the only reason we have for calling these brains disordered is a metaphysical commitment to physicalism together with the claim that the property of being disordered is inherited across levels of description.” (2022, p36) (my italics)
Jefferson's goal is to describe a concept of brain disorder that avoids the problems of both the narrow view and the broad view. She thinks that this concept is also implicit in projects like the Research Domain Criteria (RDoC):
“When people like the former head of the National Institute for Mental Health in the United States claim that mental disorders are brain disorders, they are not claiming that they are conditions like neurosyphilis. They are also not saying that mental disorders are brain disorders by definition. Rather, they predict that we will, at least in some cases, find ways in which the brains of people suffering from mental health problems systematically differ from those of healthy individuals, and these differences in brain function or structure can rightly be described as dysfunctional.” (2021)
Jefferson clarifies that it is an empirical matter whether such differences will be found. The existence of systematic brain differences has to be demonstrated by scientific research. It is a matter of discovery, an open question whether we will find such systematic differences.
An important thing is that this makes the notion of brain dysfunction dependent on the presence of psychological dysfunction: “brain differences underlying mental disorders derive their status as disordered from the fact that they realize mental dysfunction and are therefore non-autonomous or dependent on the level of the mental.” (2020)
“brain differences underlying mental disorders derive their status as disordered from the fact that they realize mental dysfunction and are therefore non-autonomous or dependent on the level of the mental.” Anneli Jefferson (2020)
The hard work of deciding whether something is a disorder or a normal variation or a problem in living still has to be done in psychological terms. Merely pointing to a systematic brain difference won’t settle that question.
“… on my account we will often only be able to identify anomalies in brain processes as dysfunctional because they realize psychological dysfunction. In short, my aim is to mentalize the brain, rather than using the brain disorder label to discount the level of the mental.” (2022, p7)
Now, it may seem that finding systematic brain differences should be easy enough in many cases, and we already have many examples from the research literature, but it becomes obvious from the discussion in the book that Jefferson is aiming for more stringent criteria.
This is how Jefferson defines brain dysfunction:
“I define brain dysfunction as brain difference that realizes or causes psychological dysfunction. Much of the resistance to the brain disorder view stems from the fact that people ignore the possibility of dysfunctional realizers of mental dysfunction, focusing on preceding causes instead.” (2022, p7)
“It is sufficient for X to be a dysfunctional type of brain process if tokens of this type always realize a psychological dysfunction.” (2022, p39)
Here’s an example she discusses:
“Assume, for the sake of argument, that we find differences in the dopamine system in addiction that underlie the cravings for the drug that we are familiar with on the psychological level and that we have characterized as dysfunctional psychological processes. There is promising research in this area (Holton and Berridge 2013). Let us further assume that these differences in brain function are sufficient for the cravings, that we will not find this kind of brain difference without cravings. This means that this pattern of brain function realizes cravings. In other words, it fulfils the condition of being sufficient for mental dysfunction and realizing that dysfunction specified above. This would then be a case where a mental dysfunction is also a brain dysfunction.” (2022, p 40) (my emphasis)
Psychiatry at the Margins is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
Let me identify two distinct but related questions:
1) “When are we justified in extending psychological norms/standards, by which decide that psychological processes are not operating as they are supposed to, to the brain?”
In Jefferson’s words: “Our question is how to extend the notion of brain dysfunction to those brain processes that realize the psychological dysfunctions we find in mental disorders.” (p 39)
2) “When are we justified in calling a psychological phenomenon a brain phenomenon?” “When is a psychological difference also a brain difference?”
For Jefferson, the answer to the second question is that an identifiable psychological state must be realized by an identifiable brain state. A psychological difference must correspond to a brain difference that we can identify, and this brain difference must be sufficient for the psychological difference to exist.
To restate: The brains of individuals with a psychological difference X from must be distinguishable from the brains of individuals without the psychological difference X. AND The brain difference must be sufficient to realize the psychological difference. Brain difference Y should always be accompanied by the psychological difference X.
If we have identified a brain difference that is sufficient to realize a psychological difference, and if the psychological difference is a psychological dysfunction (based on psychological norms), then according to Jefferson we are justified in extending the notion of dysfunction based on psychological norms to the brain processes. (Provided we recognize that the status of brain differences as disordered is derived from the mental.)
I think Jefferson is right to develop a notion of brain disorder where brain dysfunction is dependent on psychological dysfunction, and her articulation of this notion allows for the mental disorder–brain disorder debate to advance by breaking the impasse between the narrow and the broad views. However, the requirements of realization and sufficiency strike me as quite arduous to satisfy, to a point where I am uncertain that they can be met for mental disorders.
There is something awkward about the notion of realization itself. As Shoemaker has recognized, “the notion of realization is entangled with the history of functionalism in the philosophy of mind” (2007, p 2) and it is difficult to work with the notion if we are not operating within this history.
In words of Shoemaker: “The relation between a realizer and what it realizes is a constitutive relation – the having of a realized property consists in the having of whatever property realized it on that occasion. The occurrence of realized states is “nothing over and above” the occurrence of their realizers.” (2007, p 2)
The notion that a psychological difference X is nothing over and above a brain difference Y makes me uncomfortable. I generally accept that the mental is nothing over and above the physical (broadly understood), but if psychological processes are in a mereological relationship with neurophysiological processes in the context of an organism in interaction with the environment (as, e.g., Sanneke de Haan posits within an enactive framework), and if the physiological and the experiential are two different ways of looking at the same process, then I am not sure it makes much sense to say that the psychological processes are nothing over and above the neurophysiological processes.
I am not a philosopher. I’m not quite sure how realization works in a mereological context and whether it is compatible with what Jefferson has in mind. (If you have a better conceptual grip on this and are able to explain in an accessible manner, please do!) This article Realization Relations in Metaphysics by Umut Baysan does outline a category of Mereological Realization where complex functions of large systems are realized by less complex functions of the parts of these systems and where qualitatively distinct higher-level properties are realized by lower-level properties, but the article also cites Shoemaker “we need an account of realization that gives a role to the properties of micro-entities other parts of macroscopic objects … [but] the cure for this is not to count the properties of parts of macroscopic objects as the realizers of properties of the macroscopic objects.”
Philosophical technicalities of “realization” aside, my other worry is that it is very difficult to demonstrate that a brain difference Y is sufficient for – is always accompanied by – a psychological difference X. We have not been able to demonstrate such a relationship even for bona fide brain disorders such as Alzheimer’s disease and Parkinson’s disease. Brain changes such as amyloid and tau deposition aren’t even sufficient for cognitive impairment, let alone disturbances in mood and perception that are often seen in Alzheimer’s. Have we demonstrated what brain differences realize psychosis in Parkinson’s disease psychosis? We know that Parkinson’s disease pathology causes psychosis, but the requirements for realization appear to be more stringent.
It is worth pointing out that even for conditions like Huntington’s disease the requirements of the pathological change being sufficient and always associated with clinical presentation of interest aren’t met. There are many forms of Huntington’s disease mutations that have reduced penetrance (with CAG repeat length 36-38 apparently having penetrance as low as 6-20%).
What if we identify consistent brain differences but they are associated with a psychological dysfunction only 70% or 80% of the time? Or what if we identify consistent and systematic brain differences but only at the group level and not at the individual level? A classic example of that would be schizophrenia, where elevated in vivo markers of presynaptic striatal dopamine activity have been consistently reported, and the elevation in dopamine synthesis capacity shows a large effect size (Howes, et al. 2012; Howes et al. 2013). This is a substantial, reliable, and systematic difference, but it falls short of the brain difference being sufficient for psychological dysfunction.
What are we to make of such empirical findings? This is not the anarchy of extreme multiple realizability where no differences are identifiable, and this is not the sufficient realization relationship that Jefferson has in mind. Can we posit another Goldilocks solution to the justification question in addition to the one proposed by Jefferson?
We can consider the pros and cons of justifying calling mental disorders as brain disorders on the basis of substantial, reliable, and systematic brain differences, but I see no obvious reason why we should dismiss it as a possible answer to the question.
Update — See follow-up to this post here:
Comments on this post are open to all subscribers.
Psychiatry must be a big mess if we ask if mental disorders are true "brain disorders". Do gastroenterologist question whether hepatitis is a disorder of the liver or do urologists challenge the concept of an enlarged prostate? They don’t. Yet, psychiatrists keep wondering if there is something ethereal hiding behind amygdalae and basal ganglia. No wonder other physicians roll their collective eyes listening to us.
Of course, mental disorders are disorders of the brain. What else could they be?
But I understand how someone might get confused into thinking that these are societal, not biological. Take voyeurism, for instance. It’s not pathological to achieve sexual arousal by looking at naked or undressing people. It turns into a disease when one can’t stop spying on non-consenting people and is arrested and prosecuted for it. ‘The confused’ fails to recognize that voyeurism is not the disease of sexual arousal (the drive) but a mismatch with weak biological inhibition (the brakes) in the face of obvious negative consequences. By analogy, if someone crushes a car with a Corvette engine and Fiat 500 brakes, it would be a car and the driver problem (one or both must be fixed) and not the freeway with rules and speed signs.
Even though society labeled voyeurs mentally ill, voyeurism is not a social disease. Instead, it’s the inability to control the drive. Voyeurism is the disease of brain neurocircuits, not society. The emergence of novel diseases with social and economic advances is not new. For example, there was no common issue with presbyopia during the Stone Age; no learning disabilities and ADHD before universal education; pedophilia was not a problem before it became illegal; and fewer were labeled as exhibitionists after the opening of nudist beaches.
Most of the confusion in psychiatric taxonomy and diagnosis comes from the DSM, which avoids talking about the nature of mental diseases, leaving it to the readers to define. Psychiatric diagnosis would look less mysterious and confusing if one applies a sensible medical model. For that, we must accept the brain as a collection of separate, closely related units, a.k.a. neurocircuits. Ultimately, their dysfunction (it can be deficit, excess, instability, or mismatch) determines thoughts, emotions, and behaviors that we identify as pathological, leading to mental disorders.
In medicine, the evaluation starts with a “review of systems” (cardiovascular, GI, endocrine, nervous, etc.) Grouping symptoms and linking them back to the “units” that generate them is the standard medical practice. In psychiatry, we should start an evaluation with a review of functional neurocircuits. The brain has myriads of them, each serving specific functions, but not all are relevant to psychiatric diagnosis. Mostly, these are the circuits for arousal, mood, automaticity, executive control, reality testing, social relatedness, and information processing. In medicine, the treatments target the organs. In psychiatry, the target is neurocircuits, not descriptive diagnoses, despite the “indications.”
American psychiatrists have been taught to gather bits and pieces from patients’ narratives and turn them into a descriptive label (e.g., major depression, post-traumatic stress disorder, autistic disorder, schizophrenia, etc.). Then, treat the label as a whole. That’s how DSM avoids mentioning the brain and neurocircuits. In contrast to medical classifications, the Manual cuts off the link between symptoms and the organs that generate them. That is its major failure. However, it will be impractical to “repeal and replace” DSM. Instead, we should begin with deconstructing DSM diagnoses. First, separate and then group the jumbled and sometimes contradictory criteria according to contributing neurocircuit.
There is more; unfortunately, blogs' format has limitations. Another time, perhaps.