Brain Disorders in Entangled Brains

More on Mental Disorders as Brain Disorders

This is a follow-up to “When are we justified in calling mental disorders "brain disorders"?” and builds on crucial details from my prior discussion.

When are we justified in calling mental disorders "brain disorders"?

We can't talk about what a brain disorder is and how psychiatric disorders might be brain disorders without taking into account how neuroscientists understand the relationship between brain and behavior. Many neuroscientists have challenged long-standing assumptions of the field in recent years, and these include assumptions that are relevant to the ‘are mental disorders brain disorders’ debate. Without getting into specific disagreements, we can say that there are a number of broad principles that are supported by many prominent neuroscientists. (I know this is a topic that is still being debated, and I don't want to give the impression that the debate is over.) Luiz Pessoa’s latest book, The Entangled Brain (2022) illustrates these developments in neuroscience quite well. Pessoa summarized the main thesis of the book for an article in the Journal of Cognitive Neuroscience, which I am using as my reference.

Here’s how Pessoa explains the intended meaning of the term “entangled”:

“… the sense of “entangled” is one in which brain parts dynamically assemble into coalitions that support complex cognitive–emotional behaviors, coalitions composed of parts that “jointly” do their job. Thus, an entangled system is a deeply context-dependent one in which the function of parts (such as a brain region, or a population of cells within a region) must be understood in terms of other parts: an interactionally complex system”

Pessoa understands the brain as a complex system where organizational relationships and dynamic interactions among parts of the system lead to novel, collective properties. He invokes 3 principles of brain organization:

1. massive combinatorial anatomical connectivity

2. highly distributed functional coordination

3. networks/circuits as functional units

Here are several quotes from Pessoa’s article offering further elaboration:

“The overall picture is one of massive interconnectivity, leading to “combinatorial” pathways between sectors. In other words, one can go from point A to point B in a multitude of ways. We propose that, combined, connectivity systems spanning the entire neuroaxis (cortical forebrain, subcortical forebrain, midbrain, and hindbrain) provide the basis for both broadcasting and integration of diverse signals linked to the external and internal worlds. Such crisscrossing connectional systems support the interaction and integration of signals that are typically associated with standard mental domains, including emotion, motivation, perception, cognition, and action (Pessoa, 2013) but, critically, in a manner that does not abide by putative boundaries between these categories (see below). I propose that this general architecture supports a degree of computational flexibility that enables animals to cope successfully with complex and ever-changing environments” (my emphasis)

“The complexity of anatomical pathways allows signals to flow across the brain in a staggeringly large set of ways. Anatomy provides a backbone that constrains function, but the structure–function relationship is anything but simple… the relationship between signals in disparate parts of the brain is not determined by structural pathways in a straightforward manner.”

“In a highly interconnected system, to understand function, we need to shift away from thinking in terms of individual brain regions: The network itself is the functional unit, not the brain area…”

“Networks should be considered inherently overlapping and dynamic. Parts of the brain (say, populations of neurons within areas) affiliate dynamically with other elements in a highly context dependent manner driven by the current endogenous and exogenous demands and opportunities present to the animal. Critically, network properties are novel (with respect to that of individual regions), and key functions are distributed across regions or neuronal populations.”

“Categories such as perception, cognition, action, emotion, and motivation organize how we understand and study brain function. But are such mental domains consistent with the framework described here? In a nutshell, “no.” The standard decomposition adopted by neuroscientists requires an organization that is fairly modular, which is inconsistent with the principles of the anatomical and functional neuroarchitecture discussed. In general, mental processes of interest cut across domains and do not respect putative boundaries between traditional systems (e.g., emotion, cognition). In fact, crisscrossing anatomical/functional connectional systems dissolve potential lines of demarcation.” (my emphasis)

---

Now let’s consider another paper, published just last week in Trends in Cognitive Sciences, by Lisa Feldman Barrett’s group: “Improving the study of brain-behavior relationships by revisiting basic assumptions.” Westlin et al. challenge 3 traditional assumption of brain-behavior relationships and offer 3 revised assumptions.

Traditional assumptions

1. Localization assumption: instances of a psychological category can be localized to a dedicated neural ensemble.

2. One-to-one assumption: neural ensembles correspond one-to-one with psychological categories.

3. Independence assumption: a stimulus will reliably evoke activity in a specific neural ensemble that produces an instance of the specific psychological category of interest.

Revised assumptions:

1. Whole-brain signals contribute to mental events: instances of a psychological category arise from activity across the entire brain, not from a separable neural ensemble.

2. Many neural ensembles for one psychological category: there are degenerate (many-to-one) mappings between neural ensembles and a psychological category.

3. Mental events emerge as a complex ensemble of signals: an instance of a psychological category emerges from a complex ensemble of signals from the brain, body, and world. These signals can only be understood in relation to the rest of the ensemble; i.e., each may have a weak effect on its own, but a strong effect when considered collectively.

The paper generated a lot of debate on neuroscience twitter, which I observed with great fascination as a neuroscience outsider. Some neuroscientists seemed particularly irked by Westlin et al. to a degree that mystified me (I’m quite mindful that even within my field of psychiatry, conceptual issues that appear to be of minor significance to outsiders can generate very heated debates within the profession), but some responses were very insightful.

Kevin Mitchell, for instance, tweeted the following, and I think he’s right on the money: (I have converted his tweets into paragraph form, you can see the thread here)

“IMO, some pushback against simplistic ideas of function localisation is reasonable (though I think the field has moved on from ‘blobology,’ to be fair). By simplistic, I mean taking brain region A to “do X” (based on activity patterns or lesion info), which seems to imply sufficiency, not just necessity. There’s an analogous pitfall in genetics with the common construction: a “gene for X” (as opposed to the more careful: a gene required for X). However, that caveat doesn't invalidate the specificity of activation patterns and connectivity and lesion effects, etc., that support differential involvement of brain structures in different functions. And the fact that such functions heavily rely on interconnectivity and can be modulated by all sorts of contextual information doesn't mean it's some kind of neural free-for-all.”

---

At this point you might be beginning to wonder how all this relates to the issue of whether mental disorders are brain disorders. Let’s recall that Anneli Jefferson insists on a criterion of sufficiency: “It is sufficient for X to be a dysfunctional type of brain process if tokens of this type always realize a psychological dysfunction.” For differences in brain function to be sufficient for a psychiatric problem, the brain difference cannot exist without the psychiatric problem also being experienced. A psychiatric condition must correspond to a brain difference that we can identify, and this brain difference must be sufficient for and always realize the psychiatric condition. According to Jefferson, it is only this stringent criterion of sufficiency that allows us to attribute disorderedness to brain differences underlying mental disorders (i.e., that allows us to transpose the psychological norms by which we judge a behavior to be dysfunctional onto the brain).

What’s the issue? The problem is that if neuroscientists such as Pessoa, Feldman Barrett, and Mitchell are correct, this stringent criterion of sufficiency is scientifically impossible. Our best neuroscientific understanding of the brain-behavior relationship tells us that this sort of sufficient one-to-one correspondence, as a matter of general principle, is not going to be demonstrated. An account of brain disorders that fails to take this into account just isn’t going to be very useful.

Now, a philosopher could perhaps maintain that it can be theoretically possible for a brain difference to be sufficient for a mental difference, provided we possess an absolute and perfect knowledge of each and every detail of the brain states involved (à la  Laplace's demon), such that even the knowledge of the influence of past brain states and knowledge of contextual influence from the body and world is also included/incorporated in the brain difference. I am not advocating for the viability of this thought experiment, but I want to say that even if it were viable, such knowledge is inaccessible to us. Even if it is metaphysically possible, it’s not epistemically possible.

What can we reasonably expect instead? Similar to Mitchell, I’ll say that we can expect differential involvement of brain functions – expressed, for example, with reference to brain networks and their dynamic activity – in different psychiatric conditions. These relationships will neither be necessary nor sufficient; they may not even be highly specific.

In my previous post, I proposed that we could consider psychiatric disorders as brain disorders if there are associated substantial, reliable, and systematic brain differences. The preceding discussion supports this criteria, but there is an element missing. Even for normal psychological phenomena, establishing specific and sufficient relationships can be an elusive goal. For example, consider Feldman Barrett’s assertion: “Even after a century of effort, scientific research has not revealed a consistent, physical fingerprint for even a single emotion.” [How Emotions Are Made. 2017. p 11]

Yet, it would be very odd for us to debate – as we often debate in the case of psychiatric conditions – whether emotions have a basis in the brain. Emotions are realized in the brain, they are made possible by the brain. Even if we cannot identify a brain state that always realizes an emotion, the brain-based nature of emotions is not in dispute. We instead focus on how to best formulate and theorize the relationship between emotions and brain networks. The relationship is such that it involves many areas of the brain, there is no one-to-one mapping, and the emotions emerge from a computational process that involves complex dynamic interactions in a highly context dependent manner. This theoretical framework of the brain-behavior relationship is the element missing from my prior definition.

This then is the aspirational goal of the mental-disorders-are-brain-disorders project: a psychiatric disorder can be said to be a brain disorder if there are empirically identifiable systematic brain differences associated with it and these associations are contextualized within a scientifically robust theoretical understanding of the relationship between brain and behavior. Consistent with the spirit of Jefferson’s account, this is still an empirical matter. It is an open scientific question whether this goal will be achieved and for which conditions. It is also a goal that meaningfully aligns with the research agenda of biological psychiatry and frameworks such as RDoC.

Before I end, I want to briefly make a point about why we insist on distinguishing between mental disorders and brain disorder in the first place. As argued by Matthew Broome and Lisa Bortolotti (2009), mental disorders have “distinctive features [that] can be adequately characterized only by using the vocabulary of the mental. We do not deny that psychiatric disorders can be described as disturbances of neurobiological mechanisms, but we insist that they are pathological in virtue of their manifestations, and they manifest as disturbances of the mind.” [Mental illness as mental: in defence of psychological realism. Humana Mente]

It is this distinctively mental nature of psychiatric disorders that has to be acknowledged and respected by any account of mental-disorders-as-brain-disorders. The status of psychiatric disorders as mental disorders is secure, even if and when we are justified in characterizing them as brain disorders.