What are the Units of the Mind?

Diagnosis sets a problem to which treatment must propose a solution.  In empirically mature medical specialities, both are framed in units of of physical force, energy, or spatial extension.  Unwanted growth and its subsequent shrinkage, high pressure and its reduction, irregular rhythm and its normalization: each is described in millimeters, milliliters, cells per deciliter, calories, millivolts, hertz, or any of a dozen other standard scientific units.  These units are the means by which medicine connects math to the physical world, and ensures that the patient’s body is carried along for the ride when treatments that change the equation of disease are described. For example in modern medicine coronary artery disease (CAD) is defined as arterial lumen narrowing, and this diagnostic premise leads deductively to the conclusion that a balloon catheter might be inserted into the narrowed artery and then inflated to increase lumen width

.  Such deductive logic is deployed every day by hundreds of thousands of health professionals, on behalf of millions of patients, across a huge range of medical illnesses, in virtually every medical specialty


Psychiatry is not among them.  Nor will it be until at least 2020, when the reign of the forthcoming edition of the Diagnostic and Statistical Manual (DSM-5) comes to an end.  Even then, there is no guarantee that psychiatry will join the medical mainstream; the Research Domaine Criteria (RDoC) may not be ready to take over yet, and DSM may continue to do then what it always has done: define psychiatric illness using words.  Nowhere will these words be mathematically tractable or attached to units of physical force, energy, or spatial extension.  As such psychiatry will continue to lack what it has always lacked: a reasonable explanation of how the words “depression,” “mania,” “anxiety,” “addiction,” or “psychosis” might be connected to the organ that produces them, the brain.

Bizarrely, psychiatry routinely prescribes somatic treatments that are linked to the brain by units in math; as receptor agonists are prescribed in milligrams, ECT is delivered in volts, and deep brain stimulation is delivered to precise anatomical coordinates.  The fact that treatments are somatic does not solve any problems; on the contrary, it causes a new.  We call this the “deductive problem,” which is that  psychiatry’s diagnoses are not illogically related to its somatic treatments.  To see this, consider the current consensus that serotonin reuptake inhibitors (SRI) improve depressed mood by increasing BDNF secretion and dendritic arborization in the hippocampus.  This conclusion does not follow logically from the DSM definition of depression, which does not mention the hippocampus, dendrites, or serotonin, let alone quantify normal values for them.

This is not to say psychiatry’s somatic treatments do not work.  They frequently do.  It is merely to say that in the context of the DSM they do not make any logical sense.

Psychiatry’s inability to attach the mind to the brain is a point of pride in some quarters, and considered merely nonproblematic in others.  But within mainstream medicine it is a source of embarassment, and for most of the 20th century has burdened psychiatry with second-class citizenship within the profession.  Medicine regards it as a first principle that biological processes must be causally attributed to the organs that produce them – heartbeats to the heart, respiration to the lungs, and digestion to the duodenum.  Psychiatry’s failure to explain how the brain produces the mind is thus a violation of medical etiquette.

In fact this rudeness affects all medical specialties.  It is caused by what the philosophy of consciousness calls “the hard problem,” which is the challenge of explaining causally how the brain gives rise to the mind and its qualia – joy, fear, the color red.   However should cardiology attempt to explain how a heart attack causes chest pain, it will get only as far as describing which pain fibers are stimulated during muscle hypoxia.  This correlation between nerve cell activity and subjective experience is what the philosophy of consciousness calls “the easy problem,” and solving it is nontrivial but philosophically simple, because it merely involves correlation, not causality.  Psychiatry is well ahead of all other fields in addressing the easy problem. But cardiology cannot explain how or even why pain fiber firing causes conscious pain in the first place.  In this respect, it has nothing on psychiatry; they are equally stymied.

Which brings us to a central question: why is cardiology not embarassed by its inability to solve the hard problem?  Here we suggest a simple answer: it is not trying to.  Cardiology has organized the relationship between diagnosis and treatment so as to omit any need to explain how subjectivity arises. Psychiatry has not.  A diagnosis of a heart attack does not require the symptom of conscious pain.  It is as easy for an electrocardiogram (ECG) to detect S-T elevation in an unconscious patient as in a conscious one.  Thus cardiology need not worry why heart attack patients feel pain; it can simply go about detecting and then treating their heart attacks using instrumentation.  In contrast, it is impossible for psychiatry to detect depression in an unconscious patient.  There are both philosophical and practical reasons for this.  Philosophically, psychiatry believes it cannot avoid the hard problem, has attempted to answer it, and has failed.  It has thus has reasonably concluded that it cannot possibly know with which units to measure the de facto immaterial mind. This leads directly to the practical problem: unable to choose brain waves, cortical thickness, membrane potential or any other mathematically tractable physical candidate for the task of measuring the mind, psychiatry naturally lacks a diagnostic instrument with which to detect mental illness. Where pulmonolgy has the stethescope, opthalmology the opthalmoscope, and cardiology the ECG, psychiatry has only the chair. Psychiatry cannot rely on its own judgment in reaching diagnosis.  The rest of medicine can.  Psychiatry must talk to its patients. The rest of medicine need not. Psychiatry is thus inherently relational; medicine is not.

But the hypocritical score of other specialties is certainly not a good basis upon which to psychiatry should redouble its efforts to measure the mind in scientific units. Patient care is. And here we point out that psychiatry’s inability to measure the mind negatively affects the health of patients in myriad ways. Several of these are relatively obvious, as they are due solely to the absence of technological and laboratory tests for psychiatric illness.  Such tests would allow the resolution of diagnostic debates and create standardization and consistency in diagnosis across clinicians; tests would permit the detection of illness too subtle to be noticed during an interview; tests could be done without the presence of a psychiatrist, with only patients with positive results getting appointments, thus rationalizing allocation of care while increasing screening; tests are likely to destigmatize psychiatric illness and thereby increase the use of psychiatric care, thereby lowering the burden of psychiatric illness, as the public tends to regard physical findings as legitimate medical problems while dismissing purely subjective conditions; tests might allow the evaluation of treatment efficacy and thus speed patients’ progression through ineffective treatment and the identification of useful ones; and tests might reduce the psychiatrist’s reliance on the patient’s capacity for sensitive, specific, accurate introspection, which may vary across the population.

However a more fundamental reason for adopting units by which to measure the mind is that this will allow the detection of unconscious processes, long the dream but never the reality of field.  Indeed it often escapes the attention of practitioners that as presently constituted, psychiatry can postulate but not detect unconscious contributions to mental life.

There is abundant clinical evidence that most patients are unaware of maladaptive motivational processes that damage their wellbeing, even as their friends and family gossip easily about how they “don’t want to grow up,” or “are afraid of success,” or “always sabotage their relationships by falling in love with unavailable women.”  Most psychotherapists spend the majority of their time attempting to make patients aware, years into treatment, of problems the therapist noticed in the first few sessions; clinically the unconscious is patently obvious.

Further, there is overwhelming neuroscientific evidence that that vast majority of neural processing is unconscious, that suboptimal processing can occur in these unconscious regions, and finally that this suboptimality can lead to psychiatric illness.   Prudence would appear to dictate a means by which to understand the informational content of unconscious processing, which moving forward we will call the informational correlates of neurons (ICN). It is worth noting how entirely unradical and frankly conventional this emphasis on neuroscience’s potential to discover unconscious causes of mental illness is.  In fact it is so conventional that for decades the overwhelming majority of neuroscience articles have simply assumed that it is the job of neuroscience to observe the physical correlates of unconscious mental processes: they attempt to analyze the informational implications of neural processes and structures that, as when in lower animals, clearly lack conscious correlates. In academic psychiatry, the movement to predict the development of psychiatric illness by measuring cortical thickness, white matter connectivity, and subcortical activity is an index of the field’s impulse to look beyond what consciousness can tell us about the human experience

.  As discussed below, there has been much talk about in the philsophy of science about the neural correlates of consciousness (NCC).  Here we talk about a distinct, and logically dubious second category of correlation, the neural correlates of unconscious mental processes (NCU). For the remainder of this article, we will assume that these are the necessary and sufficient subcategories of a more general correlation, the neural correlates of information (NCI). But because the DSM provides no mechanism by which such unconscious processes, and therefore unconscious errors, might be detected, unconscious mental is a black hole in the universe of diagnoses under the current system: no information can escape them.  Neuroscience potentially offers a clear view of the unconscious through visualization of the brain, if only the units of the mind can be settled upon, and technology for detecting them at the necessary resolution developed


The preceeding in mind, we have arrived at an intellectual crisis from which it seems we will fail to escape.  On the one hand we have established the clinical utility of measuring the mind in mathematically tractable scientific units.  On the other hand is the hard problem; there seems no known way to determine which mathematically tractable scientific units we should use.

We propose to solve this crisis by ignoring it. Instead we turn psychiatry’s mesmerized attention to the deductive problem. In doing so, we follow the rest of medicine which has done exactly the same. Moreover, like the rest of medicine, we will solve the deductive problem with deduction – which can be done from an armchair – thus eliminating the need for any new empiricism. Everything psychiatry needs it already has before it.  The trick is to use it.

At the practical level, deduction hinges on language, and the remainder of this article is a long exposition on the remarkable uses we can find for a mere eighteen words, nine from the language of neurobiology and physics, and nine from the language of information theory and cybernetics.  By claiming not that they cause or correlate with one another, but merely define one another, we unleash explanatory power that psychiatry otherwise lacks. Readers are almost certain to object at first to the assertion that we are defining informational words with neurobiological structures. We are not. We are defining informational words with neurobiological words. At the article’s end we will turn to the philosophical problems induced by this maneuver, but for now we ask the reader simply “go with it,” and accept what they Wittgenstein showed and they intuitively know: that words that describe physical objects are not, in fact, physical objects, but are just words. 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

About Peter Freed, M.D.

I am a psychiatrist (psychopharmacology and psychotherapy) specializing in the so-called "personality disorders," particularly narcissistic and borderline personality disorders. I was a Fellow and then an Assistant Professor of Clinical Psychiatry at Columbia from 2004- 2011. I am currently in private practice in NYC.