The Circular Logic Argument

 

I thought I would do a quick post on this because I am interested in rhetoric and this is mind-numbingly simple rhetoric.  It goes like this:

Being depressed means that you have major depression and you are depressed because you have major depression…..

Having motor restlessness means that you have attention-deficit hyperactivity disorder (ADHD) and having ADHD means that you have motor restlessness

This has been presented as though it is an indictment of psychiatric descriptive diagnosis – but you don’t have to think about it too long to figure out why this is a fallacy.  By analogy

Having a cough means that you have COVID and having COVID means that you have a cough.

The circular logic fallacy obviously does not consider the biological complexity of medical and psychiatric diagnoses.  We can rewrite them more accurately using that knowledge.  For example:

Being depressed means that you may have one of hundreds of medical, neurological or psychiatric conditions causing depression or that you may have completely normal mood reactivity or you may have one of thousands of pluralistic causes and having any one of hundreds of medical, neurological and psychiatric conditions or normal mood reactivity or one of thousands of pluralistic causes means that you have depression.

In other words – there is no 1:1 mapping of clinical depression onto the symptom of depression.  The diagnostic process returns a hypothesis about a condition that may be responsible for depressive symptoms.  I hope that illustrates how fallacious this argument is. The problem with rhetorical arguments like this is that they are generally advanced by people who have not gone to psychiatry school or who may have done it but poorly. It is reinforced by business practices and what I would call the necessity of low-quality research.

Starting with the research issue first.  Practically all studies of depression in the literature do not consist of psychiatric diagnoses of depression. Large GWAS studies typically use a ratings scale like the PHQ-9 as the depressive phenotype of interest.  There is no assurance that the patient would be diagnosed with depression by a psychiatrist or have had any of the other thousands of causes of depression considered.  In some of those studies there is a more general diagnostic screen administered to research subjects by non-psychiatrists and if screening criteria are met – the inclusion criteria for the study are met. None of this is assurance that the subjects’ studied would be diagnosed with depression (and not something else) by a psychiatrist.  The low-quality diagnosis in this case is necessitated by massive databases.  For example, the UK Biobank has data on a half million individuals and that would require at least a million hours of interviews by research psychiatrists to make a clinical diagnosis of depression.  That would probably require several hundred full-time psychiatrists working their entire 35 year career to complete.

The business practice of treating depression has similar problems.  It is almost a universal experience today to take anxiety and depression rating scales in primary care clinics.  The primary care experience may be even more crude than the research experience because the PHQ-2 may be administered instead of the PHQ-9.   The PHQ-2 consists of the following 2 ratings over the past 2 weeks:

1:  Little interest or pleasure in doing things

2:  Feeling down, depressed, or hopeless

These screening methods were initiated to show that managed care plans were interested in treating depression.  Since there will never be enough psychiatrists to assess and treat depression, these proxy screenings were felt to be an adequate replacement for psychiatry and they generally result in a diagnosis and treatment of depression even though (once again) there is no guarantee that a psychiatrist would have made that diagnosis.  Just from a purely rhetorical standpoint – it is a syllogistic fallacy to conclude that 1 and 2 above are adequate premises to establish a diagnosis of depression.  The debate at that point may be: “Well the clinician seeing the score will engage in a more elaborate diagnostic interview to make the diagnosis.”  If that is the case – what prevented them from doing that in the first place?  There is an expected paucity of data related to this practice – but I suspect there are many cases of antidepressant overprescription and “treatment resistant depression” based on the wrong diagnosis.  I recently offered to analyze the data from a large health plan for free and they were not interested in looking at it.

The most recent commentary on circular reasoning apparently came from a paper (1) claiming that causal language about psychiatric disorders is the result of a logical error and leads to a confused public and intellectual dishonesty.  The authors make several errors along the way as they develop this argument including:    

Ideally, a medical diagnosis both provides a precise term for a given condition and identifies its etiological mechanism

This is a rhetorical construct that ignores what has been known for decades and that is according to Merskey (2): “Medical classification lacks the rigor of either the telephone directory or the periodic table.  It is exceptionally untidy but it is taken to reflect in some way “the absolute truth” or at least the wonderful truth as it is known to its best practitioners.”  Merskey elaborates on how the medical classification system has several conceptual parameters – most independent of etiological mechanism. In fact, if etiological mechanisms were known – all categories would be mutually exclusive and that is another property that does not exist in medical classifications.  The medical terms "diagnosis" and "disease" are anything but precise and that leaves them open to attack by anyone providing a restricted definition.

“By contrast, diagnostic categories in psychiatry are currently defined only by symptoms.”

The DSM classification has a significant number of disorders where the precipitating and etiological factors are known. The hundreds of causes of organic mental disorders are a case in point as well as an entire section of neurocognitive disorders where the pathology is at least as precise as examples that the authors give.  There is a universe of medical and neurological disorders that are polygenic quantitative disorders with no specific etiology like psychiatric disorders.  Psychiatric disorders are also comprised of clear reproducible signs including sleep and appetite disorders and motor disorders that produce measurable results.  

“While it would be entirely correct to say that the human experiences that the diagnostic criteria describe can feel like an illness, it is different from claiming that an identified external biomedical pathological entity is really causing the symptoms.”

The authors trivialize depression as a mere feeling. I have never seen a person who came in for an assessment based on a mere feeling. They are typically experiencing a disruption in many aspects of their life and have difficulty functioning on a day-to-day basis.  Most patients seeing psychiatrists also have considerable medical comorbidity.

“By contrast, psychiatric diagnoses are not conceptually independent of their respective symptom lists.”

 The authors contrast psychiatric disorders and their symptoms with a lung tumor and a cough and suggest that because psychiatric diagnoses “cannot exist” without symptoms and this is proof that a purely descriptive syndrome cannot be a “cause” of the symptoms.  They also make the error in suggesting that a person must “meet criteria” for depression to be diagnosed with depression.  The problem is that depression, mania, and psychosis existed for centuries before there was a DSM.  These conditions existed long before there were psychiatrists.  They are obvious to non-psychiatrists (the authors apparently excepted).  The only reason psychiatry exists today is to treat syndromes that have been systematically observed and recorded by both psychiatrists and non-psychiatrists.  The medical side of things is described well by DeGowin and DeGowin in their physical diagnosis text (3).

"For several thousand years physicians have recorded observations and studies about their patients.  In the accumulating facts they have recognized patterns of disordered bodily functions and structures as well as forms of mental aberration.  When such categories were sufficiently distinctive, they were termed diseases and given specific names.”

To this day – medical practice is largely based on recognizing sufficiently distinctive categories and not pathophysiology.  There is always a lot of speculative pathophysiology and I have witnessed that all changing over the course of my career.  The pathophysiology learned in medical school – even if based on Nobel Prize work – is not the pathophysiology that applies today.  These diagnoses were independent of symptom lists for centuries and to this day they still are - in that no experienced psychiatrist is treating depression based on symptom lists or “meeting criteria”.

“Some authors therefore emphasize that depression can be described as an adaptive response or a functional signal to adverse circumstances.  Contrary to the erroneous causal beliefs that circular claims promote, this approach underlines that low mood and/or loss of pleasure are often meaningful reactions to life events, and that they can be meaningfully understood.”

This is a potentially erroneous causal belief and the authors apparently have no problem with circularity in this case or the potential lack of investigation of associated causes.  They also seem to misunderstand the idea that to have a disorder – there has to be some form of altered functioning beyond what would be expected.  Most people have that knowledge.  This is also a naive statement from the perspective of assessment and treatment of suicide risk. Can suicidal thinking associated with loss be explained away as a “meaningful reaction to life events” or does something more definitive need to be done?  Before anyone dismisses the idea as rhetorical - some of these same authors have suggested that psychosis is an adaptive response.  Finally – they include a quote from authors on the adaptive response theory as if psychiatrists have not been involved in theories, clinical observations, and developing therapies of these phenomena for decades (4-11).  

Rather than continuing a point-by-point analysis – a look at the rhetoric is probably a better summary.  From the diagram, the authors argue using a typical biomedical psychiatry conflation combined with controlling the premise. The top of the diagram illustrates that when all of psychiatry (in this case depression) is condensed or conflated into a monolithic nondescript biomedical model  - it is easy to demonstrate not only circularity but also how clueless psychiatrists are.  This should come as a surprise to no psychiatrist since this is really a longstanding rhetorical approach to the deconstructive criticism of the field.

A more realistic assessment can be seen in the lower graphic. I labelled it clinical depression since in this case the authors’ use of biomedical psychiatry is largely pejorative.  Every psychiatrist I worked with in acute care would not consider it to be a problem – since we were confronted with hundreds of conditions that had depressive symptoms that we had to figure out.  We were good at it and looked forward to it.  The emphasis is on multiple etiologies.  Numerous psychiatric disorders have depressive symptoms as well as medical and neurological disorders that psychiatrists need to be able to diagnose.  There are known biological causes as noted in the DSM, but many psychiatric disorders are complex polygenic disorders with no specific etiology.  With rule out diagnoses – that means that depression can cause depressive symptoms that can be addressed at the pluralistic level.  The authors suggest that “guild issues” may be a reason that biomedical psychiatry is defended as causal of depressive symptoms. Psychiatry in fact has produced a solid literature (4-11) of various etiologies of depression and how to treat them that easily encompass the authors’ suggestion that meaningful events may have a role to play. That theme has been present in psychiatry for decades prior to this paper.

Anyone reading a paper like this one needs to have an awareness of biology and human biology as a subset.  As I tried to point out in previous posts – for many reasons biological classifications will be imperfect.  That is true for biology without human constraints like speciation in all living organisms.  It is also true for disease classifications and I hope to have more on this soon. Any argument that there exists a standard for categories, diagnoses, or disorders in medicine or psychiatry that is perfect or even unidimensional should be considered rhetorical.

 

George Dawson, MD, DFAPA

 

1:  Kajanoja J, Valtonen J. A Descriptive Diagnosis or a Causal Explanation? Accuracy of Depictions of Depression on Authoritative Health Organization Websites. Psychopathology. 2024 Jun 12:1-10. doi: 10.1159/000538458. Epub ahead of print. PMID: 38865990.

2:  Merskey H. The taxonomy of pain. Med Clin North Am. 2007 Jan;91(1):13-20, vii. doi: 10.1016/j.mcna.2006.10.009. PMID: 17164101.

3:  DeGowin, EL, DeGowin, RL. Bedside Diagnostic Examination. United Kingdom: Macmillan, 1976.

4:  Sifenos PE.  Short-term Dynamic Psychotherapy.  New York.  Plenum Medical Book Company, 1979.

5:  Klerman GL, Weissman MM, Rounsaville BJ, Chevron ES.  Interpersonal Psychotherapy of Depression, New York: Basic Books, 1984.

6:  Yalom ID.  Existential Psychotherapy.  New York: Basic Books, 1980.

7:  Beck AT, Rush JA, Shaw BF, Emery G.  Cognitive Therapy of Depression.  New York: Guilford Press, 1979.

8:  Bennett D.  Social and community approaches.  In:  Paykel ES (ed).  Handbook of Affective Disorders.  New York:  Guilford Press, 1982:  pp.  346-357.

9:  Arieti S.  Individual psychotherapy.  In: Paykel ES (ed).  Handbook of Affective Disorders.  New York:  Guilford Press, 1982:  pp.  298-305.

10:  Stein A.  Group therapy.  Paykel ES (ed).  Handbook of Affective Disorders.  New York:  Guilford Press, 1982:  pp.  307-317.

11:  Viederman M. The psychodynamic life narrative: a psychotherapeutic intervention useful in crisis situations. Psychiatry. 1983 Aug;46(3):236-46.

Explanatory Note:  When I use the terms psychiatric, neurological, and medical diagnoses - I am referring to medical as including all internal medicine specialties (Infectious Disease, Endocrinology, Nephrology, Cardiology, Rheumatology, Allergy and Immunology) as well as general Internal Medicine and Family Medicine.  Neurology and Psychiatry generally have non-overlapping conditions but there is a considerable amount of comorbidity from the medical fields.