The Sentence Every CBT Therapist Repeats Without Knowing It
I recorded a podcas tinterview with Donald Robertson (see above), the cognitive-behavioral psychotherapist who has spent two decades arguing that Stoicism is the unacknowledged parent of modern CBT. Robertson (2019) opened the case with a clinical detail most therapists have never heard. Albert Ellis, in the 1950s, used to hand patients a single line from Epictetus during the orientation phase of treatment and build the entire ABC model around it.
“He said, read this quote from Epictetus, one of the main stoic teachers, it’s passage five of the Enchiridion, a handbook of Epictetus that says, it’s not things or events that upset us, but our opinions about them. And Ellis taught that, because it encapsulated what he called the ABC model.”`
Read that again. The cognitive model of emotion, the load-bearing assumption of every CBT manual on the market, was lifted intact from a Greek slave who died in 135 CE. Beck (1976) was equally explicit when he traced the philosophical origins of cognitive therapy to the Stoic school. Robertson (2019) documents both attributions in detail in his Philosophy of Cognitive-Behavioural Therapy.
The mechanism is the same one our patients are running every minute of every session. Stimulus arrives, appraisal fires, autonomic state shifts, behavioral output follows, downstream consequence locks the loop in place. Epictetus interrupts the loop at the appraisal step. Beck interrupted it at the appraisal step. The technology is identical. The licensure paperwork is different.
What follows is an argument I have been making in clinical consultation for the better part of a year, and which Robertson sharpened across two hours of conversation. The cognitive model has a documented prior author. Modern psychotherapy has systematically forgotten the prior author. The forgetting costs us decades of clinical refinement, and I can name the specific decades.
How a Therapist-Friendly Philosophy Got Sidelined Anyway
Robertson noticed the asymmetry early. Laypeople found Stoicism. Therapists did not.
“When I first started getting into stoicism, it was really a nerdy thing. And I didn’t have any little stoic friends at the time… weirdly, they seem to have kind of lagged behind a little bit the wave of popular interest in stoicism, which kind of surprised me.”
Two explanations sit on top of each other. Robertson (2019) attributes the first to clinician reading budgets. CBT practitioners spend their professional hours inside outcome research and treatment manuals, leaving the philosophical literature to specialists. The second is linguistic. The lowercase English adjective “stoic” means emotionally shut down, which reads as the opposite of what therapy is supposed to deliver. Most clinicians I have consulted with carry that conflation without noticing.
The conflation does material damage. Robertson made the corrective explicit in our conversation:
“What they call a value, which is an intrinsic quality of action and character, seems to me a weirdly convoluted way of just referring to what any student of philosophy would immediately recognize is arete or virtue… They’ve kind of reintroduced it as it were. Right now it’s like a hot topic again in psychotherapy.”
Hayes, Strosahl, and Wilson (2012) built acceptance and commitment therapy around values clarification as a third-wave innovation. The construct of values in ACT is arete in Stoicism. The framework was waiting on the shelf for two millennia, and Hayes rediscovered it under a new branding cycle without ever having to credit the source.
I will name the pattern at the center of this essay. I call it the Citation Amnesia Problem: a discipline that re-derives its own techniques every two generations because its institutional memory only reaches back as far as the most recent peer-reviewed manual. The pattern is not aesthetic. It is iatrogenic.
Why “CBT Already Does This” Misses What Got Left Behind
The standard pushback when I introduce Stoic frames in clinical consultation runs like this: the relevant techniques are already inside CBT, so the philosophical scaffolding is decoration. Robertson conceded that Ellis and Beck imported pieces of Stoicism into the clinic. He also named what they left out.
“There’s a difference between a bunch of techniques, or therapy is more than that, but it’s kind of moving in that direction and generalizing the whole thing so it becomes part of your identity and way of life… nowadays we’re more interested in prevention rather than cure. Resilience building is kind of the holy grail of psychotherapy… it’s dawning on people that if you want to do that, then you need to re-conceptualize the whole thing more in terms of a philosophy of life rather than a group of techniques or an intervention.”
Stoicism delivers cognitive technology inside a coherent meaning structure. Modern CBT, in its lean evidence-based forms, delivers the same technology inside a fifty-minute reimbursement code. The patient learns to dispute automatic thoughts, scores lower on the PHQ-9 at termination, and drifts back to baseline at six months because the technique was never embedded in anything that asks why a person should bother regulating affect at all.
Robertson framed it in evolutionary terms without quite using the word. Stoicism was designed as resilience training for a population that expected catastrophe as a baseline.
How Behavioral Activation Got Lost for Forty Years
The pattern of rediscovery has a specific, traceable case study. Beck, Rush, Shaw, and Emery (1979) published the standard treatment manual for depression with three major components stacked on top of each other: activity scheduling, automatic-thought disputation, and core-belief work targeting underlying schemas. Jacobson et al. (1996) ran the canonical dismantling study, randomizing 150 outpatients with major depression to the full Beck protocol, the protocol minus schema work, or behavioral activation in isolation. Across termination and six-month follow-up, Jacobson and colleagues found no advantage for the multi-component treatment over behavioral activation alone.
Dimidjian et al. (2006) replicated and extended the finding in a four-arm trial. Behavioral activation now sits in its own evidence base as a first-line treatment for major depressive disorder. Lewinsohn (1974) developed the underlying reinforcement theory of depression years before Beck’s manual appeared in 1979.
Robertson surfaced the lineage problem this exposes, and his tone shifted when he said it:
“Beck’s seminal book on cognitive therapy, nowhere in that entire book does he mention Peter Lewinson, who developed behavior therapy for clinical depression a few years earlier… We already had like, you know, this was effective in itself. What Beck added to it works, but it didn’t really add anything to the efficacy. And so that kind of blinded us for decades to thinking, CBT for depression is this multi-component treatment that combines activity scheduling and cognitive disputation… We could have had half a century of ongoing research in psychotherapy.”
Forty years.
Patients with major depression received the full Beck package across that span, including the cognitive components that Jacobson (1996) eventually showed contributed no incremental benefit. Researchers who could have spent the same four decades refining the active ingredient spent them tracking redundant scaffolding. The mechanism that produced the drift was citation failure. Beck did not credit Lewinsohn. The field assumed the package was load-bearing in all its parts. The Citation Amnesia Problem then closed the loop.
Robertson made the same point about Dubois (1909), the Swiss psychiatrist whose Psychic Treatment of Nervous Disorders described what we would now call cognitive psychotherapy. Dubois prescribed Seneca’s letters as bibliotherapy. He ran what we would now call cognitive disputation. The field absorbed Freud instead, forgot Dubois entirely, and then rediscovered the same territory under Ellis and Beck in the 1950s and 1960s. Robertson dates the loss at half a century.
The Patient Who Cannot Stop Worrying, and the Sentence That Was Already There
Picture a patient with generalized anxiety disorder. She presents with chronic perseverative worry, sleep-onset insomnia, somatic tension across the shoulders and jaw, and a metacognitive conviction that her worry is uncontrollable and dangerous. She has tried thought-stopping. She has tried distraction. She arrives in the consulting room expecting another technique to fail.
Wells (2009), the developer of metacognitive therapy at the University of Manchester, opens treatment with an assessment move that should look familiar to anyone who has read the Enchiridion. He asks the patient to rate, from zero to one hundred percent, how much she believes she has voluntary control over the worrying itself. Patients with GAD score near zero on this rating. Wells then disproves the belief experimentally, through detached mindfulness drills and stimulus-control postponement, until the patient discovers that perseverative thinking is a high-level cognitive process she can interrupt.
Robertson made an observation in our interview that I have not been able to stop thinking about:
“Clients will say, I can’t control my worry… we can prove to them that actually they can take control over it. Like it’s a high level strategic cognitive process that you can actually, to some extent, stop doing or detach yourself from… people massively underestimate how much control they have over some of the things that are going on in their head. But equally, they’ll be trying to ignore, suppress, block out, automatically thoughts and feelings that they should just let be.”
Epictetus, operationalized. The opening line of the Enchiridion partitions phenomena into what is up to us and what is not (Epictetus, ca. 125 CE/2018). My patient with GAD has the partition exactly backwards. She is expending voluntary effort on phenomena outside voluntary control (the intrusive image, the heart rate, the future outcome) while disclaiming voluntary control over the elaborative process that she can in fact interrupt (the rumination chain itself). The treatment task is to flip the partition.
What Robertson said next deserves a slow read:
“That’s so simple. You could write the instructions for it in the back of a business card… You can get them to evaluate their progress in applying it. It’s only really one basic strategy that they have to employ. It’s eminently testable, right?”
The dichotomy of control is testable. Wells (2009) built metacognitive therapy around a structurally identical move and accumulated outcome data competitive with standard CBT for GAD. Nobody has yet isolated the dichotomy of control as a standalone intervention and run a dismantling study against multi-component CBT. Whether Epictetus-as-protocol would outperform Beck-as-package remains an open empirical question. The Citation Amnesia Problem closed it before anyone thought to ask.
Pseudo-Passions and the Polyvagal Patient
The objection I hear most when I import Stoic frames into trauma work is that the philosophy demands emotional suppression. The textual record contradicts that reading. Seneca, in De Ira, described what he called propatheiai, proto-passions: involuntary first movements that precede cognitive appraisal. His example was the reflexive blink when a finger approaches the eye. Robertson summarized the construct exactly as it appears in the Stoic literature:
“Seneca says, it’s kind of like, his example is interesting. He says, it’s like if someone went to poke their finger in your eye and you blinked. He said, we have emotions that are like that… But the stoics, and even Beck, would say that the important thing is what happens next.”
The mechanism Seneca described in the first century maps almost perfectly onto contemporary affective neuroscience. LeDoux (2015) describes a rapid subcortical thalamo-amygdalar pathway that activates autonomic reactivity before cortical appraisal completes. Porges (2011), in the polyvagal model, describes neuroception, an unconscious detection process that recruits autonomic state before any deliberate cognition arrives. Restate Seneca’s position in modern terms and you get the operating assumption of every credible trauma-informed clinician working today.
Robertson told the story of an unnamed Stoic teacher caught in a storm at sea, recorded by Aulus Gellius. The teacher turned pale and shook with the rest of the passengers. After the ship reached harbor, Gellius confronted him. The teacher pulled a lost book of Epictetus from his satchel and explained that the autonomic reaction was shared with animals and seasoned sailors alike. The Stoic distinction was not the absence of the proto-passion. The Stoic distinction was the refusal to add rumination on top of it once the storm had passed.
Two layers of intervention follow from this. The first-movement layer (the autonomic surge, the somatic intrusion, the reflexive startle) requires bottom-up work. Shapiro (2018) built EMDR around exactly this layer, using bilateral stimulation to uncouple somatic and emotional content from the source memory through the adaptive information processing mechanism. The second-movement layer (the interpretation, the elaborative meaning the patient assigns to the reactivation) requires cognitive work. Stoic-informed appraisal exercises target this layer directly.
Conflating the two layers is a clinical error in both directions. I have watched cognitive-only protocols fail trauma patients because they pushed appraisal work onto first-movement reactivations the patient could not yet interrupt. I have also watched somatic-only protocols stall because they treated the second-movement appraisal as a fixed property of the body rather than a modifiable cognitive elaboration. The Stoics had the two-layer model two thousand years ago. The trauma field is still arguing about which layer to start with.
What Stoicism Was Originally For
Every clinical observation in this essay points back to one functional question. What did Stoicism, as a cultural artifact, originally do?
Robertson’s answer reframes the entire renaissance of the philosophy. Stoicism functioned as a daily resilience-training regimen for a population that expected catastrophe as a baseline condition. Marcus Aurelius (ca. 170 CE/2002) wrote the Meditations during the Antonine plague, which killed an estimated five million people across the empire, while simultaneously prosecuting border wars and surviving a civil war launched by Avidius Cassius, one of his most senior generals.
“Marcus must have felt his own mortality all the time, like he was surrounded by death. And so when in that book, when he talks over and over about coming to terms with his own mortality, it’s not a kind of abstract scholarly thing… he watched many of his friends die. He lost half of his children. He had about 12 or 14 children and about half of them died before he did.”
The Stoic regimen was prophylactic. The view from above, premeditatio malorum, the daily dichotomy-of-control exercises, the regular contemplation of mortality, were all installed in fair weather so they would operate in foul. Robertson made the prophylactic logic explicit when he addressed how a Stoic would have responded to the COVID-19 pandemic:
“Their main piece of advice would be nothing’s changed… You should have prepared for this in advance… The wise man or the wise women knows that. And they know to prepare for it long in advance.”
Modern Western patients arrive in my consulting room without that scaffolding. The first major narrative violation hits, whether a diagnosis or a divorce or a death, and the patient enters treatment in active crisis with no prepared cognitive infrastructure. We then deliver crisis-only protocols, get partial symptom reduction, and discharge the patient without the resilience training that would have made the next narrative violation survivable. The Stoics would have called this clinical practice incomplete.
Fox (2020) argued that interrupted goal pursuit creates a completion drive that drives perseverative cognition through the Zeigarnik mechanism. Stoic prophylaxis can be read as the inverse operation. The practitioner installs completed cognitive structures (a settled stance on mortality, a settled stance on what is and is not under one’s control, a settled stance on the indifference of external goods) before the disruptive event arrives. When the event arrives, the Zeigarnik signal does not fire as hard because the cognitive container is already there. This is the mechanism I want trauma-prevention researchers to test.
What I Want Therapists to Do With This
Three recommendations follow from the lineage I have laid out, and I want them on the record.
First, read the primary source. Hays (2002) translated the Meditations for Modern Library in the most accessible English available. Hard and Gill (2011) published the most scholarly modern edition through Oxford World’s Classics. Either translation makes visible the through-line that the contemporary psychotherapy curriculum has obscured.
Second, run the dichotomy-of-control intervention as a standalone construct with your patients with generalized anxiety. Wells (2009) has already shown that the metacognitive frame works. The Stoic version sits one layer beneath the metacognitive version, costs nothing to deliver, and can be assessed with a single visual analog scale. Track the data. Publish the data.
Third, stop treating each new third-wave protocol as if it had emerged from the laboratory air. When Hayes et al. (2012) publish on values, ask what arete meant to Chrysippus. When the mindfulness literature cites Kabat-Zinn, look at Perls, Hefferline, and Goodman (1951) for the same exercises under a different name. The Citation Amnesia Problem only operates if we let it.
Read Epictetus. The cognitive model was waiting there the whole time.
References
Beck, A. T. (1976). Cognitive therapy and the emotional disorders. International Universities Press.
Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979). Cognitive therapy of depression. Guilford Press.
Dimidjian, S., Hollon, S. D., Dobson, K. S., Schmaling, K. B., Kohlenberg, R. J., Addis, M. E., Gallop, R., McGlinchey, J. B., Markley, D. K., Gollan, J. K., Atkins, D. C., Dunner, D. L., & Jacobson, N. S. (2006). Randomized trial of behavioral activation, cognitive therapy, and antidepressant medication in the acute treatment of adults with major depression. Journal of Consulting and Clinical Psychology, 74(4), 658–670. https://doi.org/10.1037/0022-006X.74.4.658
Dubois, P. (1909). The psychic treatment of nervous disorders (the psychoneuroses and their moral treatment) (S. E. Jelliffe & W. A. White, Eds. & Trans.). Funk & Wagnalls. (Original work published 1904)
Epictetus. (2018). The Enchiridion (G. Long, Trans.). Dover. (Original work composed ca. 125 CE)
Fox, J. G. (2020). Recovery, interrupted: The Zeigarnik effect in EMDR therapy and the adaptive information processing model. Journal of EMDR Practice and Research.
Hayes, S. C., Strosahl, K. D., & Wilson, K. G. (2012). Acceptance and commitment therapy: The process and practice of mindful change (2nd ed.). Guilford Press.
Jacobson, N. S., Dobson, K. S., Truax, P. A., Addis, M. E., Koerner, K., Gollan, J. K., Gortner, E., & Prince, S. E. (1996). A component analysis of cognitive–behavioral treatment for depression. Journal of Consulting and Clinical Psychology, 64(2), 295–304. https://doi.org/10.1037/0022-006X.64.2.295
LeDoux, J. (2015). Anxious: Using the brain to understand and treat fear and anxiety. Viking.
Lewinsohn, P. M. (1974). A behavioral approach to depression. In R. J. Friedman & M. M. Katz (Eds.), The psychology of depression: Contemporary theory and research (pp. 157–178). Wiley.
Marcus Aurelius. (2002). Meditations (G. Hays, Trans.). Modern Library. (Original work composed ca. 170 CE)
Marcus Aurelius. (2011). Meditations with selected correspondence (R. Hard, Trans.; C. Gill, Intro. & Comm.). Oxford University Press. (Original work composed ca. 170 CE)
Perls, F., Hefferline, R., & Goodman, P. (1994). Gestalt therapy: Excitement and growth in the human personality. Gestalt Journal Press. (Original work published 1951)
Porges, S. W. (2011). The polyvagal theory: Neurophysiological foundations of emotions, attachment, communication, and self-regulation. W. W. Norton.
Robertson, D. (2019). The philosophy of cognitive-behavioural therapy (CBT): Stoic philosophy as rational and cognitive psychotherapy (2nd ed.). Routledge.
Robertson, D. (2019). How to think like a Roman emperor: The Stoic philosophy of Marcus Aurelius. St. Martin’s Press.
Seneca, L. A. (2010). Anger, mercy, revenge (R. A. Kaster & M. C. Nussbaum, Trans.). University of Chicago Press. (Original work composed ca. 45 CE)
Shapiro, F. (2018). Eye movement desensitization and reprocessing (EMDR) therapy: Basic principles, protocols, and procedures (3rd ed.). Guilford Press.
Wells, A. (2009). Metacognitive therapy for anxiety and depression. Guilford Press.
Get full access to Next Level Psychology at psychfox.substack.com/subscribe
