Use of Logic-Based Therapy to Encode Emotional Reasoning on the Ventromedial Prefrontal Cortex

Print Friendly, PDF & Email
Elliot D. Cohen, PhD


The Ventromedial Prefrontal Cortex (vmPFC) of the brain has been linked to value-based decision-making and regulation of negative emotions. Further, one study of the vmPFC utilizing functional magnetic resonance imaging (fMRI) has helped to refine the nature of the reasoning process by which such activities as value-based decision-making proceed, especially in contexts involving negative emotions (Goel & Dolan, 2003).  Such research suggests that a cognitive-behavioral therapeutic approach that identifies the logical structure of patients’ reasoning involving emotional content may be more aligned with the way the vmPFC operates in performing the latter functions than an approach that does not directly identify such reasoning structures.    Accordingly, this article suggests that a form of Rational-Emotive Behavior Therapy (REBT) known as Logic-Based Therapy (LBT), which identifies patients’ emotional reasoning dispositions, may be more suitable as a cognitive-behavioral approach than traditional REBT, which identifies the causes of negative emotions. The article explores the implications of the latter hypothesis for the treatment of Posttraumatic Stress Disorder (PTSD).

Logic-Based Versus Traditional REBT

LBT uses logic to identify patients’ dispositional, irrational emotional reasoning in order to change these irrational dispositions into rational ones (Cohen, 1987, 1992, 2003, 2006, 2014, 2017).  Whereas traditional REBT utilizes the ABC theory to identify patients’ irrational beliefs (Ellis, 2001, 1991, 1990, 1971; Ellis & Grieger, 1977; Ross, 2006), LBT uses the valid syllogistic inference form of modus ponens (If p then q; p/∴q) to help patients identify their dispositional emotional reasoning, that is, the reasoning patients do when they are experiencing emotions such as anxiety, depression, anger, and depression (Cohen, 2016).  

The ABC Model

The ABC model proceeds in terms of three psychological points (Ellis, 1971). For example, according to REBT’s ABC Theory, the case of a patient, John, who has lost his job and is depressed because he believes he is a failure, can be represented as follows:

  • Point A: Activating event:  John having lost his job.
  • Point B: Belief system:  I’m a failure
  • Point C: behavioral and emotional Consequence:  Depression

This model provides a causalanalysis according to which A & B jointly cause C (Depression).  As such, REBT identifies the patient’s irrationalbelief(or set of beliefs) at point B that contributes causally to the patient’s self-defeating emotion at point C (Ellis, 2001).

The Logic-Based Model

In contrast, LBT’s model provides a logical analysis in terms of the patient’s inference having emotional content (emotional reasoning):

  • Rule Premise: If I lost my job then I’m a failure
  • Report Premise: I lost my job
  • Conclusion: ∴I’m a failure

The above LBT analysis is intended to capture the patient’s emotional reasoning (Cohen, 2016), that is, the inference the patient is disposed to make when the patient is depressed. The rule premise formulates the inference rule that the patient uses to validate the inference. This conditional rule consists of a statement of the patient’s emotional object (O) as antecedent (“I lost my job”), and a statement of the patient’s rating(R) of this object as consequent (“I’m a failure”). R includes the emotional contentof the inference (for example, John’s reference to himself as “a failure”). This rule is typically assumed rather than explicitly stated or thought by the patient.  The LBT analysis makes this premise explicit (Cohen, 2005).  The second report premise subsumes O under the inference rule. The conclusion, in turn, contains R, as deduced from the two premises (Cohen, 2013). Thus, the explicit form (Cohen, 2016) of this inference is:

  • If O then R
  • O
  • ∴R

LBT therapists identify the patient’s emotional reasoning by finding the patient’s O and R components and insert them into the explicit form; they then help the patient identify irrational premises (for example, self-damnation in the rule premise).

The Ventromedial Prefrontal Cortex and Emotional Reasoning

Recent fMRI, lesion, and electrophysiological brain studies provide a large body of evidence to show that the ventromedial prefrontal cortex (vmPFC) is linked to value-based decision-making and regulation of negative emotions.  Not only does this evidence show that the vmPFC inhibits negative emotions, there is evidence that it also plays a role in generatingnegative emotions (Hiser & Koenigs, 2018). Further, there is evidence that, when subjects engage in reasoning/inferences having emotional content, ventromedial prefrontal cortices (VmPFC) are active. In one study Goel and Dolan (2003) produced fMRI brain scans of 19 subjects while they engaged in a deductive reasoning activity which required them to determine the logical validity of 60 emotionally charged syllogisms and 60 emotionally neutral syllogisms. An example of each type of syllogism is as follows:

Emotionally Charged Syllogism:
All child molesters are perverse.
Some child molesters are priests.
∴Some priests are perverse.

Neutral Syllogism:
No poisons are sold at the grocers.

Some mushrooms are sold at the grocers.
∴Some mushrooms are not poisonous.

Subjects were asked to rate the emotionality of the syllogisms on a scale ranging from +5 to -5, where 0 = neutral. The scale defined high emotionality as making subjects feel “stimulated,”“tense,” and “excited”; and low emotionality as making subjects feel “relaxed,” “calm,” and “dull.” Based on the examples provided, the syllogisms appear to have included mostly, although not entirely, negatively charged terms—“murderous,” “criminals,” “perverse,” “not innocent,” “pimps,” “handicapped,” “not smart,” “rapists,” and “expendable” (Goel & Dolan, 2003, p. 2316). To distinguish between subjects’ emotional responses that were functions of the inferenceactivity itself rather than simply responding to the emotional content of the syllogism, Goel and Dolan (2003) provided a baseline condition for both emotional and neutral syllogisms in which the conclusion was clearly irrelevant to its premises so that no reasoning was needed to determine the invalidity of the argument.  Examples of such baseline syllogisms are as follows:

Emotionally Charged Baseline Syllogism:

Some wars are not unjustified.
All wars involve raping of women.
∴Some Indians are dishonest.
Neutral Baseline Syllogism:

Some Canadians are not children.
All Canadians are people.
∴Some babies are curious.

Goel and Dolan (2003) found that subjects’ fMRIs in the case of emotionally neutral syllogisms, relative to the neutral baseline, were correlated with blood oxygenation changes in the left dorsolateral prefrontal cortex (dlPFC) whereas fMRIs in the case of emotionally charged syllogisms, relative to the emotional baseline, were correlated with changes in bilateral vmPFC activity. The vmPFC responses to the emotionally charged syllogisms were stronger when the content was rated by subjects as more emotionally charged, whereas the dlPFC responses to the neutral syllogisms were stronger when the content was rated as lessemotionally charged.   The researchers concluded that “[The] VMPFC is engaged by the reasoning processin the presence of emotional saliency” (Goel & Dolan, p. 2318, italics added). Conversely, the dlPFC is engaged by the reasoning process in the case of emotionally neutral syllogisms.  Further fMRI study also suggests that the dlPFC may be engaged in neutral reasoning involving reappraisal or assessment of emotional reasoning (Winecoff, et al, 2013).
In addition, there is evidence that an intact vmPFC is necessary for the functional capacity to engage in emotional reasoning. In one controlled study, the reasoning capacity of patients with damage to the vmPFC was impaired only in the case of emotional reasoning, but not in the case of reasoning with neutral content (Nicolle & Goel, 2013). Further, there is fMRI evidence showing that during negative emotions such as depression, the vmPFC is active, and inactive when the depression resolves (Koenigs & Grafman, 2009a).  Corroboratively, the vmPFC appears to be a natural locality for emotional reasoning to occur because it has bidirectional connections to major limbic system structures including amygdala and hypothalamus, which are active in emotional responses (Siddiqui et al, 2008).

Significance of Findings for the Efficacy of Logic-Based Therapy Interventions

These findings suggest that, insofar as LBT successfully recreates the patient’s dispositional emotional reasoning, it proximately keys into corresponding dispositional vmPFC inferential activity requiring therapeutic manipulation.  For example, LBT’s construction of John’s emotional reasoning when he is depressed about losing his job would encode corresponding inferential activity conducted on his vmPFC.  More formally, where VmPFC1is the inferential activity conducted on a patient’s vmPFC at time twhen the patient experiences a negative emotion, and ER1is LBT’s construction of the patient’s irrational emotional reasoning at t, then

(VmPFC1àER1) & (ER1àVmPFC1)

where “à” represents encoding of VmPFC1to ER1, and conversely.  As such manipulation of ER1to alternative emotional reasoning that avoids the irrational emotional reasoning can also manipulate VmPFC1:


~ER1[more rational emotional reasoning]

For example, ~ER1might be ER2:
If I lost my job then this can give me an opportunity to finally pursue something else I might like even better.
I lost my job
∴I now have the opportunity to pursue something else I might like even better.

In LBT, cognitive restructuring involves working cognitively and behaviorally with the patient to replace the irrational syllogism with a more rational one (Cohen, 2017, 2013), for example, replacing ER1with ER2so that the disposition for VmPFC1is replaced with a set of neural connections that resolves the depression.

Possible Clinical Evidence

Clinical application of the LBT model often involves having patients engage in emotive imagery where patients are asked to imagine that they are in the actual situation that evokes the emotion under investigation (Cohen, 2016).  For example, John may be asked to imagine how he experiences depression about the loss of his job.  While in this state, based on information that the patient discloses about the O and R elements of the emotional reasoning in question, the therapist attempts a formulation of the patient’s emotional reasoning and asks the patient if the formulation is accurate.  If the patient indicates that the formulation is not accurate, the therapist works further with the patient to revise and reformulate the patient’s emotional reasoning until the patient affirms that the formulation is a match. Almost invariably, this therapist has noted that patients are quite confident about whether a formulation is a match.  Based on the empirical data described above, it may be hypothesized that the match is between the engagement of the patient’s vmPFC and the proposed formulation of the patient’s emotional reasoning.  In other words, encoding the inferential activity conducted on the patient’s vmPFC would, under the described clinical conditions, yield the given formulation.  Further fMRI studies are necessary to confirm this clinical hypothesis.

Implications for Post-Traumatic Stress Disorder (PTSD)

fMRI studies have provided evidence for the hypothesis that PTSD is due to the inability of the vmPFC to inhibit over-activity in the amygdala. Such studies have confirmed that PTSD patients exhibit vmPFC hypoactivity, and hyperactivity in the amygdala (Koenigs & Grafman, 2009b; Hayes et al, 2012).  However, studies of patients with lesions of the vmPFC have also shown that lesions to this area slightly reduce the likelihood of developing PTSD, which is the opposite of what would be expected if PTSD were entirely due to the inability of the VmPFC to regulate amygdala activity (Koenigs & Grafman, 2009b).
As such, it is possible that the correlation of hypoactivity in the vmPFC and hyperactivity in the amygdale, in cases of PTSD, represents a correlation without causation. Along these lines, Koenigs and Grafman (2009b) have proposed a hypothesis about the causal role of the vmPFC in PTSD that is supported by lesion studies. According to the latter researchers, the role of the vmPFC in PTSD may be related to its functional role in self-insight and self-reflection.  For example, patients with vmPFC lesions experience diminished levels of cognitive-affective symptoms such as self-damnation and guilt.  The general hypothesis appears to be that such patients are more disposed to having their vmPFC hijacked by their amygdala as a result of suffering traumas.  However, as stated previously, there is also evidence to show that the vmPFC is also involved in generatingnegative emotions as well as inhibiting them (Hiser & Koenigs, 2018). Thus, it is possible that a non-functional vmPFC no longer plays a role in generating certain negative emotions associated with PTSD such as guilt, shame, and empathy. This, in turn, may explain why at least some patients with lesions of the vmPFC are less likely to develop PTSD.
The involvement of the vmPFC in generating negative emotions associated with PTSD may also explain why the use of cognitive-behavioral interventions can be helpful in addressing PTSD (American Psychological Association, 2017) because such interventions aim at changing the cognitions and behavior that sustain these emotions. Further, the possibility of encoding vmPFC inferential activity into the emotional reasoninginvolved in negative emotions suggests that LBT may be a preferred mode of cognitive intervention for patients with PTSD.  For example, consider a veteran who suffers moral injury as a result of a traumatic event suffered during active duty in which the patient has come to feel guilty about the deaths of others.  Let’s say her emotional reasoning is ER3:

If the others were killed in the explosion, then I deserved to die too.
I was the only one who survived; the others were killed when the device exploded.
∴I deserved to die too.

Notice that the major premise also entails that the others who died in the explosion also deserved to die, which is absurd insofar as no one deserved to die.  In such a case the therapist could help the patient refute the troublesome emotional reasoning and replace it with a more rational syllogism such as ER4:

If I didn’t deserve to die in the explosion, then I shouldn’t condemn myself for surviving.
I didn’t deserve to die in the explosion
∴I shouldn’t condemn myself for surviving.

Where ER4encodes corresponding vmPFC inference activity, replacing dispositional ER3with dispositional ER4could help the patient overcome PTSD symptoms related to the former disposition, for example, that of avoiding contexts associated with the traumatic event. Such replacement would be a function of helping the patient to first identify ER3; identify its self-damning major premise; refute this premise (show that it is irrational); construct antidotal ER4; and assign cognitive-behavioral activities (keeping track of emotional reasoning, shame attacking, behavioral assignments, bibliotherapy, rational-emotive imagery, exposure therapy, etc.) aimed at reinforcing the patient’s behavioral and emotional acceptance of ER4(Cohen, 2013, 2016).


There are two related issues that require investigation. First, whereas LBT formulates the patient’s emotional reasoning using conditional syllogisms (modus ponens), the study by Goel & Dolan (2003) used more complex categorical syllogisms. (The latter sort of syllogisms relate classes of things such as “humans,” “mortals,” and “men” using quantifiers such as “all,” “some,” and “no” instead of logical connectives such as “if then,” “and” and “or.”)  For instance, in a meta-study PradoChadha, and Booth(2011) found that different portions of the brain are more consistently activated by categorical syllogisms than by conditional ones  However,Noveck, Goel and Smith(2004) confirmed that more basic inference forms, particularly modus ponens, which is used by LBT, activate the same brain regions that were studied by Goel and Dolan (2003). Further confirmation about the role of conditional arguments using emotional content is provided by Marling (2015) studying the effect of brain lesions on conditional reasoning in traumatic brain injury patients. In this study, subjects with lesions in their left prefrontal cortex were found to be impaired in emotionally charged conditional reasoning. This further supports a meta-analysis of neuroimaging studies conducted by Goel (2007) according to which conditional arguments are primarily associated with left-lateralized activation in the parietal and frontal lobes.
Second, LBT formulates the suppressed major premise of syllogisms whereas confirmation is required that this premise is represented in the neural pattern that corresponds to the syllogism on the vmPFC. Since the major premise (If p then q) of a modus ponens inference is necessary as part of the validity of the syllogism it is reasonable to suppose that this premise is somehow contained in the neural circuits that represent a modus ponens inference.  Otherwise it is difficult to understand how the brain processes basic inference forms such as modus ponens. Further imaging studies are needed to confirm this assumption.


The emotional reasoning studies discussed above provide substantial evidence in favor of treating patients with emotional problems by utilizing a modality that identifies and examines patients’ emotional reasoning rather than simply looking for patients’ beliefs that may be disturbing them.  This is because, on the proposed hypothesis, the former approach encodes the actual inferential brain processes, and pinpoints the premises from which an emotively charged conclusion is inferred.  It catches the brain in the act of inferringsuch a conclusion (for example, “I’m a failure”) from a set of premises (“If I lost my job then I’m a failure” and “I lost my job”) via modus ponens.  In contrast, traditional REBT does not have a consistent mechanism for encoding inferential brain processes, and may only capture a part of an inference such as the conclusion (“I’m a failure”). Inasmuch as related types of cognitive-behavioral approaches, such as Cognitive-Behavior Therapy (CBT), utilize the ABC theory (Albert Ellis’ ABC model, 2018), the same may be said of these approaches.
The evidence assembled here points to the need for further investigation.  This may take the form of further imaging studies that comparatively examine vmPFC and concomitant limbic system activities in subjects exposed to LBT and traditional REBT. Suggestions for undertaking further study are respectfully welcome.


Albert Ellis’ ABC model in the cognitive behavioral therapy spotlight. (2018. In Positive Psychology Program. Retrieved from
American Psychological Association. (2017). Cognitive behavioral therapy (CBT). Clinical Practice Guidelines for the Treatment of Post-Traumatic Stress Disorder. Retrieved from
Cohen, E.D. (1987). The use of syllogism in rational‐emotive therapy. Journal of Counseling & Development.66(1) 37-39.
Cohen, E.D. (1992). Syllogizing ret: Applying formal logic in rational-emotive therapy. Journal of Rational-Emotive and Cognitive-Behavior Therapy. 10(4), 235-252.
Cohen, E.D. (2003). Philosophical principles of logic-based therapy. Practical Philosophy: Journal of the Society for Philosophy in Practice, 6(2), 27-35. Retrieved from
Cohen, E.D. (2005). Criticalthinking,notheadshrinking.In P.Raabe(Ed.), PhilosophicalCounselingandtheUnconscious(156-166). Amherst, NY: TriviumPublications.
Cohen, E.D. (2006). Logic-based therapy: The new philosophical frontier for REBT. REBT Network. Retrieved from
Cohen, E.D. (2013).Theory and practice of logic-based therapyIntegrating critical thinking and philosophy into psychotherapy. New Castle Upon Tyne, UK:  Cambridge Scholars Publishing.
Cohen, E.D. (2014). Logic-based therapy to go. Psychology Today.  Retrieved from
Cohen, E.D. (2016). Logic-based therapy and everyday emotions: A case-based approach.Lanham, MD:  Lexington Books.
Ellis, A. (1971). Emotional disturbance and its treatment in a nutshell. Canadian Counsellor, 5(3), 168-170. <ahref=””>
Ellis, A., & Grieger, R. (1977). RET:  Handbook of rational-emotive therapy. New York: Springer Publishing.
Ellis, A. (1990). Rational and irrational beliefs in counselling psychology. Journal of Rational-Emotive & Cognitive-Behavior Therapy, 8(4), 221–233.
Ellis, A. (1991). The revised ABC’s of rational-emotive therapy (RET). Journal of Rational-Emotive & Cognitive-Behavior Therapy, 9(3), 139-172.
Ellis, A. (2001).  Feeling better, getting better, staying better: Profound self-help therapy for your emotions.Atascadero, CA: Impact Publishers.
Goel, V. & Dolan, R. J. (2003).  Reciprocal neural response within lateral and ventral medial prefrontal cortex during hot and cold reasoning NeuroImage 20 (2003) 2314–2321
Goel, V. (2007). Anatomy of deductive reasoning. Trends in Cognitive Sciences, 11(10), 435-441.
Hayes, J.P., Hayes, S.M., & Mikedis, A.M. (2012). Quantitative meta-analysis of neural activity in posttraumatic stress disorder. Biol Mood Anxiety Disord. 2012; 2: 9. Published online 2012.
Hiser, J.S., & Koenigs, M. (2018) The multifaceted role of ventromedial prefrontal cortex in emotion, decision-making, social cognition, and psychopathology. Biological Psychiatry, 83, 638-647.
Koenigs, M., & Grafman, J. (2009a). The functional neuroanatomy of depression: Distinct roles for ventromedial and dorsolateral prefrontal cortex Behavioural Brain Research. 2009 Aug 12; 201(2): 239–243.
Koenigs, M., & Grafman, J. (2009b). Post-traumatic stress disorder: The role of medial prefrontal cortex and amygdala. Neuroscientist. 2009 Oct; 15(5): 540–548.
Marling, M.R.R. (2015). Left hemisphere lesions differentially impact conditional reasoning with familiar and emotional content (Unpublished master’s thesis).  York University, Toronto, CA. Retrieved from
Nicolle, A. & Goel, V. (2013). What is the role of ventromedial prefrontal cortex in emotional influences on reason? In I. Blanchette (Ed.), Emotion and Reasoning. Hove, UK: Psychology Press.
Noveck, I.A., Goel,V., & Smith, K.W.(2004). The neural basis of conditional reasoning with arbitrary content.Cortex, 40(4), 613- 622.
Prado, J., Chadha, A. & Booth, J.R. (2011). The brain network for deductive reasoning: A quantitative meta-analysis of 28 neuroimaging studies. Journal of Cognitive Neuroscience. 23(11): 3483–3497.!po=34.9624
Ross, W. (2006). What is REBT?  REBT Network.
Veqar Siddiqui, S, Chatterjee, U., Kumar D., Siddiqui, A, & Goyal, N. (2008). Neuropsychology of prefrontal cortex. Indian Journal of Psychiatry. 50(3): 202–208.
Winecoff, A., Clithero, J.A., Carter, R.M., Bergman, S.R., Wang, L. & Huette, S.A. (2013). Ventromedial prefrontal cortex encodes emotional value. J Neuroscience. 33(27): 11032–11039.

Elliot D. Cohen, Ph.D.(Brown University), is Director of the Logic-Based Therapy Institute, and Executive Director of the National Philosophical Counseling Association (NPCA).  He is the inventor of Logic-Based Therapy (LBT), which he began to develop in 1985 under the auspices of his mentor, Albert Ellis.  He has conducted training workshops on LBT widely, most recently in Taiwan, and regularly conducts a six-week distance training program. His books and articles have been translated into many languages including Chinese, Korean, German, and Italian.  Among his most recent books are Logic-Based Therapy and Everyday Emotions (Lexington Books, 2016), and Counseling Ethics for the 21st Century (Sage, 2018).  Dr. Cohen also writes a blog on Psychology Todaycalled, “What Would Aristotle Do?”Email: