Toronto Alexithymia Scale (TAS-20)

Abstract

The Toronto Alexithymia Scale (TAS-20) represents the gold-standard self-report psychometric instrument designed to assess alexithymia, a multifaceted cognitive-affective construct characterized by pronounced deficits in emotional awareness, cognitive processing of affects, and interpersonal communication. Developed by R. Michael Bagby, James D. A. Parker, and Graeme J. Taylor in 1994 as a refined iteration of the earlier 26-item TAS, the TAS-20 comprises 20 self-administered items organized across three distinct, theoretically coherent, and empirically validated dimensions: Difficulty Identifying Feelings (DIF; 7 items), Difficulty Describing Feelings (DDF; 5 items), and Externally-Oriented Thinking (EOT; 8 items). Each statement is evaluated on a standard 5-point Likert scale ranging from 1 (“Strongly disagree”) to 5 (“Strongly agree”), yielding a total aggregate score spanning from 20 to 100.

Extensive psychometric investigations across clinical and non-clinical cohorts globally demonstrate robust construct validity, high internal consistency (Cronbach’s alpha typically ranging between α = .80 and .86 for the full scale), and sound test-retest reliability (exceeding r = .77 over diverse intervals). Cross-cultural adaptations across dozens of languages have consistently replicated its three-factor architecture through both exploratory and confirmatory factor analyses. Clinically, the scale incorporates standard cut-off thresholds (≤ 51 indicating non-alexithymia, 52 to 60 denoting borderline or possible alexithymia, and ≥ 61 confirming pronounced alexithymia), rendering it indispensable in psychosomatic medicine, clinical psychology, psychiatry, and cognitive neuroscience for investigating somatization, substance abuse disorders, affective dysregulation, and neurodevelopmental conditions.

Keywords

Toronto Alexithymia Scale, TAS-20, alexithymia, emotional processing, emotional awareness, Difficulty Identifying Feelings, Difficulty Describing Feelings, Externally-Oriented Thinking, psychometrics, affective neuroscience, psychosomatic medicine, emotion regulation

Authors

The TAS-20 was developed through an extensive multi-phase empirical validation program spearheaded by prominent psychometricians and clinical psychiatrists in Toronto, Ontario, Canada:

• R. Michael Bagby, Ph.D., ABPP — Professor of Psychology and Psychiatry, Department of Psychology, University of Toronto, and Director of Clinical Research at the Centre for Addiction and Mental Health (CAMH), Toronto, Canada. An internationally distinguished expert in psychological assessment, personality disorders, and clinical psychometrics.
• James D. A. Parker, Ph.D. — Professor of Psychology and former Vice-President of Research and Innovation at Trent University, Peterborough, Ontario, Canada. His seminal scholarship centers on emotional intelligence, psychological assessment, and health psychology.
• Graeme J. Taylor, M.D., FRCPC — Professor Emeritus of Psychiatry, Department of Psychiatry, University of Toronto, and Psychoanalyst at the Toronto Psychoanalytic Society. A leading global authority on psychosomatic illness, affect regulation, and the psychoanalytic conceptualization of alexithymia.

Purpose

The primary purpose of the Toronto Alexithymia Scale (TAS-20) is to provide a reliable, empirically robust, and cost-effective operationalization of the alexithymia construct. Originally conceptualized in the early 1970s within clinical psychosomatic literature, alexithymia literally translates from Greek as “no words for emotions” (a = lack, lexis = word, thymos = emotion). Prior to the publication of the TAS-20, the psychological assessment of affect deficit was severely hindered by idiosyncratic interview protocols, subjective clinical impressions, and psychometrically deficient self-report inventories that conflated emotional deficits with generalized psychological distress, anxiety, or hypochondriasis.

Clinically, the TAS-20 serves to identify individuals whose capacity for cognitive emotional appraisal is fundamentally constrained. In clinical psychiatric and psychological practice, patients presenting with elevated alexithymia frequently fail to utilize standard cognitive-behavioral therapies and insight-oriented psychotherapies because they experience somatic equivalents of affect (such as tension headaches, gastrointestinal distress, tachycardia, or diffuse muscle pain) rather than distinct experiential feeling states. Administering the TAS-20 allows clinicians to tailor treatment modalities, shifting therapeutic focus toward foundational emotional literacy, somatosensory grounding, and mentalization-based interventions.

In biomedical and psychosomatic disciplines, the instrument serves to delineate how blunted emotional processing elevates vulnerability to functional somatic syndromes, autoimmune disorders, hypertension, and inflammatory bowel diseases. From a neurobiological vantage point, high scores on the TAS-20 are associated with altered functioning in the anterior cingulate cortex, insula, and amygdala, establishing the scale as a vital screening mechanism in neuroimaging paradigms investigating the neural correlates of interoceptive and affective processing.

In psychiatric research, the TAS-20 is extensively applied to investigate transdiagnostic vulnerability across major depressive disorder, post-traumatic stress disorder, substance use disorders, eating disorders, and autism spectrum conditions. By dissociating affect recognition deficits from general neurotic distress, the instrument provides researchers with an accurate quantitative index to evaluate whether emotion processing disturbances represent stable trait phenomena or state-dependent artifacts.

Psychological Construct

The construct of alexithymia encompasses a distinct cluster of cognitive, affective, and behavioral characteristics. The TAS-20 operationalizes this multidimensional construct into three distinct yet interrelated subscales, omitting a previously hypothesized fantasy dimension due to psychometric instability and overlap with general imagination:

Factor 1: Difficulty Identifying Feelings (DIF)

Comprising 7 items (Items 1, 3, 6, 7, 9, 13, and 14), the DIF dimension captures an individual’s inability to recognize and distinguish between discrete internal emotional states and the somatic sensations that accompany autonomic arousal. Individuals with high DIF scores experience physiological surges—such as an elevated heart rate, visceral butterflies, or muscle tightness—yet cannot determine whether these manifestations signify fear, sadness, anger, or a medical illness. For example, Item 6 (“When I am upset, I don’t know if I am sad, frightened, or angry”) directly assesses affective differentiation, while Item 3 (“I have physical sensations that even doctors don’t understand”) highlights the misattribution of emotional arousal to physical disease.

Factor 2: Difficulty Describing Feelings (DDF)

Comprising 5 items (Items 2, 4, 11, 12, and 17), the DDF dimension measures deficits in finding verbal labels to express subjective emotional states to others. It reflects a breakdown in the symbolic translation of emotional experience into shared linguistic codes. High scorers experience a persistent verbal blockade when asked how they feel, frequently relying on simplistic, non-specific descriptors such as “bad” or “fine.” Item 2 (“It is difficult for me to find the right words for my feelings”) and Item 17 (“It is difficult for me to reveal my innermost feelings, even to close friends”) exemplify the communication and interpersonal barriers intrinsic to this factor.

Factor 3: Externally-Oriented Thinking (EOT)

Comprising 8 items (Items 5, 8, 10, 15, 16, 18, 19, and 20), the EOT dimension measures a cognitive mode characterized by a preference for focusing on external events, concrete details, and operational activities rather than internal subjective reflections, fantasies, or psychological motivations. Individuals scoring high in EOT demonstrate pragmatic, utilitarian, and mechanically oriented thinking patterns (known in French psychosomatic literature as pensée opératoire). Statements like Item 15 (“I prefer talking to people about their daily activities rather than their feelings”) and Item 20 (“Looking for hidden meanings in movies or plays distracts from their enjoyment”) quantify this disinclination toward introspective or symbolic mental activity.

Theoretical Framework

The theoretical architecture of the TAS-20 is rooted in psychosomatic medicine and modern cognitive-affective neuroscience. The concept of alexithymia emerged through the observations of psychoanalysts Peter Sifneos and John C. Nemiah in the early 1970s. Nemiah and Sifneos observed that classic psychosomatic patients manifested an inability to describe their emotional conflicts verbally, had impoverished dream and fantasy lives, and displayed an unyielding focus on objective, external realities. Concurrently, French psychoanalysts Pierre Marty and Michel de M’Uzan formulated their theory of pensée opératoire (operative thinking), which posited that somatic illnesses arise when psychic energy cannot be elaborated through symbolic affect, discharging directly into somatic pathways.

Taylor, Bagby, and Parker integrated these early psychoanalytic observations into contemporary cognitive processing models, drawing heavily on Allan Schore’s affect regulation theory, Lane and Schwartz’s Cognitive-Developmental Model of Levels of Emotional Awareness, and Bucci’s Dual-Code Multiple Code Theory. According to the Multiple Code Theory, human emotional processing relies on three systems: non-verbal non-symbolic (visceral sensations), non-verbal symbolic (images and fantasies), and verbal symbolic (words). Alexithymia represents a systematic disconnection—a failure of referential translation—between the subcortical, visceral-sensory emotional representations and the cortical, linguistic-symbolic apparatus.

Under this theoretical model, emotional arousal generates primary biological signals via subcortical structures (such as the brainstem and amygdala). In non-alexithymic individuals, these raw sensory inputs are integrated into conscious awareness via the insula and rostral anterior cingulate cortex, where they receive cognitive appraisal, symbolic representation, and contextual verbal labeling within the prefrontal cortex. In individuals with alexithymia, this secondary integration fails: emotional arousal remains unmentalized, experienced predominantly as undifferentiated somatic stress, and is managed through behavioral discharge, compulsive routines, or somatic conversion.

Validity

The construct, convergent, discriminant, and predictive validity of the TAS-20 have been rigorously evaluated in hundreds of empirical studies spanning clinical, community, and university populations.

Construct and Structural Validity

Bagby et al. (1994a, 1994b) originally demonstrated construct validity through extensive item-selection protocols and cross-validation procedures. Across both healthy adult and clinical psychiatric samples, items loaded robustly onto their theoretical factors without aberrant cross-loadings. Subsequent confirmatory factor analytic studies (e.g., Parker, Taylor, & Bagby, 2003) across large multinational cohorts confirmed that the three-factor model consistently demonstrates superior fit compared to unidimensional or two-factor models.

Convergent Validity

The TAS-20 demonstrates robust convergent validity against complementary measures of affect regulation and emotional intelligence. Research reveals high negative correlations with the Bar-On Emotional Quotient Inventory (EQ-i) and the Mayer-Salovey-Caruso Emotional Intelligence Test (MSCEIT), where TAS-20 total and subscale scores correlate inversely (r = -.45 to -.68) with emotional perception and emotional facilitation. Furthermore, the scale correlates moderately to strongly with the Levels of Emotional Awareness Scale (LEAS), an open-ended scenario assessment, validating that individuals with elevated TAS-20 scores demonstrate objective impairments in complex affective vocabulary.

Discriminant Validity

A central psychometric challenge in measuring alexithymia is dissociating it from generalized negative affect, anxiety, and depression. Factor analytic and regression studies have firmly demonstrated that while the TAS-20 correlates moderately with the Beck Depression Inventory (BDI) (r ≈ .35 to .50) and state-trait anxiety measures, alexithymia remains a distinct construct. Longitudinal studies evaluating depressed patients before and after successful somatic or psychological treatments demonstrate that while depressive scores decline significantly, TAS-20 scores exhibit substantial trait stability, confirming that alexithymia is not merely an epiphenomenon of acute dysphoria.

Predictive and Clinical Validity

The TAS-20 has demonstrated high predictive utility across a wide range of somatic and psychiatric conditions. Elevated scores prospectively predict poor response to conventional cognitive behavioral therapy in mood disorders, heightened treatment dropout rates in substance abuse rehabilitation, and significantly elevated healthcare utilization among patients with irritable bowel syndrome, fibromyalgia, and chronic pain conditions.

Reliability

The psychometric evaluation of the TAS-20 demonstrates high internal consistency, strong inter-item correlations, and sustained temporal stability across diverse target populations.

Internal Consistency

In their seminal validation study, Bagby, Parker, and Taylor (1994a) reported a Cronbach’s alpha coefficient of α = .81 for the total scale in a normative sample, and α = .80 in a clinical psychiatric sample. Across subsequent international studies, total scale internal consistency values typically range between .80 and .86. At the subscale level, internal consistency estimates demonstrate differential strength:

• Difficulty Identifying Feelings (DIF): Displays the highest consistency, with alpha coefficients consistently ranging from α = .78 to .84 across clinical and student samples.
• Difficulty Describing Feelings (DDF): Demonstrates moderate to high reliability, with coefficients typically falling between α = .75 and .82.
• Externally-Oriented Thinking (EOT): Demonstrates comparatively lower, though acceptable, internal consistency, typically ranging between α = .66 and .74. This lower internal reliability is widely recognized in psychometric literature and reflects the conceptual breadth of externally oriented cognitive styles.

Test-Retest Stability

The TAS-20 exhibits remarkable test-retest reliability, confirming that it measures a stable psychological trait rather than fluctuating emotional states. In the initial validation cohorts, a 3-week test-retest correlation yielded r = .77 (p < .001) for the total scale. Subsequent extended longitudinal evaluations over 3-month to 12-month intervals have documented stability coefficients between r = .70 and .83 in non-clinical cohorts. Even during clinical trials where primary psychiatric symptomatology (such as depressive episodes) went into remission, test-retest correlations for the TAS-20 remained high (r > .65), reinforcing its temporal invariance.

Factor Analysis

The factorial validity of the TAS-20 has been subject to rigorous empirical scrutiny using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA).

Exploratory Factor Analysis (EFA)

During scale construction, Bagby et al. (1994a) evaluated an initial pool of candidate items extracted from the original 26-item TAS and newly generated items. Principal components analysis followed by oblique and varimax rotations consistently revealed a clear three-factor structure. Criteria for retaining items included: a primary factor loading ≥ .35, minimal cross-loadings (≤ .25 on secondary factors), and conceptual fidelity. The resulting 20 items loaded cleanly into Difficulty Identifying Feelings, Difficulty Describing Feelings, and Externally-Oriented Thinking.

Confirmatory Factor Analysis (CFA)

Subsequent cross-validation studies using structural equation modeling have tested several competing structural models, including:

• A single-factor general alexithymia model;
• A two-factor model combining DIF and DDF against EOT;
• The standard three-factor correlated model;
• A hierarchical bifactor model containing a general alexithymia factor and three orthogonal subfactors.

Across non-clinical student groups, community adult populations, and diverse clinical samples, the standard correlated three-factor model and the bifactor model have consistently outperformed alternative structural configurations. In landmark cross-cultural studies spanning 19 countries and languages (Parker et al., 2003), goodness-of-fit indices demonstrated acceptable to excellent structural replication:

• Comparative Fit Index (CFI): Ranging typically from .90 to .95;
• Tucker-Lewis Index (TLI): Consistently exceeding .90;
• Root Mean Square Error of Approximation (RMSEA): Consistently within the .04 to .06 range with 90% confidence intervals below .08;
• Standardized Root Mean Square Residual (SRMR): Falling below .06 across cohorts.

While the EOT factor occasionally exhibits weaker factor loadings (varying between .30 and .55, compared to .60 to .80 for DIF items), its retention is theoretically and structurally vital for capturing the operative, pragmatic cognitive hallmark of the clinical alexithymia syndrome.

Instrument / Measurement Tool

• Test Type: Standardized Self-Report Psychometric Inventory.
• Format: Pen-and-paper or digital questionnaire; self-administered.
• Number of Items: 20 items.
• Response Scale: 5-point Likert scale:
  • 1 = Strongly disagree
  • 2 = Moderately disagree
  • 3 = Neither agree nor disagree
  • 4 = Moderately agree
  • 5 = Strongly agree
• Subscale Architecture:
  • Factor 1: Difficulty Identifying Feelings (DIF): 7 items (1, 3, 6, 7, 9, 13, 14)
  • Factor 2: Difficulty Describing Feelings (DDF): 5 items (2, 4, 11, 12, 17)
  • Factor 3: Externally-Oriented Thinking (EOT): 8 items (5, 8, 10, 15, 16, 18, 19, 20)
• Scoring and Directionality:
  • Items are scored from 1 to 5.
  • Reverse Scored Items: Items 4, 5, 10, 18, and 19 must be reverse scored prior to summing (i.e., 1 = 5, 2 = 4, 3 = 3, 4 = 2, 5 = 1).
  • Total scale score ranges from 20 to 100, where higher scores indicate higher levels of alexithymia.
• Standard Clinical Cut-Off Scores:
  • ≤ 51: Non-alexithymia (typical emotion processing).
  • 52 – 60: Borderline / intermediate alexithymia.
  • ≥ 61: Alexithymia present (clinically significant impairment in affective awareness).

Permissions & Fee and Test Year

The Toronto Alexithymia Scale (TAS-20) was originally published in 1994. The instrument is a copyrighted, licensed psychological instrument. The copyright is held by R. Michael Bagby, James D. A. Parker, and Graeme J. Taylor. Commercial use, inclusion in fee-for-service software platforms, translation into unvalidated languages, or pharmaceutical trials require explicit written authorization and licensing agreements. Academic researchers and qualified clinicians may typically obtain permission for scholarly, non-profit empirical investigations. Prospective users should consult authorized test distributors or directly contact the primary developers to secure official licensing before administrative implementation.

References

• Bagby, R. M., Parker, J. D. A., & Taylor, G. J. (1994a). The twenty-item Toronto Alexithymia Scale—I. Item selection and cross-validation of the factor structure. Journal of Psychosomatic Research, 38(1), 23–32. https://doi.org/10.1016/0022-3999(94)90005-1
• Bagby, R. M., Taylor, G. J., & Parker, J. D. A. (1994b). The twenty-item Toronto Alexithymia Scale—II. Convergent, discriminant, and concurrent validity. Journal of Psychosomatic Research, 38(1), 33–40. https://doi.org/10.1016/0022-3999(94)90006-X
• Nemiah, J. C., Freyberger, H., & Sifneos, P. E. (1976). Alexithymia: A view of the psychosomatic process. In O. W. Hill (Ed.), Modern Trends in Psychosomatic Medicine (Vol. 3, pp. 430–439). Butterworths.
• Parker, J. D. A., Taylor, G. J., & Bagby, R. M. (2003). The 20-Item Toronto Alexithymia Scale: III. Reliability and factorial validity in a community population. Journal of Psychosomatic Research, 55(3), 269–275. https://doi.org/10.1016/S0022-3999(02)00578-0
• Taylor, G. J., Bagby, R. M., & Parker, J. D. A. (1997). Disorders of affect regulation: Alexithymia in medical and psychiatric illness. Cambridge University Press. https://doi.org/10.1017/CBO9780511526831

13. Items of the Scale (Questionnaire)