is a failure of mentalizing the core of borderline personality disorder, so fixing it treats it?

“if we're able to strengthen a patient's capacity to mentalise then you're treating the disorder essentially. That's exactly right.”

a guest, a clinical psychologist said on The Carlat Psychiatry Podcast, 2026-10-05, at 08:51. hear it in context →

the episode: Treating BPD Series, Episode 4: Mentalization-Based Treatment—From Reacting to Reflecting with Carla Sharp, PhD (at 08:51). we link the show's own page; we do not re-host audio.

first, the part that holds

people with borderline personality disorder do score far lower than healthy controls on reflective functioning, the measure most closely tied to mentalizing. one meta-analysis put the gap at d = 1.68, a very large effect. therapy built around mentalizing has also beaten usual care on self-harm and suicidality in randomized trials.

what we found when we went and looked

we read the 2020 Cochrane review of psychological therapies for borderline personality disorder, a 2025 review of all 14 randomized trials of mentalization-based treatment, a 2021 meta-analysis of mentalizing in the disorder, and the main study testing mentalizing as the mechanism of change. the deficit is real and large on reflective-functioning measures (d = 1.68). but the same meta-analysis found only a modest theory-of-mind deficit (d = 0.36), and its authors argue the problem is a maladaptive meta-social-cognitive style rather than a primary social-cognitive deficit. that makes "the core deficit" a theory, not a settled finding. the mechanism step ("strengthen mentalizing and you're treating the disorder") rests mostly on one uncontrolled inpatient study of 175 patients using a self-report questionnaire, where mentalizing gains tracked symptom decline at r = .89; its authors say mentalizing may "in part" explain change. the treatment built on the idea helps, but Cochrane rates that evidence as low certainty, and the 2025 review found every trial at high risk of bias.

assertion by assertion

both directions

for the speaker: the reflective-functioning deficit in borderline personality disorder is very large (d = 1.68). in the one dedicated longitudinal study, mentalizing improvement tracked symptom improvement closely (r = .89). a therapy designed to strengthen mentalizing reduced self-harm and suicidality compared with usual care in randomized trials. the adolescent mediation study fits the same model. the theory hangs together, and the speaker is describing the stated rationale of a recognized, evidence-based treatment.

against: "the core deficit" is the founding premise of one treatment school, not something the record has established. the meta-analysis found the more basic theory-of-mind deficit to be modest. it frames the problem as a distorted style of reading minds, including over-reading them, rather than simply too little capacity. the mechanism evidence comes from uncontrolled, correlational, self-report data, and its own authors claim only a partial explanation. the treatment's trials are all at high risk of bias, and Cochrane rates the effects as low certainty. those trials mostly did not measure whether mentalizing changed, so even where the therapy works, they cannot show it worked by strengthening mentalizing. Cochrane also found benefit across several borderline-specific psychotherapies that are not built on mentalizing, which is hard to square with any single-deficit account.

what this cannot say

"mentalizing" is measured several ways (interview-coded reflective functioning, self-report questionnaires, theory-of-mind tasks), and these give very different effect sizes, so how big the deficit looks depends on the measure. we could read only the abstracts of the Cochrane review (through PubMed), the 2025 trial review and the mechanism study, not their full texts. we could not open a 2024 pre-registered systematic review of mentalizing as a mediator across 33 papers, so we do not cite it. a moderator-mediator literature beyond what we read may exist. our verdict is on whether the claim's certainty matches the record, not on whether the theory is wrong.

sources, each one fetched and read

1. Cochrane Database of Systematic Reviews, 'Psychological therapies for people with borderline personality disorder', 2020, CD012955.pub2 (PubMed abstract) What we read there: mentalization-based treatment vs usual treatment: self-harm RR 0.62 (95% CI 0.49-0.80, 3 trials, 252 participants); suicidality RR 0.10 (95% CI 0.04-0.30, 3 trials, 218 participants); all low-quality evidence, 'the true magnitude of these effects is uncertain'. Fetched 2026-10-30 (HTTP 200). https://pubmed.ncbi.nlm.nih.gov/32368793/
2. 'Mentalization-based therapy for borderline personality disorder: state-of-the-science and future directions', Clinical Psychology and Psychotherapy, 2025 (PubMed abstract) What we read there: 14 RCTs; 'All identified RCTs were at high risk of bias'; research 'largely focused on symptom reduction'; 'MBT shows promise, but methodological shortcomings and limited generalizability constrain its evidence base.' Fetched 2026-10-30 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/40659413/
3. 'A meta-analysis of theory of mind and "mentalization" in borderline personality disorder: a true neuro-social-cognitive or meta-social-cognitive impairment?', Psychological Medicine, 2021, 51(15) What we read there: 34 studies, 1,448 people with the disorder and 2,006 controls; reflective functioning d = 1.68; overall theory of mind d = 0.36; hypermentalizing d = 0.60; 'BPD is characterized by very severe deficits in RF and modest and selective abnormalities in ToM'. Fetched 2026-10-30 (HTTP 200). https://cambridge.org/core/journals/psychological-medicine/article/metaanalysis-of-theory-of-mind-and-mentalization-in-borderline-personality-disorder-a-true-neurosocialcognitive-or-metasocialcognitive-impairment/6A91447CFE2AF354651D87A4A4D47E6F
4. 'Mentalizing as a mechanism of change in the treatment of patients with borderline personality disorder: a parallel process growth modeling approach', Personality Disorders: Theory, Research, and Treatment, 2018, PMID 28569524 What we read there: 175 inpatients in psychodynamic treatment, measured with the Reflective Functioning Questionnaire, no control group; 'improvements in mentalizing were strongly associated with the rate of decrease in symptomatic distress over time (r = .89)'; 'more research is needed'. Fetched 2026-10-30 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/28569524/
5. 'Mentalization mediates the relationship between borderline personality features and internalizing and externalizing psychopathology in adolescents', Scandinavian Journal of Child and Adolescent Psychiatry and Psychology, 2019 What we read there: in 109 adolescents, mentalization mediated about 45% of the link with externalizing and 54% with internalizing problems; the design was cross-sectional and self-report only, so causality cannot be established. Fetched 2026-10-30 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC7709936

speakers are described by role, not named. one sentence is quoted under fair use, with attribution and a link to the original at the timestamp; we do not re-host audio or video. nothing on this page is a reason to start, stop or change a medication. if you have the file behind anything marked unverifiable, the research board is where people can answer. adam