Self-Mutilation

Abstract

Self-mutilation is the deliberate, direct destruction of one's own body tissue without conscious suicidal intent and for purposes not socially sanctioned. Armando Favazza's taxonomy sorts it into major (rare, severe, usually psychotic), stereotypic (rhythmic, fixed, seen in developmental and neurological conditions), and superficial/moderate forms; the last, cutting and burning foremost, is nearly all of clinical practice and overlaps closely with nonsuicidal self-injury. The dominant account is functional, not diagnostic: Matthew Nock and Mitchell Prinstein's four-function model classifies each act by the reinforcement that maintains it, crossing an automatic/social axis with a positive/negative one. Most self-injury is automatic negative reinforcement — it ends an aversive emotional state — so affect regulation, formalized in the experiential-avoidance model, is the best-supported function. This article covers the taxonomy, the four-function model, the distinction from suicide, the epidemiology, and the benefits-and-barriers reframing, with interactive demonstrations.

Keywords: self-mutilation, nonsuicidal self-injury, four-function model, affect regulation, experiential avoidance

What Self-Mutilation Is

Self-mutilation is the deliberate, direct destruction or alteration of one's own body tissue, carried out without conscious suicidal intent and outside any culturally sanctioned practice. The three clauses do real work. Direct excludes the slow self-harm of starvation or substance misuse. Without suicidal intent separates it from a suicide attempt — the person is not trying to die, and usually expects to go on living. Not socially sanctioned excludes tattooing, piercing, and ritual scarification, which a culture endorses. What remains — most often cutting, but also burning, scratching to the point of bleeding, self-hitting, and interfering with wound healing — is the behavior Armando Favazza brought under clinical study and that the research literature now largely studies under the label nonsuicidal self-injury (NSSI) (Cipriano et al., 2017).

Favazza's enduring contribution was a taxonomy (Favazza, 1998; see Table 1). Major self-mutilation is rare, severe, and typically occurs during psychosis: enucleation, castration, amputation. Stereotypic self-injury is rhythmic and fixed in form — head-banging, self-biting — and is seen in developmental and neurological conditions such as autism and Lesch–Nyhan syndrome. Superficial or moderate self-mutilation, the third and overwhelmingly most common type, is the repetitive, low-lethality tissue damage — cutting and burning above all — that presents in clinical and community settings and that the rest of this article, like most of the field, concerns.

The taxonomy below is descriptive, not a set of mutually exclusive diagnoses; a single patient can show more than one type, and the boundaries are clinical conventions rather than natural kinds.

The Four-Function Model

The question that organizes the modern field is not what the behavior looks like but why it persists. Nock and Prinstein (2004) answered it with a functional model that is now the dominant framework. Rather than treat self-injury as a symptom of an underlying disorder, the model treats each act as a learned behavior maintained by its consequences, and classifies those consequences along two orthogonal axes. The first axis is automatic versus social: does the act operate on the person's own internal state, or on their social environment? The second is positive versus negative reinforcement: does it work by producing a desired stimulus, or by removing an aversive one? Crossing the two yields four functions.

Automatic negative reinforcement (ANR) removes an unwanted internal state — it ends or dampens overwhelming tension, numbness, or distress. Automatic positive reinforcement (APR) generates a desired internal state — it produces feeling where there was dissociative numbness, or a sense of control. Social negative reinforcement (SNR) removes an unwanted social demand — it allows escape from an interpersonal situation. Social positive reinforcement (SPR) produces a desired social response — attention, care, or help. The axes are independent, so an act can serve more than one function at once, and functions can change over the course of a person's history with the behavior. The model's power is that it ties assessment directly to treatment: identifying the maintaining reinforcement identifies what an alternative behavior must deliver instead.

Affect Regulation and Experiential Avoidance

Across studies, the automatic functions dominate, and ANR dominates among them: people overwhelmingly report that they injure to stop feeling something unbearable (Klonsky, 2007; Taylor et al., 2018). This is the affect-regulation account, and its most developed theoretical form is Chapman, Gratz, and Brown's (2006) experiential-avoidance model. On that model, self-injury is a negatively reinforced escape: an aversive emotional arousal arises, the person has no tolerable way to sit with it, the act sharply reduces the arousal, and that relief — immediate and reliable — reinforces the behavior, making it more likely the next time arousal rises. The loop is the engine of repetition. Because the relief is real and fast, the behavior is a highly effective short-term emotion-regulation strategy, which is precisely why it is so persistent and so resistant to simple injunctions to stop.

This reframes self-injury as a coping behavior rather than a manipulation or a bid for attention — the social functions, while genuine, are far less often endorsed. It also explains the close association between self-injury and the emotional disorders: where emotion dysregulation is high and distress tolerance is low, a behavior that reliably terminates distress is strongly selected (Bentley et al., 2015). Klonsky's (2007) review weighed the competing candidate functions against the evidence and concluded that affect regulation was the best supported, a conclusion the later meta-analytic work has sustained.

Distinguishing Self-Mutilation from Suicide

The defining clause — without conscious suicidal intent — marks the most important clinical distinction, and getting it wrong in either direction is dangerous. Self-injury and a suicide attempt differ in intent (to alter a feeling state versus to end life), in method (low-lethality, repeated, often the same site versus higher-lethality), in frequency (self-injury is typically chronic and repetitive), and in the person's own account of what the act was for. Nock's integrative work (2009, 2010) treats them as related but distinct behaviors with partly overlapping and partly separate causes.

Distinct does not mean unconnected. A history of nonsuicidal self-injury is one of the more robust prospective predictors of a later suicide attempt — the behaviors travel together even though their immediate aims differ. The most influential explanation of why comes from Van Orden and colleagues' interpersonal theory of suicide (2010) and its construct of acquired capability: a person does not attempt suicide on desire alone but must also overcome the powerful instinct for self-preservation, and repeated self-injury erodes that barrier. By raising pain tolerance and habituating the person to the fear and physical difficulty of self-directed violence, self-injury supplies exactly the fearlessness-of-death and pain-endurance that the capability requires, so that if suicidal intent later develops, the means to act on it are already in place. The clinical upshot is that self-injury must be taken seriously as a risk marker in its own right, without being conflated with a suicide attempt.

Epidemiology and Risk

Self-injury is common, and most common in adolescence. Swannell and colleagues' (2014) meta-analysis of nonclinical samples put pooled lifetime prevalence at roughly 17% in adolescents, 13% in young adults, and 6% in adults — a clear developmental gradient, with onset typically in the early-to-mid teens. Prevalence estimates vary widely across studies because of how self-injury is defined and measured; single-item questions yield higher rates than behaviorally specific checklists such as the Deliberate Self-Harm Inventory (Gratz, 2001), and the meta-analysis found that methodological factors accounted for a substantial part of the between-study variance.

Prediction, by contrast, is hard. Fox and colleagues' (2015) meta-analysis of longitudinal risk factors found that most single predictors — demographic, clinical, psychological — are weak and non-specific, with effect sizes that preclude accurate individual forecasting. Prior self-injury is the strongest predictor of future self-injury, which is informative clinically but limited for primary prevention. The pattern mirrors the broader suicide-prediction literature: many statistically reliable correlates, none individually strong enough to pick out who will injure. This is one reason the field has moved from asking who is at risk toward asking why the behavior is chosen — the question the functional and benefits-and-barriers models address.

Figure

Interactive Demonstrations

Three demonstrations make the functional account concrete: a classifier that places a self-injury episode in the four-function grid, an affect-regulation trace that shows why the loop reinforces itself, and a prevalence explorer for the developmental gradient. Each is deterministic and runs entirely in the browser.

Worked Example

Consider a clinician applying the four-function model to a concrete episode. A 16-year-old reports that, alone in her room after a distressing argument, she felt a rising, unbearable tension; she cut her forearm; the tension dropped sharply within minutes; no one else knew. Classify the function.

Work the two axes in turn. Automatic or social? The act operated on her own internal state and occurred in private, with no social audience and no interpersonal demand removed or response elicited — so automatic, not social. Positive or negative? The salient consequence was the removal of an aversive state (the tension dropped) rather than the production of a desired one — so negative, not positive. The intersection is automatic negative reinforcement (ANR) — the affect-regulation function, and the single most commonly endorsed cell in the grid (Taylor et al., 2018).

The classification is not academic. Because the maintaining reinforcement is the termination of aversive arousal, an effective alternative behavior must deliver the same thing — rapid, reliable down-regulation of distress — which is precisely what distress-tolerance and emotion-regulation skills target. Had the same adolescent injured in a way that reliably summoned a caregiver's help (social positive reinforcement) or ended an intolerable social demand (social negative reinforcement), the treatment implication would differ: the skill to build would be interpersonal, not intrapersonal. The grid converts a description into a treatment target.

Discussion

The shift from Favazza's descriptive taxonomy to Nock and Prinstein's functional model is a shift in the kind of question being asked. The taxonomy names what a behavior is and where it sits among related phenomena; the functional model names what it does for the person and therefore what would have to replace it. The two are complementary — a clinician needs the taxonomy to recognize superficial/moderate self-injury as the common, treatable form distinct from the rare major type, and needs the functional model to work out why this particular patient's behavior persists. Affect regulation is the modal answer, but it is not the only one, and the model's discipline is that the function is assessed per person, not assumed.

The reframing also corrects two persistent errors. The first is treating self-injury as a failed or disguised suicide attempt; it is a distinct behavior with its own, usually non-lethal, aims, even as it raises later suicide risk. The second is treating it as manipulation — a reading the evidence does not support, since the automatic functions far outweigh the social ones. What the behavior most often is, is an effective if costly way of ending unbearable feeling, learned and maintained because it works. That is a more sobering picture than either error, and a more useful one, because it points directly at what treatment must provide.

Current Directions

The most important recent theoretical advance reframes the question again. Hooley and Franklin's (2018) benefits-and-barriers model starts from an observation the affect-regulation account alone does not fully explain: many people experience intense, poorly regulated distress, yet only a minority injure themselves. If self-injury simply follows from the need to regulate affect, why is it not far more common? Their answer is that the benefits of self-injury (chiefly affect regulation) are widely available, but most people are held back by barriers — the pain, the aversion to one's own blood, the self-image of someone who does not do this, the social and physical difficulty of the act. Self-injury emerges when those barriers are lowered, and the model organizes the known risk factors by whether they raise the benefits or lower the barriers.

This integrative turn connects to active work on the mechanics of barrier erosion — how repeated self-injury raises pain tolerance and habituates the person to self-directed harm — which is also the proposed bridge to elevated suicide risk. Alongside the theory, the field continues to refine measurement and to sharpen the boundary between nonsuicidal self-injury and suicidal behavior, including the debate over NSSI as a distinct diagnostic entity (Klonsky et al., 2014). The through-line is a move from cataloguing correlates toward explaining the behavior's selection: not only who is at risk, but why, among those at risk, this particular behavior is the one that is chosen.

Common Misconceptions

Self-injury is a suicide attempt.

By definition it is carried out without conscious suicidal intent — the aim is usually to change a feeling state, not to end life. It is a distinct behavior, though it does raise the risk of a later suicide attempt (Nock, 2010).

It is mainly attention-seeking or manipulation.

The evidence is the opposite: automatic (internal-state) functions are endorsed far more often than social ones, and most self-injury is done in private. It is primarily a way of regulating one's own emotion (Taylor et al., 2018).

Only people with borderline personality disorder self-injure.

Self-injury is transdiagnostic and common in community samples, especially adolescents. It is associated with many emotional disorders and with none — it is a behavior, not a diagnosis (Swannell et al., 2014).

The severity of injury shows how serious the problem is.

Tissue damage is a poor index of psychological distress or risk. Frequency, function, and the loss of alternative coping matter more than how deep a given wound is (Klonsky, 2007).

Glossary

Key Researchers

Armando R. Favazza

(University of Missouri). Psychiatrist whose work established the modern clinical study of self-mutilation and whose taxonomy of major, stereotypic, and superficial/moderate self-injury remains the field's organizing frame. Wikipedia · Wikidata

Kim L. Gratz

(University of Toledo). Clinical psychologist who developed the Deliberate Self-Harm Inventory and, with Alexander Chapman, the experiential-avoidance model of self-harm, grounding the behavior in emotion dysregulation. Scholar

Jill M. Hooley

(Harvard University). Clinical psychologist whose benefits-and-barriers model, developed with Joseph Franklin, reframes self-injury as a behavior whose barriers are lowered rather than whose benefits are unique. ORCID · Faculty · Scholar

E. David Klonsky

(University of British Columbia). Clinical psychologist whose review of the evidence established affect regulation as the best-supported function of self-injury and who developed influential models distinguishing it from suicidal behavior. Faculty · Scholar

Matthew K. Nock

(Harvard University). Clinical psychologist whose four-function model — classifying self-injury by the reinforcement that maintains it — is the dominant contemporary framework for understanding why the behavior persists. ORCID · Faculty · Scholar

Janis Whitlock

(Cornell University). Developmental researcher whose work on the epidemiology of self-injury among adolescents and young adults, and on its social and online dimensions, grounds the account of prevalence and context. Faculty · Scholar

Frequently Asked Questions

Is self-mutilation the same as a suicide attempt?

No. Self-mutilation is defined by the absence of conscious suicidal intent, and the usual aim is to change a feeling state rather than to end life. The two differ in intent, method, and frequency, though a history of self-injury does raise the risk of a later suicide attempt.

What is the difference between self-mutilation and nonsuicidal self-injury?

They are nearly the same behavior under two labels. Self-mutilation is the older clinical term covering Favazza's three types; nonsuicidal self-injury (NSSI) is the research construct, and it corresponds closely to the common superficial/moderate type.

Why do people self-injure?

Most often to regulate emotion, by ending an unbearable internal state such as tension, distress, or numbness. In the four-function model this is automatic negative reinforcement, the most commonly endorsed function; social functions exist but are far less often reported.

What is the four-function model?

Nock and Prinstein's framework that classifies each act of self-injury by the reinforcement maintaining it, crossing two axes: automatic versus social, and positive versus negative reinforcement. The four cells give four distinct functions.

Is self-injury just attention-seeking?

No. The evidence shows automatic (internal) functions are endorsed much more often than social ones, and most self-injury occurs in private. Treating it as manipulation misreads the behavior and the data.

How common is self-injury?

Common, especially in adolescence. A meta-analysis of nonclinical samples estimated lifetime prevalence at roughly 17% in adolescents, 13% in young adults, and 6% in adults, with onset usually in the early-to-mid teens.

Can we predict who will self-injure?

Poorly. Most individual risk factors are weak and non-specific; prior self-injury is the strongest single predictor of future self-injury. This limited predictability is part of why the field now focuses on why the behavior is chosen rather than only who is at risk.

Does self-injury mean someone has borderline personality disorder?

No. Self-injury is transdiagnostic, common across many emotional disorders and in people with no diagnosis at all. It is one of several criteria that can appear in borderline personality disorder, but it is neither necessary nor sufficient for it.

References

Bentley, K. H., Cassiello-Robbins, C. F., Vittorio, L., Sauer-Zavala, S., & Barlow, D. H. (2015). The association between nonsuicidal self-injury and the emotional disorders: A meta-analytic review. Clinical Psychology Review, 37, 72–88. https://doi.org/10.1016/j.cpr.2015.02.006

Chapman, A. L., Gratz, K. L., & Brown, M. Z. (2006). Solving the puzzle of deliberate self-harm: The experiential avoidance model. Behaviour Research and Therapy, 44(3), 371–394. https://doi.org/10.1016/j.brat.2005.03.005

Cipriano, A., Cella, S., & Cotrufo, P. (2017). Nonsuicidal self-injury: A systematic review. Frontiers in Psychology, 8, 1946. https://doi.org/10.3389/fpsyg.2017.01946

Favazza, A. R. (1998). The coming of age of self-mutilation. The Journal of Nervous and Mental Disease, 186(5), 259–268. https://doi.org/10.1097/00005053-199805000-00001

Fox, K. R., Franklin, J. C., Ribeiro, J. D., Kleiman, E. M., Bentley, K. H., & Nock, M. K. (2015). Meta-analysis of risk factors for nonsuicidal self-injury. Clinical Psychology Review, 42, 156–167. https://doi.org/10.1016/j.cpr.2015.09.002

Gratz, K. L. (2001). Measurement of deliberate self-harm: Preliminary data on the Deliberate Self-Harm Inventory. Journal of Psychopathology and Behavioral Assessment, 23(4), 253–263. https://doi.org/10.1023/A:1012779403943

Hooley, J. M., & Franklin, J. C. (2018). Why do people hurt themselves? A new conceptual model of nonsuicidal self-injury. Clinical Psychological Science, 6(3), 428–451. https://doi.org/10.1177/2167702617745641

Klonsky, E. D. (2007). The functions of deliberate self-injury: A review of the evidence. Clinical Psychology Review, 27(2), 226–239. https://doi.org/10.1016/j.cpr.2006.08.002

Klonsky, E. D., Victor, S. E., & Saffer, B. Y. (2014). Nonsuicidal self-injury: What we know, and what we need to know. The Canadian Journal of Psychiatry, 59(11), 565–568. https://doi.org/10.1177/070674371405901101

Nock, M. K., & Prinstein, M. J. (2004). A functional approach to the assessment of self-mutilative behavior. Journal of Consulting and Clinical Psychology, 72(5), 885–890. https://doi.org/10.1037/0022-006X.72.5.885

Nock, M. K. (2009). Why do people hurt themselves? New insights into the nature and functions of self-injury. Current Directions in Psychological Science, 18(2), 78–83. https://doi.org/10.1111/j.1467-8721.2009.01613.x

Nock, M. K. (2010). Self-injury. Annual Review of Clinical Psychology, 6, 339–363. https://doi.org/10.1146/annurev.clinpsy.121208.131258

Swannell, S. V., Martin, G. E., Page, A., Hasking, P., & St John, N. J. (2014). Prevalence of nonsuicidal self-injury in nonclinical samples: Systematic review, meta-analysis and meta-regression. Suicide and Life-Threatening Behavior, 44(3), 273–303. https://doi.org/10.1111/sltb.12070

Taylor, P. J., Jomar, K., Dhingra, K., Forrester, R., Shahmalak, U., & Dickson, J. M. (2018). A meta-analysis of the prevalence of different functions of non-suicidal self-injury. Journal of Affective Disorders, 227, 759–769. https://doi.org/10.1016/j.jad.2017.11.073

Van Orden, K. A., Witte, T. K., Cukrowicz, K. C., Braithwaite, S. R., Selby, E. A., & Joiner, T. E. (2010). The interpersonal theory of suicide. Psychological Review, 117(2), 575–600. https://doi.org/10.1037/a0018697