The Ash Twins Blog

Hagen et al. (2008) make a compelling case that deliberately harming oneself can be ecologically rational. Here is the abstract:

A long-standing theoretical tradition in clinical psychology and psychiatry sees deliberate selfharm (DSH), such as wrist-cutting, as ‘functional’ – a means to avoid painful emotions, for example, or to elicit attention from others. There is substantial evidence that DSH serves these functions. . . Economists and biologists have used game theory to show that, under certain circumstances, self-harmful behaviors by economic agents and animals serve important strategic goals. In particular, ‘costly signals’ can credibly reveal a ‘private state’ in situations where verbal claims and other ‘cheap’ signals might be disbelieved. . . The social contexts and associated features of DSH suggest that it might be a costly, and therefore credible, signal of need that compels social partners to respond. . .

Some other excerpts:

Skin cutting is the most common form of DSH, occurring in 40-70% of cases, followed by hitting or banging in 21-44%, and burning in 15-35%; pinching, scratching, and biting also occur in a few percent (Klonsky, Oltmanns, & Turkheimer, 2003; Ross & Heath, 2002).

Age of onset for DSH is typically between 14 and 24 years of age (Herpertz, 1995), and rates are high in clinical populations, ranging approximately from 20% (Briere & Gil, 1998; Joyce et al., 2006) to 35% in women with eating disorders (Paul, Schroeter, Dahme, & Nutzinger, 2002) to over 50% in women with borderline personality disorder (Brodsky, Malone, Ellis, Dulit, & Mann, 1997). There are few studies of the prevalence of DSH in the general population, but two relatively recent studies, one of the general population (Briere & Gil, 1998) and one of an adult, non-clinical population (Klonsky et al., 2003), both found that about 4% of respondents reported a history of DSH, with less than 1% engaging in chronic DSH (Klonsky et al., 2003). Community studies
of adolescents and college-aged populations uncovered much higher rates, however, ranging from approximately 14% to 17% (Ross & Heath, 2002; Whitlock, Eckenrode, & Silverman, 2006). Most studies have not seen significant sex differences in prevalence rates. . .

Shortly after Spence proposed that costs could serve to credibly signal private information in economic transactions, Zahavi (1975) proposed that costly displays, which he termed handicaps, could credibly signal private information about, e.g., physical condition, in conflictual interactions between the sexes, predators and prey, and so forth. The logic is the same as that put forward by Spence: by evolving a signal whose inherent cost differs by type or condition, organisms can send honest signals to those with whom they have potential conflicts. A large peacock’s tail is a credible signal of mate quality, for instance, because only a healthy, fit peacock could afford one. This idea has had as large an impact on evolutionary biology as it has had in economics. . .

The logic is as follows. When an individual needs help from others (i.e., is relatively powerless to unilaterally improve their own condition), and relations with key social partners are good, ‘cheap’ signals like verbal requests or crying are sufficient to elicit help because the parties trust and care about one another.

When there are severe conflicts with key social partners, however, a costly signal might be required to convince skeptical social partners the need was genuine and not a deceptive ploy to exploit them. In such cases, a behavior that was inherently less expensive for individuals in genuine need, but more expensive for individuals not in need, could serve as a credible signal of need. DSH is such a behavior.

For individuals whose lives are going well and who are obtaining substantial benefits from their relationships (i.e., not in need), DSH is very costly because it threatens one’s health and thus one’s ability to generate and obtain benefits. For individuals whose lives are not going well and who are not benefiting from their relationships, on the other hand (i.e., in need), DSH is much less costly because although it similarly threatens one’s health, there are fewer benefits to be lost. Thus, only individuals with little to lose, i.e., those genuinely in need, will exhibit DSH because they are the only ones who can afford to do so – these harmful behaviors credibly signal need. . .

Webb found that, compared to controls, self-harming adolescents had significantly more problems with family, friends, romantic partners, and school. School problems tended to involve bullying and not academics, although pressure to achieve and achievement failure were also factors. DSH adolescents did not feel more criticized by their parents, but they did feel less well understood . . . Additional discriminating factors included sexuality, feelings of past violation, family suicidality and illness, personal loss, family conflict and friend suicidality. Factors that protected against DSH included family intactness and cohesion, with cohesion being more important when the family was no longer intact. Webb concluded that family problems in combination with external social pressures play an important role in DSH.

Childhood trauma, particularly sexual abuse, is strongly correlated with DSH later in life. In a comprehensive review of the literature from 1988-1998 on the relationship between child abuse and self-harm and suicidality (29 studies in all), Santa Mina and Gallop (1998) found that there were more reports of self-harm, suicidal ideation, and suicidal behavior in clinical and community populations of adults who reported sexual and/or physical abuse in childhood than in comparison groups who did not report abuse. Four of the studies focused on DSH in abused and non-abused samples. Although overall rates of DSH varied widely between studies, childhood abuse typically increased DSH rates by factors of 1.5 to 4, or more (e.g., from 48% to 77%, or from 12% to 50%). . .

Klonsky (2007) conducted a meta-analysis of 18 empirical studies of DSH functionality. These studies included inpatient, outpatient, general and psychiatric hospital, forensic, and nonclinical populations, and used one or more of three methodologies: self-reported reasons for DSH, selfreports of the phenomenology of DSH (i.e., general descriptions of reasons for DSH by self-harmers, but not in reference to instances of their own self-harm), and laboratory studies using proxies for self-harm (e.g., measuring emotional arousal in patients who are visualizing selfharm). From these 18 studies Klonsky distilled seven potential functions of DSH: affect-regulation, anti-dissociation, anti-suicide, interpersonal boundaries, interpersonal influence, self-punishment, and sensation-seeking. . . All 18 studies found support for intrapsychic functions. Support for an affect-regulation function (to alleviate acute negative affect or affective arousal) was especially strong. Modest-to-strong support for an antidissociation function (e.g., to generate feelings, even if negative) was also found in the 10 studies that examined it. But there was clear and consistent support, as well, for an interpersonal-influence function of DSH. This function was found across inpatient, outpatient, general and psychiatric hospital, and non-clinical populations, and across two methodologies: self-reported reasons for DSH and selfreports of DSH phenomenology. Although it was it endorsed by a majority of patients only in one study, all of the 10 studies that examined an interpersonal-influence function found support for it. . .

Direct observations of DSH eliciting interpersonal benefits would obviously provide strong support for the bargaining model. For infants and children who exhibit selfinjurious behavior (SIB), primarily by head-banging, but also by biting, hitting, scratching, and other means, there is excellent observational evidence that these behaviors elicit responses from parents, which, in turn, reinforce the behaviors.

The prevalence of SIB in infants aged 9-18 months is 15%, and 9% in two-year-olds (Kurtz et al., 2003, and references therein). Behavioral studies show that SIB is largely socially maintained, either by social-positive reinforcement – the provision of favored stimulus, such as attention, food, or toys – or by social-negative reinforcement – the removal of an aversive stimulus, such as escape from an onerous task. Kurtz et al. (2003), for example, found that in a sample of young children (10 months to 5 years old) referred for SIB, 48% of SIB was socially maintained, mostly by social-positive reinforcement. Similar results have been found in older children and adults with mental retardation. In a referred sample of mentally retarded individuals ranging in age from 1-50+ years of age, Iwata et al. (1994) found that social reinforcement was a determinant of SIB in over two thirds of the sample. . .

Crying is obviously an evolved signal, yet it is not uncommon to consciously override its signaling function, e.g., by crying alone. Many adaptive signals and behaviors, such as the behavioral manifestations of anger, jealousy, and sadness, can be consciously overridden or concealed if desired. The same could be true of DSH. Some self-harmers admit to a tension between concealing and revealing DSH, with attention-seekers being seen as less ‘legitimate.’