NOTE: most of this applies mainly to stereotypical self-injury, such as that seen in retarded and autistic clients. Much work has been done in behavioral psychology in an attempt to explain the etiology of self-injurious behavior. In a 1990 review, Belfiore and Dattilio examine three possible explanations. They quote Phillips and Muzaffer (1961) in describing self-injury as "measures carried out by an individual upon him/herself which tend to 'cut off, to remove, to maim, to destroy, to render imperfect' some part of the body." This study also found that frequency of self-injury was higher in females but severity tended to be more extreme in males. Belfiore and Dattilio also point out that the terms "self-injury" and "self-mutilation" are deceiving; the description given above does not speak to the intent of the behavior.
Operant Conditioning
It should be noted that explanations involving operant conditioning are generally more useful when dealing with stereotypic self-injury and less useful with episodic/repetitive behavior.
Two paradigms are put forth by those who wish to explain self-injury in terms of operant conditioning. One is that individuals who self-injure are positively reinforced by getting attention and thus tend to repeat the self-harming acts. Another implication of this theory is that the sensory stimulation associated with self-harm could serve as a positive reinforcer and thus a stimulus for further self-abuse.
The other posits that individuals self-injure in order to remove some aversive stimulus or unpleasant condition (emotional, physical, whatever). This negative reinforcement paradigm is supported by research showing that intensity of self-injury can be increased by increasing the "demand" of a situation. In effect, self-harm is a way to escape otherwise intolerable emotional pain.
Sensory Contingencies
One hypothesis long held has been that self-injurers are attempting to mediate levels of sensory arousal. Self-injury can increase sensory arousal (many respondents to the Internet survey said it made them feel more real) or decrease it by masking sensory input that is even more distressing than the self-harm. This seems related to what Haines and Williams (1997) found: self-injury provides a quick and dramatic release of physiological tension/arousal. Cataldo and Harris (1982) concluded that theories of arousal, though satisfying in their parsimony, need to take into consideration biological bases of these factors.
