Ask Cipani!

“Hurting Oneself”


Many of the readers of this column have probably had clinical contact with children with autism and/or intellectual disabilities who engage in severe and intense forms of physical self-injury.

From the viewpoint of the casual observer, such self-destructive acts seem to defy any reasonable explanation. Common naïve explanations for a client who engages in self-injury belie an environmental analysis. “It comes out of nowhere,” or “s/he does it because they have autism/intellectual disabilities,” are often touted as suitable explanations. Even professionals in the field of developmental disorders as recent as the late 1970s contended that such behaviors defied a social learning model. How could someone “learn” how to inflict such damage to oneself? It was not until the groundbreaking research conducted at Johns Hopkins under Dr. Brian Iwata and colleagues (Iwata, Dorsey, Slifer, Bauman, & Richman, 1982)  that an environmental explanation became verified through an experimental analysis in the clinical laboratory1. The primary role of environmental contextual conditions was exposed as the causal factor in an individual’s self-injury (see multitude number of research studies in research journals such as the Journal of Applied Behavior Analysis2.

Can a functional operant model explain even cases where self-injury has reached dangerous proportions (involving tissue damage3)? How do such dangerous forms develop? Mild forms of behavior can start innocuously, and over time, become downright dangerous in terms of their effect on the person’s health and welfare. Let us use a hypothetical child to demonstrate the selective longitudinal environmental shaping of severe self-injurious behavior in children.

Initially, as a very young child, tantrum and screaming behaviors are effective (functional) in producing the desired outcome. These behaviors can both access desired items and events, as well as escape aversive conditions, i.e., effective in producing reinforcement. The adult mediating such behaviors may feel uncomfortable about “giving in.” Nevertheless, with such tantrum behaviors producing a very uncomfortable and aversive condition for the parent, s/he relents in order to get their child to stop such behavior. With the desired outcome delivered, the behavior does cease (relatively quickly). However, over time, such behaviors do not produce the same “crisis mode” in the adult as their occurrence had previously4. So milder forms of tantrum fall on deaf ears, and only an exacerbation of the volume and/or form produce reinforcement.

During one of these tantrums where the desired result has not yet been delivered, the child slaps himself in “a fit of anger.” After a brief amount of attention, e.g., “Stop doing that, you are hurting yourself,” the “true” desired outcome is produced. As you can see, hitting oneself has now become more adaptive than long duration screaming and tantrum bouts. As the milder forms of behavior fail to recruit the desired outcome (functional reinforcer) over time, behavior under the specific presenting deprivation or aversive condition becomes more varied. Regrettably, the variation selected by the social environment (i.e., the parent) is not often a more desirable behavior (e.g., whining would be preferable to alternatives involving hurting oneself). The form that achieves reinforcement is the increased intensity of self-injury. What was previously a slap to the leg now transforms to multiple slaps to a diverse area of one’s body.

If you have had multiple cases involving self-injury, you probably noticed that the form often involves hits/slaps to the face/head. Why would such be the case? Ask yourself this: Are you more likely to intervene if the hitting occurs to one’s thigh area, or the face? As is evident, tissue damage is much easier to inflict to the head area, e.g., drawing blood, contusions, etc. Such behavior makes it improbable that the prevailing social environment would be able to “not reinforce” such behavior with whatever outcome stops the onslaught.

Over the course of possibly months or years, self-aggressive behavior can change in form. Selective reinforcement of more obstreperous forms of hitting, such as banging one’s head against an object, can occur with greater frequency. If the ability of a more dangerous form of self-injury becomes more efficient (than other behavior response forms) at procuring the functional reinforcer, its probability increases under the relevant antecedent (motivative and discriminative) conditions. The selective reinforcement of these more destructive forms constitutes the operative factor in their development. In summary, one can see that even destructive forms of self-injury are often illuminated by analyzing the environmental selection of such forms.

There is another clinical population that engages in physical aggression towards self. Some adolescents, often females (but not exclusively) without any developmental disorder, cut themselves, mostly on their wrists and arms. The topography of the act is usually referred to as self-harm (or cutting) behaviors. These self-harm incidents are termed Non-Suicidal Self-Injury or NSSI to distinguish them from incidents with intentional suicide.

Why would some adolescents want to cut his or her arms, and draw blood? It is a pervasive contention in the mental health field to regard such self-harm acts as the client relieving internal emotional distress and turmoil (Andover, 2012; Walsh, 2012). However, there is a lack of empirical evidence that demonstrates, in a causal manner, that cutting oneself relieves emotional distress (contention is theory driven and anecdotal). If cutting oneself relieves emotional states, such a practice would occur routinely with couples in marriage conflict, inept Wall Street traders, and other high stress circumstances or jobs, including being a teacher in some classrooms! In contrast, there is over three decades of empirical evidence for the efficacy of function-derived interventions for a variety of challenging problem behaviors5.

I believe that a functional perspective of the individual client’s self-harm behaviors can lead to a long-term solution. I view self-harm behaviors as behaviors that effectively operate on their ecosystem, i.e., social environment6 . As a licensed psychologist in California, I privately consult and provide clinical direction for a particular educational agency’s mental health service called Therapeutic Behavioral Services (TBS). Our unique implementation of TBS involves a behavior-analytic approach for students (with IEP’s) identified as having an extreme risk of dangerous/disruptive behaviors that jeopardize their welfare in both home and/or school. In that manner, several of our clientele present with a history of self-harm behaviors. They may be referred upon the first occurrence, or some clients have had multiple episodes of such prior to the referral.

Most of these referrals come because of a relatively recent hospitalization for the latest episode. Our behavior-analytic approach is two-fold: (a) implement contrived contingencies that are operational shortly after client is released from hospital and (b) examine (post-hoc) the possible contextual conditions that make self-harm behavior functional in the person’s life.

The post hospital environment we design contrives contingency plans (i.e., behavioral contracts) that favor the absence of incidents. Such contrived contingencies arrange access to highly favored events for the absence of such incidents (e.g., contingent weekend privileges with friend(s), phone privileges, contingent home privileges, etc.). Since most clients who engage in such behavior do so at relatively low rates, such contrived contingencies produce immediate and consistent contact with such powerful (hopefully) reinforcers. The differential reinforcement program usually entails access to such events with the absence of a self-harm incident over the course of a week (i.e., a DRO). Such contingencies do not exploit any uncovered functional relationship between self-harm and specific functional reinforcer, so they can be enacted quite early post-hospitalization.

However, just as important as developing contrived reinforcement plans is determining why such a behavior exists. In fact, it is probably more important for the long-term success of treatment! In our functional behavior-analytic approach, we deem that such self-harm behavior(s) affects the social environment7 by producing a desired outcome (or the perception that such will happen). Further, it is most likely the case that other more innocuous behaviors are not effective (or not occurring) in procuring the desired outcome. The analysis of why self-harm has become an effective adaptive behavior needs to be answered8.

Self-harm behaviors, like other forms of behaviors, can serve a variety of environmental functions, involving both access (to something) and escape/avoidance (from something). For example, some of our clients are on home-hospital educational services. They do not go to a school for their special education services, but teaching is provided at their home. This in-home placement may have been ongoing for several months or up to a year upon their referral. School personnel periodically feel compelled to bring the student back to the school placement. Hence, quarterly meetings about doing such are held about how the student can be “transitioned” back to the school. When the gist of these discussions is relayed back to the student, episodes of self-harm occur subsequently. For those individual’s where the prospect of having to go back to school is dreadful, engaging in the behavior that led to their being removed in the first place proves effective in quashing such further discussions on the topic.

Why would going back to school be such a dreaded event? Is it possible that the material creates an aversive condition? If the student does well on her home assignments, it would appear that task difficulty and the amount of instructional material required are not the issue. It would appear that other school factors are creating the aversive motivational condition.

There is a big difference between home and school settings. At home, there are no peers. At school, there are many same aged students. Without knowing exactly why the presence of peers creates a feared/aversive event for this person, we consider that exposing the student to hundreds of students on Day 1 is probably not prudent. An abrupt transition would only engender the conditions that led to the functional self-harm behavior.

Our approach is to develop a target repertoire that makes the client fit into a social group that values such behaviors. In addition to developing essential interpersonal skills in the client, we consider that some marketable set of skills should also be targeted for development. We take into account the client’s preferences that could lead to achievable careers as an adult. In regards to the latter, developing a small “niche” peer group that may have the same vocational interests become our target.

While the behavioral functions of self-injurious behavior have been extensively studied in persons with developmental disabilities, such is not the case with people diagnosed with NSSI. Further, prevalent clinical practice with the latter group trivializes the role of environmental variables in explaining such behavior. I hope that in the near future, an understanding that hurting oneself constitutes operant behaviors whose function is environmentally derived transpires, irrespective of whom the client is.

1 Carr, Newsom & Binkoff (1980) conducted a series of experiments earlier on escape from task demands as the function of aggressive behavior, published prior to Iwata et al (1982).
2 Decades of published research has subsequently been conducted on self-injury and its various functions, not in the scope of this column to review such, reader enjoined to consult behavioral journals
3One should also consider that some cases of severe self-injury may possibly involve sensory outcomes that maintain such behavior independent of the social environment (direct access to sensory stimuli; DA1.1 in my classification system)
4 One might say that an adult’s tolerance of such milder forms increases!
5There is currently no studies to my knowledge that have looked at the functional properties of NSSI, so what follows is from my clinical experience in working with several clients presenting such behaviors
6There may be circumstances where cutting oneself produces a direct access function to sensory stimuli, but I have not had such a case yet
7 It is plausible that such behavior could be maintained by its sensory effect, i.e., seeing, feeling blood gush out. In our limited number of cases, we have not found that with our clientele. Examples of behavior that more commonly functions to access a sensory function and independent of socially mediated functions are: (a) skin picking or (b) trichotillomania (pulling out one’s hair). These forms of behavior are not symptomatic of our self-harm clientele.
8For a delineation of a variety of functions of problem behaviors in general, the reader is enjoined to consult chapters one and three of my text (Cipani, 2017).


Carr, E. G., Newsom, C. D., & Binkoff, J. A. (1980). Escape as a factor in the aggressive behavior of two retarded children. Journal of Applied Behavior Analysis, 13, 101-117.

Cipani, E. & Cipani, A.L. (2017). Behavioral classification system for problem behaviors in schools: Diagnostic manual. New York: Springer Publishing.

Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., & Richman, G. S. (1982). Toward a functional analysis of self-injury. Analysis and Intervention in Developmental Disabilities, 2, 3-20


About: Ask Cipani! is a periodic column that addresses a submitted case presentation from a BCBA who wants some suggestions/advice on identifying the function of a particular challenging behavior(s). Dr. Ennio Cipani's response addresses the presenting case in terms of possible motivative conditions (EO) and subsequent environmental functions to consider, and possible tests/validation strategies to conduct to verify the putative function. The advice provided in the Ask Cipani! column is provided with the caveat that its use is considered carefully by a certified or licensed behavior analyst. Since Dr. Cipani is only aware of the information presented to him, each reader must make a determination of whether such advice fits any particular case. It also is incumbent on any reader to determine if additional consultation should be sought for either program design and/or logistical implementation.

Ennio Cipani, Ph.D., is the author of the popular text, Functional Behavioral Assessment, Diagnosis and Treatment (2018, 3rd edition), which features a unique function-based classification system, the Cipani Behavioral Classification System or Cipani BCS. He co-authored a diagnostic manual for school settings on the Cipani BCS with his daughter, Alessandra Cipani, titled, Behavioral Classification System for Problem Behaviors in Schools. He also authored an e-book called "A Clinical Treatment Guide to 10 Common Behavioral Pediatric Problems" which can be downloaded for free here. Dr. Cipani partnered with BDS to bring you a companion CE course for his free e-book called, "10 Common Pediatric Problems & Solutions," available here.

Please direct all comments, inquiries about the column, or correspondence for Ennio Cipani to 

We welcome your case submissions, too! They may be included in a future column.
Please read the Ask Cipani! terms of use prior to submitting a case.