Despite its early stage of development, much may be learned
from the study of environments in terms of predicting individual violence. The
following are what appear to be the best candidates for situational or environmental
correlates of violent behavior that potentially can be of use for prediction in
the individual case. The first three can be conceptualized either as environmental
"support systems" used by an individual for coping with life stress
(President's Commission on Mental Health 1978), or as the sources of the life
stress itself (see chapter 5).
One of the best predictors of whether released mental patients will survive in
the community without being rehospitalized is the degree of support provided by
their families (Fairweather, Sauders, and Tornatzky 1974). As Stone (1975, p.
13) stated, "A principal social function of the law-mental health system
is to provide technical care for those individuals who are temporarily or permanently
extruded from society's principal caretaking unit, the family. The wisdom and
morality of this extrusion and the quality of this technical care are the bedrock
problems of the law-mental health system."
In the case of violent
behavior, the family context is crucial since family members are so frequently
the victims of violent behavior (Monahan, l977b). Skodol and Karasu (1978), as
noted previously, found that, in 77 percent of emergency commitment cases in which
the patients admitted to actively considering violence, the victims were family
members. The frequency of violence in police family-crisis interventions has been
well documented (Bard 1969; Driscoll, Meyer and Schanie 1973).
family environment may be critical because of its role in supporting or discouraging
violent behavior on the part of the family member whose behavior is being predicted.
The probability of a person being violent may be greater if he or she resides
in a family that encourages robbery as a career and where violence by other family
members is a frequent occurrence, than if he or she has support and models for
nonviolent modes of interaction and needs satisfaction. Though their prior records
may be the same, the probability of recidivism of a released offender living with
grandparents on a farm may be substantially less than that of another offender
living with alcoholic friends in an inner city.
is an enormous sociological literature on "peer group influences" on
behavior, particularly adolescent behavior. Likewise, numerous psychological studies
attest to the effects of one's friends as behavior models (Bandura 1969). There
is, in addition, ample folk wisdom about the effects of "getting in with
the wrong crowd" on criminal activity. Gang violence is probably the paradigmatic
case of peer-induced harm. To the extent that a person's violent behavior in the
past has occurred in a particular social context (rather than "as a loner,"
for example), it may be important to ascertain whether the same peers who encouraged
previous violence are likely to provide similar encouragement in the future. The
person returning to the same friends who participated in the last robbery may
have a greater likelihood of future violent crimes than the person who has broken
contact with a criminally oriented support group.
is a growing body of research on the effect of employment upon criminal behavior,
although the research generally does not separate violent from nonviolent crime
(Monahan and Monahan 1977). At monthly intervals, Glaser (1964) interviewed a
sample of 135 parolees released from Federal institutions in 1959 and 1960. In
comparing the job holding activity of the men who completed parole with that of
men returned to prison, he found that the eventual successes acquired their first
jobs sooner, and during the initial period of parole, earned a higher monthly
income than did the eventual recidivists.
Cook (1975), studying
327 male felons released from Massachusetts prisons in 1959, found that 65 percent
of those who held a "satisfactory" job (defined as a job which lasted
1 month or more) during the first 3 months of parole were eventually successful
in completing an 18-month parole period compared with a 36 percent success rate
among those who did not have a satisfactory job during the first 3 months. Seventy-five
percent of parolees holding a satisfactory job during the second 3 months of parole
were eventual successes, compared with 40 percent of those who did not hold a
satisfactory job. Eighty-nine percent of those having a satisfactory job at the
end of their first year on parole completed the parole period without revocation,
while only 50 percent of those not satisfactorily employed successfully completed
their term of parole.
Cook (1975) also found that steady job
holding was related to parole success, while frequent job changing increased the
likelihood that a parolee would recidivate. The probability of recidivism during
the second 3 months on parole increased directly with the number of jobs held
during the first 3 months, from 11 percent recidivism when one job was held to
43 percent when five jobs were held. While such data do not prove a causal relationship
between employment and crime (since some third factor may cause both the reduction
in recidivism and whether one is employed), it would appear that holding a job
that is both satisfying and supportive reduces the probability of recidivism for
at least some criminal offenders.
AVAILABILITY OF VICTIMS
as Toch (1969) has emphasized, may be thought of as an interactional concept.
It takes two for a murder to occur. Clearly, some persons are relatively indiscriminate
in the victims they choose. Mergargee (1976, p. 8) quotes a steel worker interviewed
by Studs Terkel in Working: "All day long I wanted to tell my foreman to
go fuck himself, but I can't. So I find a guy in a tavern; to tell him that. And
he tells me too . . . He's punching me and I'm punching him, because we actually
want to punch somebody else" (Terkel, 1974, p. xxxiii). Consistent with the
frustration aggression hypothesis and theories of displacement, it is likely that
both parties to this dispute would have found other "victims" had they
not chanced upon each other.
There may be other types of individuals
who are quite specific in their choice of victim and will not be violent other
than to a given victim or class of victims. Spouse murderers, for example, have
a very low recidivism rate since they have removed their source of irritation.
Incest offenders may desist when their children grow up. The now famous Tarasoff
case (1976) is a clear example of victim specific violence (Roth and Meisel 1977;
Wexier 1979). A client revealed in therapy his intention to kill a woman who had
rejected his romantic interests. The client then committed no violent acts for
2 months while the woman was on vacation. Shortly after she returned home, he
murdered her. As Shah (1978) has noted:
wish to know whether the dangerous acts are more likely to occur against some
particular persons (e.g., a spouse or girl friend, the individual's own children,
or a neighbor with whom longstanding conflicts have occurred); and/or against
some broader group of people (e.g., minor boys or girls in the case of a pedophile,
adult women in the case of certain exhibitionists or rapists, etc); and/or a more
dispersed segment of the community (e.g., the likely victims of "purse-snatchings"
and other street robberies, potential victims of recidivistic drunken drivers,
etc.) (p. 180).
AVAILABILITY OF WEAPONS
the presence of weapons has long been held to be a situational instigation to
violent behavior (Berkowitz and LePage 1967). Equally importantly, weapons may
influence not the occurrence but the severity and lethality of violent behavior
(Newton and Zimring 1970; Zimring 1977). The difference between assault and murder
frequently revolves around whether the offender had a knife or only a fist at
his or her disposal. The difference between murder and attempted murder likewise
is often determined by whether the offender has access to a gun or a knife.
as the possession of the "means" to commit suicide is a frequently used
predictor of suicide (Beck, Resnick, and Lettieri 1974), so the person who reveals
possession of a household arsenal may be more likely to harm another than the
individual without such means of destruction.
The evidence linking the excessive use of alcohol to violent
behavior was noted in the last chapter. There is a great deal of literature on
criminology relating the high frequency of violent behavior in and near bars and
taverns (e.g., Wolfgang 1958). At least for those persons whose previous violent
behavior has been associated with a state of intoxication, the easy availability
of alcohol and the presence of a support group which encourages its excessive
use (drinking buddies) may constitute a high-risk context for the occurrence of
- Monahan, John, Predicting Violent Behavior, Sage
Publications, Inc.: London, 1981.
Reflection Exercise #9
The preceding section contained information
about environmental approaches to improved clinical prediction of violence. Write
three case study examples regarding how you might use the content of this section
in your practice.
Online Continuing Education QUESTION
What concept is gang violence probably the best paradigmatic example
of? Record the letter of the correct answer the .