4-6 pg paper

The Empirical Status of the Subculture of Violence Thesis Author(s): Howard S. Erlanger Source: Social Problems, Vol. 22, No. 2 (Dec., 1974), pp. 280-292 Published by: Oxford University Press on behalf of the Society for the Study of Social Problems Stable URL: http://www.jstor.org/stable/799764 Accessed: 04-04-2016 17:26 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at

http://about.jstor.org/terms

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted

digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about

JSTOR, please contact [email protected].

Society for the Study of Social Problems, Oxford University Press are collaborating with JSTOR to digitize, preserve and extend access to Social Problems

This content downloaded from 141.210.130.85 on Mon, 04 Apr 2016 17:26:33 UTC All use subject to http://about.jstor.org/terms

THE EMPIRICAL STATUS OF THE SUBCULTURE OF VIOLENCE THESIS*

HOWARD S. ERLANGER

The University of Wisconsin

This article reviews some of the existing literature bearing on the subculture of violence thesis (Wolfgang, 1958), reports the results of a re-analysis of survey data collected for the President's Commission on the Causes and Prevention of Violence, and presents new data on peer esteem and social psychological correlates of fighting among males in Milwaukee, Wisconsin. It is concluded that although the subculture of violence thesis has not been definitively tested, the weight of the evidence is against it.

In the study of adult interpersonal violence (which may be defined as acts of physical aggression directed at persons, excluding acts under the aegis of, or directed against, politi- cal, parental, or other authority), one of the most important and most often cited theoretical statements has been the "subculture of vio- lence" thesis (Wolfgang, 1958; Wolfgang and Ferracuti, 1967). Ac- cording to Wolfgang and Ferracuti, violence results from adherence to a set of values which supports and en- courages its expression. These val- ues are seen as being in conflict with but not totally in opposition to those of the dominant culture. It is said that within the subculture, vari- ous stimuli such as a jostle, a slightly derogatory remark, or the appear- ance of a weapon in the hands of an adversary are perceived differently than in the dominant culture; in the

subculture they evoke a combative reaction.

Although violence obviously is not and cannot be used continu-

ously, Wolfgang and Ferracuti see the requirement to be violent as a norm governing a wide variety of situations. They judge the subcul- tural theme to be "penetrating and diffuse" and argue that violations of the subcultural norm are punished within the subculture. Adherence to

the norm is not necessarily viewed as illicit conduct, and "a carrier and user of violence will [generally] not be burdened by conscious guilt ... [and] even law-abiding members of the local subcultural area may not view various expressions of violence as menacing or immoral" (Wolfgang and Ferracuti, 1967:161).

When preparing the 1967 vol- ume, Wolfgang and Ferracuti could locate no data on the distribution of

values regarding violence, so they were forced to rely on inferences from available data on criminal acts of interpersonal violence. Since criminal statistics indicate that the

groups with the highest rates of homicide are males, nonwhites, lower- and working-class whites, and young adults, it is, therefore, among these groups that "we should find in most intense degree a subcul- ture of violence" (Wolfgang and Ferracuti, 1967:153). They acknowl-

* I am especially grateful to Russell Mid- dleton, James Greenly, and James F. Short, Jr. for the useful critiques they prepared on earlier drafts of this paper. I am also indebted to Hal Winsborough for his methodological advice and to Irene Rodgers for her very ca- pable research assistance.

The research reported here was supported by funds granted by the National Institute of Mental Health and by funds granted to the Institute for Research on Poverty at the Uni- versity of Wisconsin by the Office of Economic Opportunity pursuant to the Economic Opportunity Act of 1964. The opinions expressed are those of the author.

This content downloaded from 141.210.130.85 on Mon, 04 Apr 2016 17:26:33 UTC All use subject to http://about.jstor.org/terms

Subculture of Violence 281

edge that their reasoning here is cir- cular, and they agree that individual data on values are necessary for an adequate test of the theory.

In the years since the subculture of violence thesis was first intro- duced, there have been a variety of studies which directly or indirectly bring data to bear on the thesis. In the study of juvenile delinquency, for example, there has been a re- lated controversy over the value sys- tem of adolescent gangs. W. Miller (1958) has argued that these gangs reflect the "focal concerns" of lower-class culture, which he sees as including "toughness" and "excite- ment." However, the analysis of gang values by Short and Strodtbeck (1965) failed to confirm the exis- tence of these focal concerns, and a study by Lerman (1968) has ques- tioned the existence of a distinctive lower-class culture reflected in gangs. In addition, various studies (e.g., Short and Strodtbeck, 1965; Jansyn, 1966) have concluded that gang activity is related more to group processes than to a violence oriented subculture,' and later work by Miller and his colleagues does not indicate that physical aggression is an important part of lower-class gang life (Miller et al., 1961; Miller, 1966).

Some studies, such as those of Kobrin et al. (1967) and Yablonsky

(1962) have found that status within the gang is at least in part based on the criteria outlined by Wolfgang and Ferracuti, but Yablonsky has also emphasized the fluid nature of group membership and the limited ability of leaders to sanction mem- bers who do not conform (see also Matza, 1964; Short and Strodtbeck, 1965). Moreover, it is important to remember that the existence of vio- lence as a criterion of status in gangs in low-income neighborhoods is in- sufficient to establish the existence

of such norms among nongang juveniles in those neighborhoods, especially since it is generally the most extreme gangs that have been studied. When the whole juvenile population is studied, the patterns can be quite different (Hirschi, 1969).

In the study of adult interpersonal violence, research has been much more limited. Various studies and

texts in sociology (e.g., Amir, 1971; Clinard, 1973; Schur, 1969) and so- cial psychology (e.g., Akers, 1973; Toch, 1969) have stressed the sub- cultural view, but they have not used individual data to support their arguments. The idea of a subculture of violence is conspicuous by its ab- sence in various well known ethno-

graphic studies of adult lower-class communities (e.g., Liebow, 1967; Suttles, 1968; Whyte, 1955). Since these writers are not explicitly con- cerned with the issue, the absence of discussion is not definitive evi- dence against the thesis. It does, however, suggest that violence is not a major theme in the groups studied.2

1 Some critics of this manuscript have held that the group process material in the Short and Strodtbeck work supports the subculture of violence thesis. Although the material may be open to varying interpretations, Short re- ports that he "never felt that our data were supportive of the subculture of violence thesis" (personal communication). Short feels that the group process mechanisms are re- lated to subcultural variations, but that the subculture of violence thesis is not particu- larly helpful in explaining the outcomes they observed.

2 One anthropological study that does re- count many violent incidents is Lewis's biog- raphy of the Rios family in Puerto Rico and New York. But although a degree of

This content downloaded from 141.210.130.85 on Mon, 04 Apr 2016 17:26:33 UTC All use subject to http://about.jstor.org/terms

282 SOCIAL PROBLEMS

Few systematic studies of class differences in values or attitudes

among adults have been reported in the literature, and some of the most often cited are quite dated. Most studies that do exist do not spe- cifically deal with low-income groups; the lower class is either omitted or combined with the work- ing class for analysis.3 Insofar as the present author can determine, until the late 1960s no survey data on the values or attitudes of adults toward violence were available.

In a recent paper, Ball-Rokeach (1973) analyzes responses to the Rokeach Value Survey given by males with various degrees of par- ticipation in violence. She finds no important differences in the ranking of 18 "terminal values" or of 18 "in-

strumental values" by men classified as having no, a "moderate," or a "high" degree of participation in violence at any time in their life. She reports that controls for education and income, which are crucial for the examination of a subculture which is said to be class-based, do not affect the findings. A compari-

son of prisoners convicted of violent crimes and persons convicted of non-violent crimes also found no important differences in the ranking of values. Although there are some difficulties with the data used in

these studies,4 they are the only re- cent materials which attempt to measure directly value hierarchies; and they yield findings incompatible with the subculture of violence thesis.

Attitudinal data collected for the President's Commission on the Causes and Prevention of Violence in 1968 also call the subcultural thesis into question. In a national survey, for which questionnaire construction was supervised by Ball-Rokeach, respondents were asked about their general approval of the use of physical aggression in certain kinds of interpersonal in- teractions; those who gave this gen- eral approval were then asked about four or five more specific situations. The general approval questions asked whether there were "any situ- ations that you can imagine" in

machismo is clearly present, violence is often criticized by the family. Most of the family members feel hurt by the violence and depri- vation they experienced as children, and many resolve to do better with their children. In an earlier discussion of poverty in Mexico, Lewis (1961) lists "frequent resort to vio- lence" as an element of the culture of pover- ty, while in La Vida (1966) he talks more generally of lack of impulse control. In neither case does he say that the culture re- quires acts of violence.

3 One relevant study which combines class data in this way is Schneider and Lysgaard's (1953) work on the "deferred gratification pattern." Although the findings of class dif- ferences are open to criticism (see Miller, 1965), note that at any rate the differences in the use of physical violence were small and were considered unimpressive by the au- thors.

4 On the dependent variable, there are weaknesses in the indicators of the values

supporting violence. "An Exciting Life," "Pleasure," "Social Recognition," and being "Courageous" are examples of the indicators of the machismo concept; yet the phrases ac- companying these and each of the other value choices suggest a very broad interpretation, e.g., "a stimulating and active life;" "an en- joyable, leisurely life;" "respect, admiration;" "standing up for your beliefs." On the inde- pendent variable, in the national study, the "degree of participation in violence" includes both aggression and victimization, is based on the variety rather than the extent of experi- ence, and weighs childhood incidents equally with any recent ones. Even if a respondent's aggression were being estimated with some accuracy, the violence may have occurred long before the contemporary value patterns were established. Also the studies apparently do not control for age or race.

This content downloaded from 141.210.130.85 on Mon, 04 Apr 2016 17:26:33 UTC All use subject to http://about.jstor.org/terms

Subculture of Violence 283

which the respondent would ap- prove of such acts as a husband slapping his wife's face; a husband shooting his wife; a man punching (or choking) an adult male stranger; one teenage boy punching (or knifing) another. Because these items and their follow-ups are so general, acceptance of them does not imply membership in a subcul- ture of violence. But conversely, it seems reasonable to assume that

persons who are in such a subcul- ture would find it quite easy to sup- port many of the items, especially those dealing with relatively minor forms of violence. If levels of sup- port in low-status groups are rela- tively low, then the finding can be taken as suggestive evidence con- trary to the thesis.5

Preliminary analysis of these data has been reported elsewhere (Baker and Ball, 1969; Stark and McEvoy, 1970). The present author has un- dertaken a detailed analysis of these data, using cross tabulation and mul- tiple regression. My analysis does not alter the basic preliminary findings, which showed an absence of major differences by race or class6 in approval of interpersonal violence, and in general a low rate of approval. For example, marital fighting is often thought to be a characteristic of the "subculture of violence," but when approval of a husband slapping his wife's face is examined, only 25 percent of white and 37 percent of black married

men aged 18-60 say that they can imagine any situation in which they would approve, with no systematic variation by income or education. (There is an age effect, with men over 40 being sharply lower in ap- proval, but it is independent of race, education, or income.) Moreover, both the level of support and the variation by race decrease markedly when follow-up items are examined. A similar pattern is found for items relating to approval of a man chok- ing an adult male stranger; while on items relating to punching an adult male stranger, approval by whites is higher than that by blacks.

Attitudes toward machismo can

be gauged by an index made up of items relating to approval of teenage fighting. The items on this index seem to be very easy to support- "Are there any situations you can imagine in which you would ap- prove of a teenage boy punching another teenage boy?" If yes, or not sure, "would you approve if he didn't like the other boy?" ... "if he had been ridiculed and picked on by the other boy?" ... "if he had been challenged by the other boy to a fist fight?" ... "if he had been hit by the other boy?" The index was con- structed by scoring a yes response to each of the five items as 2, a not sure as 1, and a no as 0. The range is thus 0-10.

Whites tend to score higher than blacks on this index; and when par- ents with at least one teenage child are analyzed separately, only 12 percent of black parents, compared to 38 percent of white parents, score above six on the ten point in- dex. Among whites, parents with low-income score lower than those

with high-income. If a subculture of violence existed

5 Of course, this does not mean that a per- son's response to the general item directly indicates his attitude or action in some actual

instance in which he may become (or have been) involved in.

6 Social class is indicated by income and education. Occupational data were not coded.

This content downloaded from 141.210.130.85 on Mon, 04 Apr 2016 17:26:33 UTC All use subject to http://about.jstor.org/terms

284 SOCIAL PROBLEMS

among low-status adults, or if low- status adults valued the expression of violence among their children, the general trend on this index would be expected to be the reverse of that found, and the rate of sup- port at the high end of the index would have been much higher. The data and conclusions say nothing about the extent of fighting among lower-class or black teenagers; and the questions of unintentional so- cialization through the latent effects of parental behavior, or of socializa- tion to violence by teenage peers, remain open. It may well be that lower-class or black teenagers are involved in a disproportionate number of fights, and the lower rate of approval by their parents could be a result of the frequency or seri- ousness of these fights. But such a situation would only support the conclusion that lower-class parents in general, and black parents in par- ticular, do not especially like the idea of their children fighting and that teenage fighting is probably not a product of an adult value system emphasizing violence.

Some New Data: Peer Esteem and

Psychological Correlates of Fighting

In addition to the investigation of verbal support for a "subculture of violence," support and sanction in peer interactions can be examined. Wolfgang and Ferracuti (1967:160) argue that nonviolent members of a subcultural group are subject to great pressure to conform, that sanction is an integral part of the existence of a norm, and that "alien- ation of some kind ... seems to be a

form of punitive action most feasi- ble to this subculture." It seems to

follow that, conversely, persons

who adhere to the values would be

more likely than those who do not to be liked, respected, and accorded high status in the group. Data from a 1969 survey of black and white males aged 21-64 in Milwaukee, Wisconsin, give some evidence on this point.7 Physical aggression is indicated by the item "How often do you get in angry fist fights with other men?" (never, almost never, sometimes, often); perceived es- teem accorded by others is indicated by two items, "How do you com- pare with most men you know on being respected and listened to by other people?" (five point code, from much worse to much better) and "How do you compare with most men you know on being well liked by other people and having lots of friends?" (same code). Since the esteem items are double-

barreled, they are less precise than desirable. However they are useful for exploratory purposes.

Because the subcultural hypo- thesis posits statistical interaction, separate analyses were made for the "lower class" (income less than $5,000) and "nonpoor" (income over $5,000), and for blacks and whites. As a result, low-income whites have a small sample size and detailed analysis cannot be carried out for this subsample.

The bottom row of Table 1 shows

that the pattern of fighting by race

7 The data are from an ongoing study of correlates of self-esteem directed by Russell Middleton (sponsored by the National Sci- ence Foundation); I am grateful to him for permission to analyze and report the relation- ships presented here. The interviews were conducted by the Wisconsin Survey Research Laboratory and respondents and interviewers were matched by race. An area probability sample for the Milwaukee city limits was used.

This content downloaded from 141.210.130.85 on Mon, 04 Apr 2016 17:26:33 UTC All use subject to http://about.jstor.org/terms

Subculture of Violence 285

TABLE 1

EFFECT OF FIGHTING ON FEELING "RESPECTED AND

LISTENED TO BY OTHERS"

By race and income, with controls Milwaukee men, aged 21-64, 1969

White Black

<$5,000 >$5,000 Total <$5,000 >-$5,000 Total

Hypothesized Relation Positive Strong Neg. Strong Pos. Positive

Zero order r -.15 .02 -.25 -.18* Beta, net of social desirability index (SDI) -.05 .02 -.22 -.18*

Beta, net of SDI, occupation and age a .08 .02 -.14

(N) (15) (207) (222) (51) (184) (235) % who fight 47% 19% 21% 39% 29% 31%

a Because of small sample size, this coefficient cannot be computed. * Indicates beta significant at .05 or better.

and income group is consistent with the subcultural thesis; blacks are more likely to fight than whites, and the poor are more likely to fight than the nonpoor.8 (Contrary to ex- pectations, poor whites are more likely to fight than poor blacks, but the percentage for whites is unreli- able because of the low N.) How- ever, this pattern is also consistent with several other non-subcultural

theories, such as those of Henry and Short (1954), Coser (1963), Gold (1958), or Cloward and Ohlin (1960). The important question here is whether men who fight are accorded (or at least see themselves as being accorded) more esteem by others.

Although the subculture of vio- lence thesis does not make a predic- tion about the overall association

between race or economic status

and peer respect or high status among peers, it predicts that the basis of the respect of status will be different in different groups. Sub- cultural theory would seem to pre- dict a relatively strong positive cor- relation between the peer esteem item and fighting for low-income blacks, a somewhat smaller (but at least statistically significant) positive correlation for low-income whites and nonpoor blacks,9 and a rela- tively strong negative correlation for nonpoor whites.

Table 1 shows the relationship between fighting and perceiving "respect by others," in terms of zero order correlations and as the net ef-

fect (beta) of fighting on perceived esteem by others, controlling first for social desirability bias10 and then

8 The differences by class and race re- ported for the samples here are larger than those found on items in the Violence Com-

mission survey, which asked retrospectively about acts of physical aggression. See Baker and Ball (1969) or Stark and McEvoy (1970).

9 An alternative prediction would be that, because of strong norms against violence among the "black bourgeoisie," the correla- tion between violence and esteem would be negative at least for those nonpoor blacks in white-collar jobs.

10 "Social desirability bias" is indicated by

This content downloaded from 141.210.130.85 on Mon, 04 Apr 2016 17:26:33 UTC All use subject to http://about.jstor.org/terms

286 SOCIAL PROBLEMS

for social desirability bias, occupa- tion, and age.11 The findings are in- consistent with the predictions out- lined, with the betas and zero order correlations being either very close to zero or having a sign opposite that predicted.12 Table 2 shows the relationship between fighting and perceptions of being "liked by others," in terms of zero order cor- relations and the net effect of

fighting on perceived esteem. Here the findings are somewhat as pre- dicted by subcultural theory, with low-income blacks and low-income

whites showing a positive net effect of fighting on perceived esteem. But the former beta is rather small; and although the latter is larger, neither of them is statistically significant. Moreover, for nonpoor white men, the predicted strong negative corre- lation does not appear.

Although the findings here do not refute the subculture of violence thesis, taken as a whole they cast doubt on it. To the extent that vio-

lence is important to low-income or black men, and to the extent that a subcultural norm is being enforced through ostracism or peer rebuke, we would expect to find a relatively strong positive relationship between fighting and perceived general es- teem. Similarly, if a counternorm of nonviolence is important in the white middle class, a strong negative relationship should have been found. Overall the data here are not

consistent with this predicted pat- tern; and if we take statistical sig- nificance as a minimal criteria of

support, none of the predictions of subcultural theory is supported. It is possible, of course, that the avail- able indicators mask the relation-

ships predicted. For example, perhaps responses to fighting draw approval or rebuke as predicted, but these responses do not affect the overall evaluation perceived by the violent person. In this case, how- ever, we would have to conclude that violence is not as important to the subculture as hypothesized, for as the sanction gets stronger-e.g., ostracism-consequences for gen- eral esteem should follow.

As a corollary to the analysis of violence and esteem, the relation- ship between violence and feeling of well being can be examined. The subcultural thesis holds that vio- lence is normal behavior and is the product of normal group processes. Similarly, it posits that violent people do not feel guilty about their actions. An empirical inquiry could examine psychiatric records or ad- minister various personality tests (see, e.g., Ferracuti, Lazzari, and

a five item adaptation of Crowne and Mar- low's (1964) scale, which includes items which are either socially desirable but proba- bly untrue or probably true but socially unde- sirable. (For example, True or False: "I never hesitate to go out of my way to help someone in trouble.") Of the, five items, three were worded such that agreement was socially de- sirable, and two worded such that disagree- ment was socially desirable. Respondents scoring high on this scale are somewhat more likely to report that they do not get in fights and that they are held in high esteem by others.

11 Both occupation and age were indicated by sets of dummy variables. For occupation, the categories were white collar, blue collar, farm; for age they were 21-25, 26-30, 31-35, 36-45, 46-55, 56-64.

12 A competing interpretation of the zero and near zero relationships in Tables 1 and 2 is that they mask a strong reciprocal causality. The possibility of such reciprocal causality, which would be consistent with the subcul- tural thesis, has been explored for blacks using two stage least squares analysis (Er- langer and Winsborough, 1975). The results of that analysis support the conclusion in the text.

This content downloaded from 141.210.130.85 on Mon, 04 Apr 2016 17:26:33 UTC All use subject to http://about.jstor.org/terms

Subculture of Violence 287

TABLE 2

EFFECT OF FIGHTING ON FEELING "WELL LIKED BY OTHER

PEOPLE AND HAVING LOTS OF FRIENDS"

By race and income, with controls Milwaukee men, aged 21-64, 1969

White Black

<$5,000 >$5,000 <$5,000 >$5,000

Hypothesized Relation Positive Strong Neg. Strong Pos. Positive

Zero order r .22 -.01 -.12 -.01

Beta, net of social desirability index (SDI) .21 -.01 -.09 -.01

Beta, net of SDI, occupation and age a -.03 .16 -.04

(N) (15) (207) (51) (184)

a Because of small sample size this coefficient cannot be computed. NB: None of the coefficients in the table are significant at .05 or better.

Wolfgang, 1970); alternatively, var- ious measures of psychological ad- justment can be included in an in- terview schedule or questionnaire. One such measure is an index of happiness which can be constructed from items in the Milwaukee sur-

vey.13 It would seem that outside the subculture men who are violent

would be less likely to be happy than would non-violent men, both because they were receiving nega- tive sanctions for their violence and because in this group it would be the more marginal men who would be violent. By contrast, within the subculture, happiness would be positively correlated with violence, since violence is posited as not being a pathological condition and since non-violent men are hy-

pothesized to be negatively sanctioned. Table 3 shows that fighting is negatively correlated with happiness for all four subgroups, and (statistically) significantly so for blacks. Except for nonpoor whites, these findings run directly counter to the predictions of the subcultural thesis. And even for nonpoor whites, the finding of a correlation even less negative than for blacks can also be considered evidence

contrary to the thesis.14

DISCUSSION

Although much suggestive evi- dence on the subculture of violence exists, there is a clear need for further research in this area. Me-

thodologically, this research should be designed so that there is adequate representation of minori- ties and of poor whites for analysis; 13 The items were these: "On the whole,

how happy would you say you are now?" "On the whole, how happy would you say you are now compared with other men you know?" "How often do you feel very discouraged and depressed?" "How often do you get the feel- ing that life is not worth living?" Each item had four possible responses.

14 Because of the uncertain direction of causation, partial correlation coefficients may be a more appropriate measure of association here than regression coefficients. However, use of partial r's would not have changed the findings.

This content downloaded from 141.210.130.85 on Mon, 04 Apr 2016 17:26:33 UTC All use subject to http://about.jstor.org/terms

288 SOCIAL PROBLEMS

TABLE 3 RELATIONSHIP BETWEEN FIGHTING AND HAPPINESS

By race and income, with controls Milwaukee men, aged 21-64, 1969

White Black

<$5,000 I>$5,000 <$5,000 >$5,000

Hypothesized Relation Positive Strong Neg. Strong Pos. Positive

Zero order r -.09 -.07 -.36* -.19" Partial r, net of social desirability index (SDI) .08 -.07 -.33* -.19"

Partial r, net of SDI, occupation and age a -.09 -.30* -.18"

(N) (15) (207) (51) (184)

a Because of small sample size this coefficient cannot be computed. * Indicates partial r significant at .05 or better.

and it should make some attempt to cover both "streetcorner men" and more traditional householders (cf. Hannerz, 1969). A major limitation of existing survey data is that they are based only on persons in house- holds (cf. Parsons, 1972). Another is that the surveys do not have con- centrated samples in a given neighborhood. These difficulties are alleviated, but not erased, by the data from the field studies.

Substantively, work needs to be done on establishing the pervasive- ness of a subculture. At least three

quite different degrees of perva- siveness may exist. In the most ex- treme case, a large majority of the demographic group presumed to have the subcultural trait would exhibit it in some way, as opposed to a minority of members of other demographic groups. In this case, one could characterize the demo-

graphic group as subcultural. A more limited pervasiveness exists when the trait is exhibited by a minority of a particular demo- graphic group, as compared to a vir- tual absence in other groups; this

would constitute a subculture within a demographic group. Final- ly, some analysts would consider a small but statistically significant dif- ference between demographic groups as evidence of a subculture. However, it seems incorrect to characterize such a subculture as

being located in the demographic group with greater support for the value, or to characterize that de- mographic group as a subculture. Rather, the subculture would have to be defined as the group of people who hold the value, irrespective of their demographic group.15

These differences in the perva- siveness of subcultures have impor- tant implications for the public im- agery of social groups. In the case of the subculture of violence, if class or racial groups can in fact be charac- terized as being different (or if

15 Also, as Rodman (1963) suggests, this case might be better understood as one of a variation on a common cultural theme, not as a tension between the values of a dominant

culture and a subculture. Of course, even if such a difference is not considered subcul-

tural, it may still be descriptively interesting and important in the explanation of violence.

This content downloaded from 141.210.130.85 on Mon, 04 Apr 2016 17:26:33 UTC All use subject to http://about.jstor.org/terms

Subculture of Violence 289

findings are presented as though they could), popular conceptions of widespread pathology among non- whites and low-income whites would be supported. By contrast the existence of a subculture within a class or racial group, or of value dif- ferences that are statistically sig- nificant but not large, would be more consonant with the view that

there is wide variation in the values, needs, and problems of the poor and of nonwhites.

Future research should also focus more closely on the precise content of supposed subcultural differ- ences.'6 It is possible, for example, that rather than a "subculture of vio- lence," something like a "subculture of masculinity" exists, with violence being only one of many possible outlets, and not necessarily the pre- ferred one. In this case, violence may result from the blocking of al- ternative opportunities to exhibit "machismo" (cf. Miller, 1966). Another possibility is that the use of liquor may be part of a broader so- cial configuration which generates situations conducive to violence. A value system which sanctions or even encourages either drunken brawls or wild behavior on certain special occasions would not neces- sarily be the same as one which re- quires "quick resort to physical combat as a measure of daring, courage, or defense of status" in everyday interaction.17

Finally, the origins, permanence, and relationship to social structure must also be given careful consider- ation in future research. These con- siderations are especially important in the formation of social policy (cf. Banfield, 1968; Lewis et al., 1969; Liebow, 1971; Valentine, 1968).

CONCLUSION

Although the subculture of vio- lence thesis has received a certain measure of acceptance in the field, a wide variety of evidence suggests that it is questionable. All of the data available have limitations of various sorts, and the thesis cannot be said to have been definitively tested. On balance, however, more of available evidence is inconsistent with the thesis than consistent with it.

At this time we do not know how important a deviant value system is in explaining violence in the United States;"s and, if it exists, we do not know whether such a value system

16 This can probably best be done by be- ginning with relatively unstructured in-depth interviews with informants. A move in this direction is made by Toch (1969), who conducted intensive interviews with both convicts and policemen who had frequently engaged in assault. But even here the subcul- tural thesis is drawn from the literature rather

than grounded in the accounts of those inter- viewed.

17 Similar considerations hold for the

question of the existence of a subculture of violence in the American South. Many writ- ers have noted the quite disproportionately high rate of homicide in the South, and re- cently Gastil (1971), Hackney (1969), and Reed (1972) have argued that this divergence can be explained by regional differences in the acceptability of violence. But again, the exact content of the hypothesized subculture is generally unclear, and data do not support the application of the subculture of violence thesis to the South (Erlanger, 1975).

18 This analysis does not address Wolfgang and Ferracuti's contention that a subculture of violence exists in Colombia, Sardinia, Mexico, Albania, and Albanova, Italy. The case of Sardinia is explored in more detail in Ferracuti et al. (1970), who find some evi- dence in support of the hypothesis but con- cludes that "the subculture of violence in Sardinia is limited to violent offenders" (1970:110). This suggests that although it may be that violent offenders in Sardinia re- ceive support for their actions from a limited group, Sardinia itself cannot be characterized as embracing a subculture of violence.

This content downloaded from 141.210.130.85 on Mon, 04 Apr 2016 17:26:33 UTC All use subject to http://about.jstor.org/terms

290 SOCIAL PROBLEMS

can be said to be found predomi- nantly within the black or low- income white communities or whether it can be said to be rela-

tively independent of social struc- ture. But there is enough evidence to conclude that these groups are not characteristically different from the dominant society in their rate of approval of the use of physical ag- gression. This conclusion, along with a growing empirical literature on other aspects of the lives of poor and black (and other minority) per- sons in the United States, is compat- ible with the view that the social and economic deprivations experienced by members of these groups are primarily the result of social struc- tural factors, rather than the product of group pathology (cf. Goodwin, 1972; Institute for Social Research, 1974; Kriesberg, 1970; Shiller, 1973).

REFERENCES

Akers, Ronald L. 1973 Deviant Behavior. A Social Learn-

ing Approach. Belmont: Wads- worth Publishing.

Amir, Menachem 1971 Patterns in Forcible Rape. Chicago:

University of Chicago Press.

Baker, Robert K. and Sandra J. Ball 1969 Mass Media and Violence; Report

to the National Commission on Causes and Prevention of Vio- lence. Washington, D.C.: U.S. Government Printing Office.

Ball-Rokeach, Sandra J. 1973 Values and Violence: A Test of the

Subculture of Violence Thesis. American Sociological Review 38:6 (December) pp. 736-49.

Banfield, Edward C. 1968 The Unheavenly City. Boston: Lit-

tle, Brown. Clinard, Marshall B.

1973 Sociology of Deviant Behavior. New York: Holt, Rinehard, and Winston.

Cloward, Richard A. and Lloyd Ohlin 1960 Delinquency and Opportunity.

New York: The Free Press. Coser, Lewis A.

1963 "Violence and the social structure." Science and Psychoanalysis 6:30- 42. Reprinted in Shalom Endleman (ed.), Violence in the Streets. Chicago: Quadrangle (1968).

Crain, Robert L. and Carol Sachs Weisman 1972 Discrimination, Personality, and

Achievement. New York: Seminar Press.

Crowne, Douglas P. and David Marlowe 1964 The Approval Motive. New York:

John Wiley. Ferracuti, Franco, Renato Lazzari, and

Marvin E. Wolfgang 1970 "A Study of the Subculture of Vio-

lence Thesis," in Ferracuti et al., Violence in Sardinia. Rome: Mario

Bulzoni, editor. Gastil, Raymond D.

1971 "Homicide and a regional culture of violence." American Sociologi- cal Review 36(June): 412-27.

Gold, Martin 1958 "Suicide, homicide, and socializa-

tion of aggression." American Journal of Sociology 63(May): 651-61.

Goodwin, Leonard 1972 Do the Poor Want to Work?

Washington, D.C.: Brookings In- stitution.

Hackney, Sheldon 1969 "Southern Violence," in Hugh

Davis Graham and Ted Robert Gurr (eds.), History of Violence in America. Report of the Task Force of the President's Commission on Causes and Prevention of Vio- lence. New York: Bantam Books, 505-28.

Hannerz, Ulf 1969 Soulside: Inquiries into Ghetto

Culture and Community. New York: Columbia University Press.

Henry, Andrew and James F. Short 1954 Suicide and Homicide. New York:

The Free Press.

Hirschi, Travis 1969 Causes of Delinquency. Berkeley

and Los Angeles: The University of California Press.

Institute for Social Research

1974 The Changing Status of Five Thousand American Families:

This content downloaded from 141.210.130.85 on Mon, 04 Apr 2016 17:26:33 UTC All use subject to http://about.jstor.org/terms

Subculture of Violence 291

Highlights from the Panel Study of Income Dynamics. Mimeo. Ann Arbor: The University of Michi- gan.

Jansyn, Leon R., Jr. 1966 "Solidarity and delinquency in a

street corner group." American Sociological Review 31(October): 600-14.

Kobrin, Solomon, Joseph Puntil, and Emil Peluso

1967 "Criteria of status among street groups." Journal of Research in Crime and Delinquency 4(Janu- ary).

Kriesberg, Louis 1970 Mothers in Poverty: A Study of

Fatherless Families. Chicago: Al- dine.

Lerman, Paul 1968 "Individual values, peer values, and

subcultural delinquency." Ameri- can Sociological Review 33(April): 219-35.

Lewis, Oscar 1961 Children of Sanchez. New York:

Vintage. 1966 La Vida: A Puerto Rican Family in

the Culture of Poverty. San Juan and New York: Random House.

1969 et al., Review Symposium: Culture and Poverty (Charles A. Valen- tine). Current Anthropology 10(April, June): 189-201.

Liebow, Elliot 1967 Talley's Corner. Boston: Little,

Brown.

1971 "Comment on Miller's Paper," pp. 131-36 in J. Alan Winter (ed.), The Poor. Grand Rapids, Michigan: W. B. Eerdmans.

Matza, David 1964 Delinquency and Drift. New York:

John Wiley.

Miller, S. M., Arthur Seagull, and Frank Riessman

1965 "The Deferred Gratification Pat- tern: A Critical Appraisal," in Louis Ferman et al., Poverty in America. Ann Arbor: University of Michigan Press.

Miller, Walter B. 1958 "Lower class culture as a generating

milieu of gang delinquency." Jour- nal of Social Issues 14(Sum- mer): 5-19.

1961 et al., "Aggression in a boy's

street-corner group." Psychiatry 24(November): 283-98.

1966 "Violent crimes in city gangs." An- nals of American Academy of Po- litical and Social Science 364(March): 96-112.

Mulvihill, Donald J. and Melvin Tumin 1969 Crimes of Violence. Staff Report

to the National Commission on Causes and Prevention of Vio- lence. Vols. 11, 12, and 13. Washington, D.C.: U.S. Govern- ment Printing Office.

Parsons, Carole W. (ed.) 1972 America's Uncounted People.

Washington, D.C.: National Acad- emy of Sciences.

Reed, John S. 1972 The Enduring South: Subcultural

Persistence in Mass Society. Lexington, Massachusetts: D. C. Heath.

Rodman, Hyman 1963 "The lower class value stretch."

Social Forces 42(December): 205- 15.

Schneider, Louis and Sverre Lysgaard 1953 "The deferred gratification pattern:

A preliminary study." American Sociological Review 18(April): 142-49.

Schur, Edwin 1969 Our Criminal Society. Englewood

Cliffs, New Jersey: Prentice-Hall. Schiller, Bradley R.

1973 "Empirical studies of welfare de- pendency: A survey." Journal of Human Resources 8(Supple- ment): 19-32.

Short, James F. and Fred L. Strodtbeck 1965 Group Process and Gang Delin-

quency. Chicago: University of Chicago Press.

Stark, Rodney and James McEvoy III 1970 "Middle class violence." Psychol-

ogy Today 4(November): 52-4, 110-12.

Suttles, Gerald D. 1968 The Social Order of the Slum.

Chicago: University of Chicago Press.

Toch, Hans H. 1969 Violent Men. Chicago: Aldine.

Valentine, Charles A. 1968 Culture and Poverty-Critique and

Counter-Proposals. Chicago: Uni- versity of Chicago Press.

This content downloaded from 141.210.130.85 on Mon, 04 Apr 2016 17:26:33 UTC All use subject to http://about.jstor.org/terms

292 SOCIAL PROBLEMS

Whyte, William F.

1955 Street Corner Society. Chicago: University of Chicago Press.

Wolfgang, Marvin E.

1958 Patterns of Criminal Homicide. Philadelphia: University of Penn- sylvania Press.

1967 and Franco Ferracuti. The Subcul- ture of Violence, Towards an In- tegrated Theory in Criminology. London: Tavistock-Social Science Paperbacks.

Yablonsky, Lewis 1962 The Violent Gang. New York:

Macmillan.

LABELING, DETERRENCE, AND RECIDIVISM: A STUDY OF POLICE DISPOSITIONS OF

JUVENILE OFFENDERS*

MALCOLM W. KLEIN

University of Southern California

Police departments with high rates of juvenile diversion did not yield different re- cidivism rates than those with low diversion rates unless comparisons were made be- tween first offenders and multiple offenders. Then, the two sets of departments differed substantially: high diversion departments had lower subsequent recidivism rates for first offenders than for multiple offenders, while low diversion departments did not exhibit such differences. The emergence of differences in multiple but not in simple recidivism and among high diversion but not among low diversion departments tentatively sup- ports both labeling theory and a deterrence approach in interaction with departmental and offender variables. There is also a suggestion that the effects of delinquent stigmati- zation are cumulative with each arrest, supporting Lemert's secondary deviance concep- tion, at least among first offenders in high diversion departments.

INTRODUCTION

The past several years have seen a veritable explosion of publications dealing with labeling theory, recent summaries of which make a review here unnecessary (Gove, 1970; Schur, 1971). Within the mental health and criminology fields, the emphasis on labeling has taken on an applied rhetoric around the concept of "diversion," i.e., the mechanics of reducing client and offender insertions into the formal

organizational systems of mental health and criminal justice (Lemert,

1971; NIMH, 1971; Klein, 1973; Vorenberg and Vorenberg, 19731.

Significantly, however, the mate- rials appearing to date are almost entirely theoretical, conceptual, speculative, or exhortatory in the absence of-and often without re-

gard for-empirical data which might describe, investigate, or test labeling and diversion propositions against their counterparts in the lit- erature on the deterrent effects of sanctions. The bulk of the criminal justice data pertinent to these issues falls roughly into three areas. First, there are the studies exemplified by Schwartz and Skolnick (1964), by Buikhuisen and Dijksterhuis (1971), and by Rosenthal and Jacob- son (1968), which confirm that labeling can indeed affect the re-

* Revised from a paper delivered at the 1972 ASA meetings, New Orleans. I am very grateful to Daniel Glaser, Solomon Kobrin, and Carol A. B. Warren for their critiques of an earlier draft.

This content downloaded from 141.210.130.85 on Mon, 04 Apr 2016 17:26:33 UTC All use subject to http://about.jstor.org/terms