Q3: Summarize the primary argument of the Richardson and Weatherby article “Belief in an Afterlife as Symbolic Sanction.” How persuasive is the analysis? Cite at least one strength and one weaknes

Religious Research Association, Inc. Belief in an Afterlife as Symbolic Sanction Author(syf - R K Q * 5 L F K D U G V R Q D Q G * H R U J L H $ : H D W K H U E y Source: Review of Religious Research, Vol. 25, No. 2 (Dec., 1983yf S S 9 Published by: Religious Research Association, Inc. Stable URL: https://www.jstor.org/stable/3511493 Accessed: 02-05-2019 08:06 UTC REFERENCES Linked references are available on JSTOR for this article: https://www.jstor.org/stable/3511493?seq=1&cid=pdf-reference#references_tab_contents You may need to log in to JSTOR to access the linked references. 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]. Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at https://about.jstor.org/terms Religious Research Association, Inc. is collaborating with JSTOR to digitize, preserve and extend access to Review of Religious Research This content downloaded from 69.41.96.244 on Thu, 02 May 2019 08:06:45 UTC All use subject to https://about.jstor.org/terms 162 BELIEF IN AN AFTERLIFE AS SYMBOLIC SANCTION John G. Richardson Georgie A. Weatherby Western Washington University Review of Religious Research, Vol. 25, No. 2 (December 1983yf . This paper examines the correlates of belief in an afterlife. Afterlife is theoreti- cally proposed as representing a symbolic sanction and relevant to the perception of negative consequences accompanying the violation of traditional marital and family behaviors. Data taken from the NORC General Social Survey for 1978 are tested utilizing discriminant function analysis. The discriminant analysis of attitu- dinal variables is compared to earlier research which found religious affiliation and church attendance to be the main predictors of belief in heaven and the afterlife. This study finds a prohibitive attitude toward premarital sexuality, abor- tion, and toward divorce laws to be significant predictors. Their meaning in relation to religious affiliation and church attendance is discussed. In his now classic Protestant, Catholic and Jew, Herberg (1955yf Q R W H G W K H L U R Q \ L n American religious life when he suggested that increasing secularization may not necessarily mean a corresponding decline in religiosity. In his empirical study of the religious factor, Lenski (1963:59yf V X F F L Q F W O \ G H I L Q H G W K L V L U R Q \ $ P H U L F D Q V D U H E H - coming more religious while at the same time becoming more secular. Transcendental faith is gradually being transformed into cultural faith." Lenski's empirical study tended to affirm Herberg's observation by demonstrating that church attendance, while one index of religiosity, may increasingly decline as a determinant of "doctrinal orthodoxy." More recently, Bellah (1975yf K D V W K H R U H W L F D O O \ L Q W H U S U H W H G V X F K H Y L G H Q F H R I L Q F R Q V L V - tency" as reflecting the broader, historical tension between "biblical religion" and "utilitarian individualism." Nonetheless, examples of inconsistencies of behavior and belief illustrate well the relative strength of a religious interpretation resting on "con- science" or a utilitarian interpretation resting on the maximization of personal "in- terest." Recently, Hertel (1980yf H [ S O R U H G R Q H V X F K H [ D P S O H R I L Q F R Q V L V W H Q F \ R I E H O L H I L Q a discriminant analysis of stance toward heaven and the afterlife. The strongest predictor was found to be Protestant and Catholic affiliation. In testing a number of social- demographic variables, none was found to be significant. It is likely that the strong discrimination produced was derived from the strong association between religious affiliation and belief in both heaven and afterlife. Thus, the percent of cases correctly predicted was 96 percent for both, 80 percent for heaven only, and then drops to 46 percent for belief in an afterlife only, and a low of 29 percent for belief in neither. Evidently the combination of heaven and afterlife represents a predictable social category for it is bounded by the fact that such believers nearly always hold a religious affiliation. This study reports on a similar empirical analysis of the correlates of belief in an afterlife based on a national probability sample (NORC, 1978yf 7 K H V W X G \ I R F X V H V R Q O y on avowed belief in an afterlife, and is not able to separate out the possible confounding influence of belief in a heaven. This study departs from, yet attempts to extend Hertel's approach by the test of specific attitudinal variables which are considered theoretically relevant to belief in an afterlife. The relevance of these variables is formulated from a theoretical interpretation of the afterlife as a symbolic sanction. This content downloaded from 69.41.96.244 on Thu, 02 May 2019 08:06:45 UTC All use subject to https://about.jstor.org/terms 163 THE AFTERLIFE AS SYMBOLIC SANCTION Beyond the prima facie importance of religious affiliation as a correlate of belief in an afterlife, it is useful to consider the relevance of variables which tap elements of the "cultural faith" noted by Lenski. We may take as the starting point the reminder by Weber (1964: 1yf W K D W W K H V \ P E R O L F F R Q V H T X H Q F H V R I U H O L J L R Q D U H G L U H F W H G W R W K L V Z R U O G . Although the social psychology of a world religion may encompass largely mystical elements and may be oriented to "other-worldly" events, the practical consequences of this orientation hold for behavior conducted within this world. As others have noted (Ross, 1915:130-131; Goode, 1951:50; Skinner, 1953:352-355yf W K H V R F L R O R J L F D O V L J - nificance of supernatural beliefs lies in their power as symbolic sanctions. Belief in an afterlife may be seen as a private, intimidating belief which when held collectively sustains a minimum level of social control. The element of uncertainty about one's salvation which historically linked moral conduct to acquisitive economic behavior (Weber, 1958yf P D \ E H H [ W H Q G H G W R E H O L H I L Q W K H D I W H U O L I H Z K H U H R Q H V P H P E H U V K L S L V V H H n as enhanced through proper conduct in this life. As a form of punishment through superempirical expectations, belief in the afterlife may sanction behavior with consider- able economy. Where features of belief in the afterlife more often catalogue negative than positive consequences, they function to diffuse among believers "the vague feeling that somehow the course of things is against him who breaks the social interdict" (Ross, 1915:129yf , W L V W K H V X E V W D Q F H R I W K H V R F L D O L Q W H U G L F W Z K L F K H [ S O D L Q V W K H E H O L H I Q R W W K e belief which explains the social interdict. A critical question remains: What is the form of the social interdict? As Ross implies, the social interdict is akin to a Durkheimian common conscience which exerts social control by linking individual conduct to perceived inclusion in the organization of collective beliefs. In this manner, belief in the afterlife is an internalization, a "repre- sentation" of the common conscience. Yet, as Durkheim argued (1964:167yf W K e "common conscience consists less and less of strong, determined sentiments," for its evolution rests on the progressive effacement of local arrangements and traditional behaviors.To the extent that the common conscience is antecedent to a belief in the afterlife, we propose that it is an individual's location relative to the direction and impact of major social trends which will shape the meaning of that belief as a symbolic sanction. Groups which find themselves opposed to the major direction of social trends will perceive such trends as secular disturbances of a kind which alter the link between this-worldly conduct and another-worldly destiny. Opposition to these trends will represent both a symbolic and instrumental means whereby one preserves access to an afterlife. To the extent that belief in an afterlife is linked to the anticipation of negative consequences befalling a violation of the social interdict, we may specifically hypothe- size that belief would be associated with a commitment of traditional socially acceptable behaviors. One's commitment to an ethic of social purity and to circumspect conduct in this world would hold practical consequences for one's visualization of an afterlife. Attitudinal positions on matters pertaining to traditional marital and family behaviors may be theoretically most relevant on two grounds. First, such behaviors are most widely institutionalized and culturally reinforced. They are buttressed, in effect, by learned social scripts which are preceded by a prolonged and directed socialization (Laws and Schwartz, 1977yf 6 H F R Q G D V E H K D Y L R U D O V F U L S W V W K H \ Z L O O E H D F F R P S D Q L H d by learned beliefs as to the consequences of their violation. Anticipated sanctions from violations of marital or family behaviors represent the social interdict. This content downloaded from 69.41.96.244 on Thu, 02 May 2019 08:06:45 UTC All use subject to https://about.jstor.org/terms 164 The relevance of a traditional stance toward marital and family behaviors entails two dimensions from which measures are identified. The first is the more proximate arena of marital and family behaviors; the second is the broader role behaviors which are grounded in traditional marital and family relations. Thus, norms of social purity which prescribe circumspect behaviors before and during marriage can be seen as determi- nants of extrafamily roles. A married woman's occupational participation represents major secular trends of recent decades (Oppenheimer, 1977yf 6 L P L O D U O \ F R Q Y H U J H Q F H s of an approving stance toward premarital sexuality is an apparent national secular trend (Mahoney, 1978yf 7 U D G L W L R Q D O V D Q F W L R Q V U H O D W L Q J W R P D U U L D J H D Q G I D P L O \ U R O H V D U H O L Q N H d to broader, societal changes and become, in effect, competitive to these changes. The social interdict as a traditional common conscience is implicated in secular changes which are perceived as consequential for traditional marital and family behaviors. The objective of the empirical analysis which follows is to identify the relative effect of variables as predictors of belief in an afterlife. The analysis separates two sets of variables: one consisting of those tested by Hertel, and the other consisting of specific attitudinal measures designed to tap perceived negative sanctions. DATA AND METHOD Sample The data are taken from the National Opinion Research Center's General Social Survey for 1978. The General Social Surveys (GSSyf D U H Q D W L R Q D O S U R E D E L O L W \ V D P S O H V R f English speaking persons age 18 and above who are in noninstitutional settings in the United States. For the 1-978 survey, the total sample interviewed comprised 1,532 respondents. Measures Dependent Variable. Belief in an afterlife was measured by a simple dichotomy, indexed as (1yf E H O L H I L Q O L I H D I W H U G H D W K D Q G \f no belief. The total responses to the item are skewed, where 70 percent of the respondents answered yes and 21 percent no. Nine percent were undecided or supplied no answers. Independent Variables. The utilization of the GSS makes possible a comparison to the results from Hertel's study where Gallup Poll data were analyzed. Care was taken to include as accurately as possible the same variables included by Hertel in his analysis. These variables are: religious affiliation, church attendance, occupational presitge, educational level, size of hometown, age, sex, race, income and region. In comparison to the model variables reported by Hertel, five attitudinal variables are tested which relate to the conceptualization of an afterlife as a symbolic sanction. Those variables which relate to proximate marital and family behaviors are four: Attitude Toward Premarital Sexuality, indexed (1-4yf D O Z D \ V Z U R Q J W R Q R W Z U R Q J D W D O O ; Attitude Toward Divorce Laws, indexed (1-3yf V K R X O G E H P R U H V W U L F W W R D E R X t right"; Attitude Toward Pornography Laws, index (1-3yf V K R X O G E H O D Z V D J D L Q V t whatever the age" to "there should be no laws against it"; Attitude on Abortion, asking respondents whether or not they thought it should be possible for a pregnant woman to obtain a legal abortion if she is married and does not want any more children, indexed as (1yf Q R \f "yes." This item elicits the most discretionary stance, free of economic or health constraints which are predominantly seen as legitimate grounds for abortion. For This content downloaded from 69.41.96.244 on Thu, 02 May 2019 08:06:45 UTC All use subject to https://about.jstor.org/terms 165 each of these variables, a restrictive stance is taken as signifying a circumspect orientation toward traditional marital and family matters and is specifically hypothe- sized to be linked to belief in life after death. The final variable measures more distant behaviors. This measures respondent's attitudes toward women's work roles. This item asks "Do you approve or disapprove of a married woman earning money in business or industry if she has a husband capable of supporting her?" Responses to this were indexed as (1yf D S S U R Y H \f "disapprove." The item conceptually deals with a stance on the independence of married women and thus measures the willingness to accept a change in or a violation of traditional marital roles. Method The methodological objective here is to identify the predictors of belief in an afterlife. The dependent variable is represented by two nominal categories. The appropriate method under these circumstances is discriminant function analysis. This multivariate technique was employed by Hertel as well and thus provides for a sound comparison. As a form of multiple regression, discriminant analysis identifies the linear composite of variables which best discriminates two or more criterion groups. The merit of discrimi- nant analysis is its ability to identify those variables which make a statistically signifi- cant contribution to the discrimination or separation of the two or more nominal categories. Moreover, the method reveals the relative contribution of variables within the composite of discriminant function. When only two criterion groups are analyzed, the standardized discriminant coefficient may be interpreted in the same way as are beta coefficients in normal regression.2 The analysis measures the effects of those variables given in the Hertel study and the attitudinal variables identified above. Stepwise analyses are performed which identify those variables which make a statistically significant change in Rao's V, a generalized distance measure (Huberty, 1975:562yf Z K L F K J L Y H V D F R Q V H U Y D W L Y H W H V W I R U W K H D E L O L W \ R I a variable to maximally separate the two groups. A .05 level of significance is set. Those variables which do not meet the minimal F value for inclusion are not included in the analysis. RESULTS The results of the discriminant analysis of the two variable sets are given in Table 1. In the analysis of the Hertel set consisting of fourteen variables, eight are significant. Protestant affiliation is the strongest predictor, followed by church attendance. The strength of Protestant affiliation in this analysis is similar to that found by Hertel. The divergence from the Hertel results lies in the significance of race, occupational prestige and the northeastern and north central regions. In this analysis, white respondents express belief in an afterlife significantly more than do black respondents. A cross- tabulation analysis reveals that 78 percent of whites in contrast to only 63.5 percent of black respondents believe in an afterlife. Additionally, it is noteworthy here that occupational prestige is significant and is positively associated with belief in an af- terlife. A cross-tabulation analysis of occupational prestige finds that 83 percent of high prestige levels compared to 73 percent of low prestige levels expressed a belief in an afterlife. Although Hertel found no effect for occupation in the discriminant analysis, it This content downloaded from 69.41.96.244 on Thu, 02 May 2019 08:06:45 UTC All use subject to https://about.jstor.org/terms 166 Table 1 STANDARDIZED DISCRIMINANT FUNCTION COEFFICIENTS FOR BELIEF IN AN AFTERLIFE Variable Coefficient Variable Coefficient Affiliation Protestant .66a Premarital Sex .75a Catholic .54_a holiew .2a Divorce Laws .41a Other .32a Abortion *32a Attendance -.58a Women Working -- Occupation -.25a Pornography - Education .14 Size of Hometown -- Age -- Sex (1--male; 2=femaleyf 9 Income - Race (l=white; 2=nonwhiteyf a Region South -- North Central .12 Northeast .24a RC .369 R .210 X2 129.660 X2 39.950 P < .001 < .001 Group Sizes: Yes = 670 No = 221 aSignificant change in Rao's V (P<.05yf . is worth noting that his cross-tabulations report a demonstrable tendency for the professional, business and executive levels to believe in the afterlife "only,"' in contrast to the unskilled and skilled levels who express a stronger belief in heaven "only" (Hertel, 1980:179yf , W L V H Y L G H Q W W K D W K H D Y H Q D Q G D Q D I W H U O L I H D U H V H S D U D W H U H I H U H Q W V I R r members of different occupational levels. In the analysis of the second variable set consisting of the attitudinal measures, three of the five are significant: premarital sex, divorce laws, and abortion. In the analysis, attitude toward married women working and toward pornography fail to enter the analysis. It is a prohibitive stance toward premarital sexuality which emerges as the strongest predictor. A restrictive stance on divorce laws and on abortion is linked to belief in an afterlife in the predicted direction, and is significant. Although both analyses produce significant discriminant functions, the comparison of This content downloaded from 69.41.96.244 on Thu, 02 May 2019 08:06:45 UTC All use subject to https://about.jstor.org/terms 167 the two shows a greater predictive strength of the first set, with the strength primarily in religious affiliation and church attendance. This largely conforms to the findings reported by Hertel. However, the strength of the three attitudinal variables, as demon- strated in the second analysis, suggests that they may have independent effects when included in an analysis containing the first variable set. To test this, all significant variables from both sets were entered simultaneously in a stepwise discriminant analy- sis. The results of this analysis did not alter the pattern found in the first variable set: the attitudinal variables failed to enter the analysis and religious affiliation and church attendance remained the significant predictors. As a consequence of this, a final analysis was performed which tested the possibility that the attitudinal variables and the varia- bles of religious affiliation and church attendance may be measuring common portions of the variance of the dependent variable. This is suggested by the results of zero order relations between the significant attitude variables and church attendance. The coeffi- cients for premarital sexuality, abortion, and divorce laws are r = -43, -.31, and -. 13 respectively. Clearly, a greater attendance at church is significantly related to a prohibi- tive stance on premarital sexuality and moderately related to a prohibitive stance on divorce laws and abortion. To test this possibility, the significant variables of the first set were removed one at a time and the attitudinal variables were then included in the analysis. A stepwise procedure was performed as a means to test whether one or more of the attitudinal variables would enter into the analysis and reach significance after one of the variables in the prior set was removed. This analysis revealed important results not evident by the simultaneous inclusion of all variables.When church attendance is removed, all three attitudinal variables enter the analysis and premarital sex is significant (P less than .001yf D V Z H O O D V G L Y R U F H O D Z V P less than .05yf 0 R U H R Y H U W K H F D Q R Q L F D O F R U U H O D W L R Q L V R Q O \ V O L J K W O \ U H G X F H G U H D F K L Q J 5 c = .35. For the remaining variables, only the removal of race permitted the inclusion of divorce laws and abortion, with divorce laws alone reaching significance (P less than .05yf . The results of this final analysis reveal that church attendance, and a strict stance on premarital sexuality and on divorce laws, contain similar information about belief in an afterlife. Moreover, it is this behavioral variable and not simply religious affiliation which suppresses the effects of attitude toward premarital sexuality and divorce laws. In addition, the results reveal that a strict stance on divorce laws is systemically related to race. The inclusion of divorce laws in an equation containing race will not reveal its significance. The finding points to the historical importance of different marital and divorce patterns between white and nonwhite families, and their relation to religious affiliation. At minimum, the finding is consistent with research which emphasizes the stronger link between marital stability and religious affiliation for whites, whereas for blacks the causes and consequences of marital instability are "virtually unrelated to religious participation" (Glenn and Gotard, 1977:450yf . DISCUSSION AND CONCLUSIONS The results of this analysis find that belief in an afterlife is a consequence of both a doctrinal affiliation and of behavioral attendance at church. These findings concur This content downloaded from 69.41.96.244 on Thu, 02 May 2019 08:06:45 UTC All use subject to https://about.jstor.org/terms 168 essentially with those of Hertel. In addition, this analysis finds occupational prestige to be a significant predictor and to be positively related to belief in life after death. This contrasts with the inverse relation for education. While a higher education decreases the probability of belief in an afterlife, a higher occupational level increases this probability. The significance of this measure, independent as it is of prohibitive stances toward traditional marital and family relations, may tap those who are more favorably situated relative to major secular trends and for whom belief in an afterlife represents a positive affirmation of this-worldly success.3. They are those for whom the "broad cultural shift" responsible for some decline in church attendance and in adherence to traditional behaviors "was in the direction of greater individualism, personal freedom, tolerance of diversity and distance from many traditional institutions" (Hoge and Roozen, 1979:328; see also Wuthnow, 1976yf ) X U W K H U P R U H D W H V W R I W K H D V V R F L D W L R Q E H W Z H H n occupational prestige and church attendance finds the two to be statistically indepen- dent, a point which enhances the independent effect of occupational prestige. This analysis has attempted to further specify the interpretation of a belief in an afterlife by proposing that such a belief holds important secular consequences for some believers. These consequences are symbolic sanctions and are perceived as relevant to proximate marital and family behaviors. The desire to maintain traditional marital and family relations is tied to a belief that their preservation derives from circumspect behaviors which are deemed socially virtuous and legitimate. A prohibitive stance toward premarital sexuality preserves and legitimates marital and family relations. In addition, belief in the strictness of divorce laws is a belief directed to the preservation of these relations. To loosen divorce laws is to loosen the sanctions which comprise the social interdict. The finding that these attitudinal measures enter the analysis when church attendance is removed clarifies, in part, the effect of attendance. While no interaction effect is evident, this analysis suggests that what influence church attendance has on belief in an afterlife is tied to its conserving effect on prohibitive positions on premarital sexuality, abortion and divorce. Finally, the interpretation of the relationship between church attendance and attitude toward premarital sexuality is potentially enhanced by a consideration of the signifi- cance found for race in this analysis. Belief in an afterlife is more significantly related to white than to black respondents. When religiosity and premarital sexuality are consid- ered in light of the role of white and black churches, the effects of church attendance, race and premarital sexuality become sociologically understood. One function of the white Protestant denominations or Catholic churches has been, at least historically, to maintain morality and to control sexual comportment. The function of the black church has historically differed from this. In addition to its closer involvement with economic circumstances of local communities (see Light, 1972:127-151yf W K H E O D F N F K X U F K K D s been a base of and sometimes catalyst for political action (Nelsen et al., 1975yf : K H U e the sect may exact symbolic sanctions similar to those exercised by white churches (Light, 1972:134yf E O D F N F K X U F K D W W H Q G D Q F H F D Q U H L Q I R U F H D F W L R Q G L U H F W H G W R L Q Q H U - worldly, rather than other-worldly matters. NOTES 'In contrast to Hertel, this analysis utilizes occupational prestige scores, believing it to be a more adequate measure of socioeconomic level. The measurement of religious affiliation and region conform to the dummy variable analysis employed by Hertel, where "no" affiliation is left out of the analysis, and "West" is omitted also. This content downloaded from 69.41.96.244 on Thu, 02 May 2019 08:06:45 UTC All use subject to https://about.jstor.org/terms 169 Interpretation of standardized discriminant coefficients permits a comparison of the relative discriminating power of each term in the discriminant function. Thus, a value of .6 indicates twice the discriminating power as a value of .3 'At the zero order level, there is a weak significant relation between occupational prestige and the attitudinal variables, where higher prestige is associated with a more liberal stance. However, in the multiple regression analysis, the exclusion of the occupation variable does not permit the inclusion of the attitudinal variables. This suggests that occupational prestige is a partial determi- nant of one's acceptance of these secular trends, and will therefore have a differential effect on one's perception of life after death. For higher levels the perception is more personal, for lower levels it is implicated in prohibitive stances toward the social trends which are seen as weakening traditional arrangements and behaviors. REFERENCES Bellah, R. N. 1975 The Broken Covenant: American Civil Religion in Time of Trial. New York. Seabury Press. 1976 "The new consciousness and the crisis in modernity." Pp. 333-52 in C. Y. Glock and R. N. Bellah (Eds.yf 7 K H 1 H Z 5 H O L J L R X V & R Q V F L R X V Q H V V % H U N H O H \ 8 Q L Y H U V L W \ R I & D O L I R U Q L a Press. Durkheim, Emile 1964 The Division of Labor in Society. New York: The Free Press. Glenn, N. D., and E. Gotard 1977 "The religion of blacks in the United States; some recent trends and current characteris- tics." American Journal of Sociology 83 (Septemberyf . Goode, W. J. 1951 Religion among the Primitives. Glencoe: Free Press. Herberg, W. 1955 Protestant-Catholic-Jew. New York: Doubleday. Hertel, B. R. 1980 "Inconsistency of beliefs in the existence of heaven and afterlife." Review of Religious Research 21 (Springyf . Hoge, D., and D. A. Roozen (Eds.yf 1979 Understanding Church Growth and Decline 1950-1978. New York: The Pilgrim Press. Huberty, C. 1975 Discriminant Analysis. Review of Educational Research 45 (Fallyf . Laws, J. L. and P. Schwartz 1977 Sexual Scripts. Hinsdale, Ill.: The Dryden Press. Lenski, G. 1963 The Religious Factor. Garden City: Anchor. Light, I. H. 1972 Ethnic Enterprise in America. Berkeley: University of California Press. Mahoney, E. R. 1978 "Gender and social class differences in changes in attitudes toward premarital coitus." Sociology and Social Research 62 (Januaryyf . National Opinion Research Center (NORCyf 1978 General Social Survey. Chicago: University of Chicago. Nelsen, H. M., T. W. Madron, and R. L. Yokley 1975 "Black religion's Promethean motif: orthodoxy and militancy." American Journal of Sociology 81 (Mayyf . Oppenheimer, V. K. 1977 "The sociology of women's economic role in the family." American Sociological Review 42 (Juneyf . Ross, E. A. 1915 Social Control. New York: Macmillan. Skinner, B. F 1953 Science and Human Behavior. New York: Macmillan. Weber, M. 1958 The Protestant Ethic and the Spirit of Capitalism. New York: Scribner's. 1964 The Sociology of Religion. Boston: Beacon Press. Wuthnow, R. 1976 "Recent patterns of secularization: a problem of generations?" American Sociological Review 41 (Octoberyf . This content downloaded from 69.41.96.244 on Thu, 02 May 2019 08:06:45 UTC All use subject to https://about.jstor.org/terms