A Study of Criminal Conversations
- Author:
Norman Powell
Classification Clinic, Sing Sing Prison The first quantitative study of the content of adult conversations appears to have been made by H. T. Moore 2 in 1922. Since that time conversation has become, as Gardner Murphy states in his jExperimental Social Psychology, “one of the best explored of group phenomena.” The assumption is made in such studies that what people talk about is to a considerable degree a reflection of their essential interests and thoughts. It is felt that the present investigation is significant in that it gives at least an indication as to what the incarcerated criminal thinks about and throws some light on the criminal mind.
Evidence is presented here concerning four questions: 1. What is the relative popularity of subjects in a criminal population ?
What is the relationship between the subject under discussion and the age, race, occupational level, and time in the institution of the speaker?
How does this institutional group compare with extrainstitutional groups?
4. Is there a tendency for given conversational topics to have larger or smaller discussion groups ?
The technique employed here is the sampling method, analyzing 485 bits of conversations among inmates in Sing Sing Prison. All data were gathered and tabulated by an inmate assistant in our clinic. Because it seemed inadvisable on account of the physical danger involved, for the inmate to record exactly the particular iA paper read at the annual meeting of the Association of Consulting Psychologists at the University of Pennsylvania, Philadelphia, May 6, 1933. Appreciation is expressed to Dr Amos T. Baker, Director, Classification Clinic; Dr Y. C. Branham, Deputy Commissioner of Correction; Mr. Lewis E. Lawes, Warden, and to Dr W. H. Thayer, Commissioner of Correction for encouragement of research at Sing Sing Prison. I am deeply indebted to inmate 85183 for his sincere and energetic assistance in carrying out this study. 2 Moore, H. T., Further Data Concerning Sex Differences, J. Abn. and Soc. Psychol., 1922, 17, 210-214.
snatches of conversations, a schedule for topical classification was drawn up prior to the investigation. The inmate was given practice in the classification and tabulation of imaginary snatches of conversation until it was felt that the possibility of error had been practically eliminated. Accordingly, conversations were classified and tabulated as soon as heard. At the same time the age level, occupational status, race, time in institution of the speaker, and the number in the group were recorded. All records were taken during recreation periods?all of Saturday afternoon, April 8th, 1933 and the entire next day, Sunday. The population was entirely unaware of the progress of the experiment. Conversations were classified under the following fifteen headings. Edited somewhat, a sample is given for each topical division:
Home.?My family lives in New York.
Weather.?This is a fine day.
Institution.?I got a “report” yesterday for smoking.
Reduction of Sentences.?The one-third bill didn’t pass.
Legal Release from Institution.?I go out in September.
Recreation.?Last night’s movie was not so good.
Sex.?Women don’t mean a thing to me.
Cultural.?I like to read Ibsen.
Politics.?Roosevelt ought to be given a free hand.
10. Crime.?I was too careless in my last job when I stuck that guy up. 11. Escapes.?His was a clean getaway. 12. Health.?I’m feeling a lot better now. 13. Economics.?The market’s going down steadily. 14. Religion.?I don’t believe in God. 15. Food?I couldn’t eat the mush to-day. The relative popularity of topics for the population is shown in Table 1. For the entire group, Home, comprising 14 per cent of the total, is the most frequent topic of conversation. Weather is second with 12 per cent and Health and Religion are least frequent with 3 per cent and 1 per cent respectively. If the several topics relating to getting out of Sing Sing in one way or another are added, a total of 22 per cent is obtained. A decided trend is thus observable. The criminal population is most interested in becoming a non-institutional population and in re-establishing contact with the outside world, but these interests are not nearly so powerful and omnipresent as might have been expected. Some of the results of the study of 500 conversations in ColumA STUDY OF CRIMINAL CONVERSATIONS 63
Table 1 Relative Popularity of Topics Subjects All No Per cent Race White Rank Per cent Negro Rank Per cent Mixed Negro and White Rank Per cent Home Weather Institution Reduction of Sentence… Release from Institution. Recreation Sex Cultural Politics Food Crime Economics Escapes Health Religion 66 56 54 51 42 39 36 25 24 23 21 15 15 13 5 14 12 11 10 9 8 7 5 5 5 4 3 3 3 1 1 2 3 4i ^2 6 81 14 7 41 14 10 81 11 14 15 20 14 9 8 7i < 2 5 3! 6 8 3! 4! 5 31 3! 7 4 1 6 2 4 4 10! 14 8 9 15 12 10| 13 9 11 14 9! 12 11 11 2f 1 2 6 4 0 2? 2f 1 m 2 4 1 11 5 3 6 8 7 11 11 11 14! 11 0 15 10! 20 3 9 12 71 ? 2 4! 6! 3 3 3 0 3
Subjects Home Weather Institution Reduction of Sentence. . Release from Institution. Recreation Sex | Cultural Politics Food Crime !.!!!. Economics Escapes Health Religion Age 21 yrs. and below Rank Per cent 1 17 4! 9 7 8 2! 14 2! 14 6 8 4! 9 13! 1 11 2 9 5 8 6 13! 1 10 4 12 2 15 0 22 yrs. to 40 yrs. Rank Per cent 3 11 1 15 2 13 5! 9 12! 3 5! 9 7 8 4 9 8! 5 10 5 15 1 8! 5 14 1 11 3 12! 3 41 yrs. and over Rank Per cent 3! 10 3! 10 1 14 7 9 3! 10 8! 6 13 3 8! 6 3! 10 10! 4 6 9 13 3 10! 5 13 3 15 0 64 THE PSYCHOLOGICAL CLINIC Subjects Occupation P. & B. Rank Per cent S. & S.S. Rank Per cent Labor Rank Per cent Time in Institution Under 1 yr. Rank Per cent 1-1.99 yrs. Rank Per cent 2 yrs. and over Rank Per cent Home Weather Institution Reduction of Sentence Release from Institution. . Recreation Sex Cultural Politics Food Crime Economics Escapes Health Religion 3? 12 12 3* 5 12 n 7 9! 14 15 4 3 8 13! 12 6 10| 13| 14 10^ 15 8! 18 12 10 11 8 1 2 8! 3! 1 2 3! 1 3! 2 8i 5 3! 14 10? 10! 7 14 10 8! 11 15 12 13 4 7 12 9 13! 7 7 11 10 13! 15 20 8 12 11 9 6 3 6 2 6 6 4 4 2 1 7 2! 6 9 2i ^2 11 9 12! 12! 14! 14! 3 12 12 14 7 14 9 3 14 3 3 2 2 1 1 4 7 2 8 11! 8 13! 11! 13! 9 15 15 10 9 7 11 5 4 5 3 4 3 5 2
bus, Ohio, which Landis and Burtt3 published in 1924 are comparable to ours. Business and money are about sixteen times as popular a topic in their civilian group as it is in our criminal group; sports and amusements are twice as popular, while weather is only one-fourth as popular. About the same proportion of persons in both groups talk about health. Differences with regard to occupational level are even more pronounced than for the group as a whole. Business people and students outside discuss financial matters about ten times as frequently as the business and professional group in Sing Sing. The outside group discusses sports and amusements nineteen times as often, health about eight times as often, and the weather just as often.
Similarly, in the industrial group, finance is much more popular as a conversational topic among extra-institutional workers?43 per cent as against 1 per cent in our group of skilled and semiskilled workers, sports and amusements are about the same, and the weather is nine times as popular inside as outside?18 per cent as against 2 per cent.
If, then, it is assumed that the extra-institutional conversations of inmates in these two groups would have approximated those 3 Landis, M. H., & Burtt, H. E., A Study of Conversation, J. Corny. Psychol1924, 4, 81-90.
Table 2 Comparison between Landis & Burtt and Sing Sing Groups Group as a Whole Subject Landis & Burtt Sing Sing Prison (Man to man) Business and Money 49% 3% Sports and Amusements 15% 8% Health 2% 3% Weather 3% 12% No. of cases 195 485 Subject Divided According to Occupation Landis & Burtt (Business and Students) Sing Sing (Business and Prof.) Landis & Burtt (Industries) Sing Sing (S. & S. S.) Business and Money Sports and Amusements. Health Weather 79% 38% 4% 5% 8% 2% 0.6% 5% 43% 9% 2% 2% 1% 11% 3% 18% No. of cases. 94 164 47 154
found by Landis and Burtt, it may be said that, measured by conversations, the business and professional group is, on the whole, more affected by incarceration than the industrial group, indicating that the more intelligent the individual the greater is the influence exercised by institutionalization.
Unfortunately, non-institutional results are not available with which to compare our group with respect to the influences of race and age. However, an internal analysis of our group with reference to race reveals that home, most popular in the white group, drops to seventh place in the Negro group and drops further to a tie with health for the last place with 0 per cent in the mixed group. A general trend is observable for the incidence of personal topics to be reduced when Negroes and whites converse. Moreover, there is little relationship among the ranks of the subjects discussed by whites, the Negroes and the mixed white and Negro group. The racial differences are rather marked. The correlation coefficient between the ranks for the whites and the Negroes is .35; for the whites and mixed .18, and for the Negroes and mixed .53. It is seen that the order for the Negroes resembles the order for the mixed group more closely than the order for the whites resembles the mixed. In other words, the addition of whites to the Negro group affects the conversation less than the addition of Negroes to a white group. Employing a similar technique of analysis with our group divided into three age levels, we find that the youngest group resembles the oldest group more than the middle group resembles the oldest. With regard to occupation, it is found that the industrial group shows a striking resemblance to the laboring group, and only a slight resemblance to the business and professional group. If occupational level be taken as a rough measure of intellectual level, it may be said that the conversations of those of average intelligence resemble the stupid more closely than they resemble the conversations of the bright.
With respect to analysis by length of time in institution, it is found that whereas home leads in conversations of persons less than one year in the institution, it drops to ninth place with those who have been in Sing Sing from one to two years and rises again to sixth place with those who have been in Sing Sing for more than three years, giving a reversed J-shaped curve. There is a
Table 3 Relationship Between Subject under Discussion, and Age, Eace, Occupational Level, and Time in Institution of the Speaker i Race Negroes Mixed Whites 35 .18 Negroes .53 Age 22 to 40 41 and above 21 and below 39 .60 22 to 40 .49 Time in Institution 1 to 1.99 yrs. 2 yrs. and over Below 1 year 51 .61 1 to 1.99 yrs .61 Occupation Skilled and Semiskilled Labor Business and Professional .26 .18 Skilled and Semiskilled .74
1 The probable errors for the several correlation coefficients are relatively large and quite meaningless. According to H. E. Garrett in his Statistics in Psychology and Education (pp. 142-145) the probable error measures only errors arising from a defective sampling of a population. The probable errors in this case, then, indicate exactly nothing. The population here is the number of topical divisions of the conversations and, presumably, constitutes a complete sampling.
slight unreliable trend with increasing stay in Sing Sing for a decrease in the number of conversations relating to release from the institution. Three groups divided according to increasing length of time in the institution have 24, 23, and 21 per cent of their respective conversations relate to getting out of Sing Sing. All inmates without regard to length of stay in Sing Sing seem to be about equally anxious to obtain their liberty. The median size of each of the discussion groups varies between two and three. There appears to be some tendency for topics relating to institutionalization to be discussed by the larger groups. There are five topics which may properly come under this head (Institution, Reduction of Sentence, Crime, Escapes, and Legal Release from the Institution) and only one of these, that dealing with legal release, has a median size of two. This seems to be the only trend observable. The dispersions for the various discussion groups appear, on the other hand, to be more irregular than the medians. The quartile deviations range from .05 in the case of Home to 1.00 in the case of Economics. Thus, while the median size of each of these two groups is two persons, it is seen that the distribution for Home is much narrower than that for Economics and that there is a distinctly greater spread in the distribution of the discussion groups for Economics. The data for each topic are itemized in Table 4.
In conclusion, two points should be stressed. (1) The differences between civilian and criminal conversations are very great. Table 4 Size of Discussion Group and Conversational Topic Subjects Median1 Quartile Deviation Home 2 .05 Weather 2 .07 Institution 3 .53 Reduction of Sentence 3 .52 Release from Sing Sing…. 2 .22 Recreation 3 .44 Sex 2 .32 Cultural 2 .13 Politics 3 .50 Food 3 ‘ .45 Crime 3 .49 Economics 2 1.00 Escapes 3 .66 Health 2 .28 Religion 2 .38 1 Refers to median number of individuals engaging in particular subject of conversation.
These differences are larger than any yet obtained between a criminal and non-criminal group. Though the obtained differences are almost certainly due to the incarceration itself, it is possible that some are true differences. (2) Analysis of conversations of inmates with special reference to time in institution may prove a fruitful method of determining how the penal institution affects the thoughts, the interests, and the personalities of the criminal population.
Disclaimer
The historical material in this project falls into one of three categories for clearances and permissions:
Material currently under copyright, made available with a Creative Commons license chosen by the publisher.
Material that is in the public domain
Material identified by the Welcome Trust as an Orphan Work, made available with a Creative Commons Attribution-NonCommercial 4.0 International License.
While we are in the process of adding metadata to the articles, please check the article at its original source for specific copyrights.