- 著者
-
星野 周弘
- 出版者
- THE JAPAN SOCIETY OF EDUCATIONAL SOCIOLOGY
- 雑誌
- 教育社会学研究 (ISSN:03873145)
- 巻号頁・発行日
- vol.30, pp.61-72,en179, 1975
1.Two phncipal approaches to the problem of definitions are usually recognized as the legal and the behavioral or sociological. There seems to be extensive agreement that delinquency consists of behaviors recognized as undesirable or behaviors formally prohibited by law. However, a number of questions always arise as to this seemingly simple way of defining term. One of such questions is concerned with the problem of what is involved in the idea of delinquency. Is it a single act or must there be a series of related acts, a pattern ofbehavior in order to establish the fact of delinquency? There are also questions about whois a delinquent and when one becomes a delinquent. Whether causal analysis of delinquency is possible depends on how these basic questions are answered.<BR>2. Three fundamental perspectives on delinquency dominate the current scene. They are strain or motivational theories, control or bond theories and cultural deviance theories. Although most current theories of delinquency contain at least two and occasionally all three of these perspectives, reconciliation of assumptions is very difficult. Each investigator should begin framing hisperspective in order to analyze causes of delinquency.<BR>3. There are three principal requirements that an empirical investigator must meet in order to be able to say that A causes B:<BR>1) A and B are statistically associated.<BR>2) A is causally prior to B.<BR>3) Theassociation between A and B does not disappear when the effects of other variables causally prior to both A and B are removed.<BR>4. There have been many arguments among proponentsof "general theory" or "multiple factor" approaches. Multiple factor adherents should state more explicitly the reasons for their choice of particular itemsfor analysis and the general theorists should examine and make more extensive use of data.The causes of delinquency must be discussed more in probabilistic terms than in deterministic models.