Definitions favorable and unfavorable to crime are the conceptual core of Edwin Sutherland’s differential association theory, the specific cognitive and normative content that Sutherland argued is transmitted through associational learning and that ultimately determines whether an individual violates the law. Within the Criminology Theories silo, this article isolates the definitional concept from the broader propositional system treated elsewhere, asking what exactly a definition is in Sutherland’s technical sense, where favorable and unfavorable definitions originate, how they combine into the ratio specified in proposition six, and how later researchers have measured and refined the concept. Definitions, in this account, are not isolated beliefs about particular criminal acts but broader orientations toward the legitimacy and bindingness of law, oriented positively or negatively, that accumulate through an individual’s associational history and ultimately tip the balance toward or away from law violation.
Introduction
Sutherland introduced the concept of definitions in proposition five of his nine-proposition statement, specifying that the specific direction of motives and drives is learned from definitions of the legal codes as favorable or unfavorable, and that individuals in modern society are typically exposed to both kinds of definitions simultaneously rather than to a single unambiguous normative environment (Sutherland, 1947). This simultaneity is central to understanding why the concept of definitions does the theoretical work it does: if everyone were exposed exclusively to either favorable or unfavorable definitions, criminality would be a simple function of environment, but because most individuals encounter a mixture, the theory requires a mechanism, the ratio specified in proposition six, for explaining how mixed exposure resolves into behavior.
This article proceeds by examining the concept of a definition as Sutherland used it, distinguishing it from the closely related but distinct concepts of technique and rationalization treated in the companion article on The Nine Propositions of Differential Association, before turning to the sources and content of favorable and unfavorable definitions, the ratio that combines them, the empirical research that has attempted to measure them, and the definitional concept’s continued life within contemporary social learning theory.
The Concept of Definitions in Differential Association Theory
Definitions as Cognitive and Normative Orientations
A definition, in Sutherland’s technical usage, is not a single discrete belief about the wrongness of a specific act but a broader cognitive and normative orientation toward the legitimacy of legal codes in general, an orientation that shapes how an individual interprets and responds to opportunities for law violation across a range of situations (Sutherland, 1947). This breadth distinguishes Sutherland’s concept from a simple attitude scale measuring approval or disapproval of particular crimes, since a definition in the fuller sense encompasses beliefs about whether law generally deserves compliance, whether particular categories of law are legitimate or illegitimate, and whether violation under specified circumstances is excusable, justified, or simply irrelevant to one’s self-concept as a law-abiding or law-violating person.
Matsueda’s influential reconstruction of the theory clarified that definitions operate simultaneously at a general level, shaping overall orientation toward legal authority, and at a more specific level, shaping evaluation of particular acts or categories of acts, a distinction that later proved central to Akers’s reformulation of the concept within social learning theory (Matsueda, 1988). This general-specific distinction helps explain empirical findings in which individuals hold generally pro-social attitudes toward law while simultaneously endorsing specific definitions favorable to particular offenses, such as underage drinking or minor tax evasion, that do not generalize to a broader anti-law orientation, a pattern more consistent with Sutherland’s fuller conception of definitions than with a single undifferentiated measure of criminality.
Distinguishing Definitions from Techniques and Rationalizations
Sutherland’s proposition four specified that criminal learning includes techniques, which are the practical skills required to commit an offense, alongside the motives, drives, rationalizations, and attitudes captured by the concept of definitions, and it is important analytically to keep these components distinct even though they are typically transmitted together in the same associational relationships (Sutherland, 1947). Technique concerns the how of criminal conduct, while definitions concern the whether and why, addressing the normative permission an individual requires before deploying a technique for criminal rather than legitimate purposes; an individual might acquire considerable technical skill relevant to fraud, for instance, through entirely legitimate accounting training, without that skill translating into offending absent an accompanying definition favorable to its criminal application.
Rationalizations, examined in greater empirical depth through Sykes and Matza’s techniques of neutralization, represent a specific subtype of definition activated situationally, at or near the moment of offending, to resolve the tension between an individual’s generally conventional self-concept and a specific criminal act, rather than a generalized orientation held continuously (Sykes & Matza, 1957). This distinction between generalized definitions and situationally activated neutralizations has proven analytically useful for explaining why individuals who hold broadly conventional attitudes toward law can nonetheless offend under specific circumstances, since the neutralization operates as a temporary, act-specific definition rather than as evidence of a generally criminal orientation.
Sources and Content of Favorable Definitions
Verbalizations That Neutralize or Justify Law Violation
Favorable definitions typically take the form of verbalizations, whether expressed aloud within a group or internalized as private justification, that redefine a prohibited act as excusable, necessary, or not genuinely wrong under the relevant circumstances, a process documented in detail by Cressey’s embezzlement research showing that trusted employees required access to a specific vocabulary of motive before committing an offense that violated their own generally conventional self-concept (Cressey, 1953). Nicole Piquero and colleagues’ later survey research on corporate managers found that endorsement of neutralizing beliefs about regulatory violation, including the view that such violations cause no real harm, predicted self-reported willingness to engage in corporate offending even after controlling for perceived detection risk, extending Cressey’s occupational findings to a broader contemporary managerial population (Piquero, Tibbetts, & Blankenship, 2005). Carl Klockars’s earlier life-history study of a professional fence similarly documented an elaborate, internally consistent vocabulary of motive that allowed a career property offender to maintain a self-image as a fundamentally reasonable businessman despite decades of knowingly trafficking in stolen goods (Klockars, 1974). Sykes and Matza’s subsequent typology of neutralization techniques, including denial of responsibility, denial of injury, denial of the victim, condemnation of the condemners, and appeal to higher loyalties, catalogued the specific verbal formulas through which favorable definitions circulate within delinquent subcultures, providing offenders with ready-made justifications that reduce the psychological cost of law violation (Sykes & Matza, 1957).
These verbalizations are learned in the same associational contexts that transmit technique, and their availability within a given peer group or occupational culture substantially shapes whether members of that group experience law violation as morally costly or as routine and justified; organizational research on white-collar and occupational crime has documented how industry-specific vocabularies of motive, often transmitted informally among colleagues, normalize regulatory violations that would be defined as clearly wrongful in a different professional context (Cressey, 1953). The specific content of favorable definitions therefore varies considerably by offense type and social setting, ranging from broad rejections of legal authority in some criminal subcultures to narrow, offense-specific justifications confined to a single category of violation in otherwise conventional populations.
Specific vs. General Definitions Favorable to Crime
Akers’s later elaboration of the definitional concept within social learning theory distinguished general definitions, broad orientations favorable or unfavorable to law violation overall, from specific definitions, orientations attached to particular offenses or categories of offense, a distinction with substantial empirical support (Akers, 1998). Survey research applying this distinction has found that specific definitions, such as approval of marijuana use or minor property offenses, often predict the corresponding specific behavior more strongly than measures of generalized anti-law orientation, suggesting that individuals frequently compartmentalize their normative orientation toward law rather than holding a single undifferentiated attitude applicable across all offense categories.
This specific-general distinction has practical implications for prevention research and program design, since interventions targeting generalized respect for law and legal authority may prove less effective at reducing a particular behavior, such as underage drinking or academic dishonesty, than interventions targeting the specific definitions favorable to that particular behavior within the relevant peer or organizational context. Ronald Akers and colleagues’ early test of the general theory using adolescent substance-use data found that specific definitions favorable to smoking and drinking predicted the corresponding specific behavior more strongly than a generalized measure of respect for law, an early empirical demonstration of the specific-general distinction that later became a standard feature of social learning research (Akers, Krohn, Lanza-Kaduce, & Radosevich, 1979). Marvin Krohn and colleagues extended this substance-specific approach to adolescent cigarette smoking, finding that definitions favorable to smoking, transmitted through differential association with smoking peers, predicted smoking onset and continuation independent of more general delinquent orientation (Krohn, Skinner, Massey, & Akers, 1985). The distinction also helps reconcile differential association theory with everyday observation that most offenders, including many involved in serious criminal careers, hold broadly conventional attitudes toward most legal norms while maintaining specific definitions favorable to violation within a narrower domain relevant to their particular pattern of offending.
Sources and Content of Unfavorable Definitions
Conventional Socialization and Legal Legitimacy
Unfavorable definitions, the counterweight in Sutherland’s ratio, are transmitted through the same associational mechanisms as favorable definitions but originate in relationships and institutional contexts, including family, school, religious organizations, and conventional peer groups, that affirm the legitimacy of legal codes and model compliance rather than violation (Sutherland, 1947). Sutherland did not treat unfavorable definitions as a mere residual category defined negatively against favorable ones, but as a substantive normative content in their own right, encompassing beliefs about the fairness and social necessity of law, respect for legal authority as an institution, and internalized conventional self-concepts that make law violation psychologically costly.
Research on conventional socialization has generally found that strong, consistent transmission of unfavorable definitions within the family during early childhood tends to establish a normative baseline that later peer influence must substantially outweigh before delinquent behavior emerges, a pattern consistent with Sutherland’s priority concept treated in the companion article on frequency, duration, priority, and intensity elsewhere in this silo. This baseline effect helps explain why children raised in households with unusually strong, consistent transmission of unfavorable definitions demonstrate greater resistance to delinquent peer influence during adolescence than children whose earlier associational history has been more mixed or ambiguous. Charles Tittle and colleagues’ large-scale multivariate test of differential association theory across several American communities found that measures of definitions favorable to law violation performed among the strongest predictors of self-reported crime relative to competing theoretical variables drawn from strain and control perspectives, lending broad empirical support to Sutherland’s definitional emphasis even within a study designed to test multiple rival theories simultaneously (Tittle, Burke, & Jackson, 1986).
Definitions Unfavorable Within Otherwise Criminal Environments
Sutherland’s framework also accounts for the presence of unfavorable definitions even within environments generally characterized by high rates of criminal association, since few social environments transmit exclusively favorable definitions without any countervailing conventional influence; even individuals embedded in criminal subcultures typically encounter unfavorable definitions through some combination of family members, media exposure, formal education, or personal experience with the costs of offending, including arrest and incarceration (Sutherland & Cressey, 1978). This partial presence of unfavorable definitions within otherwise high-risk environments helps explain the well-documented phenomenon of resilience, in which individuals exposed to substantial criminal association nonetheless refrain from serious offending, a pattern the theory attributes to a favorable ratio maintained through sufficiently intense or prioritized unfavorable definitions from other sources.
The coexistence of favorable and unfavorable definitions within the same individual’s associational history also explains intermittent or situational offending more readily than models assuming a fixed, unitary criminal orientation, since an individual whose ratio of definitions sits close to the tipping point may offend under some circumstances, when situational factors temporarily activate favorable definitions or neutralizations, while refraining under others, when unfavorable definitions remain more salient. This situational variability, examined further in relation to specific offense contexts in the dedicated treatment of the theory’s applications elsewhere in this silo, illustrates why differential association theory treats the ratio of definitions as a probabilistic rather than deterministic predictor of behavior.
The Ratio and Balance of Definitions
Operationalizing the Excess of Favorable over Unfavorable Definitions
Proposition six’s claim that criminality results from an excess of favorable over unfavorable definitions has proven, as discussed in the companion treatment of the nine propositions, among the most difficult elements of the theory to operationalize precisely, since Sutherland never specified a formula for combining the number, frequency, duration, priority, and intensity of favorable and unfavorable definitions into a single measurable ratio (Matsueda, 1988). Early empirical researchers, including James Short, approximated the ratio through relatively simple measures of the proportion of an individual’s close friends who approved of or engaged in delinquent behavior, a simplification that yielded results broadly consistent with the theory’s prediction while sacrificing much of the conceptual richness of the original formulation (Short, 1957).
Later researchers attempted more elaborate operationalizations incorporating both the favorable and unfavorable sides of the ratio explicitly, rather than measuring delinquent peer association alone, on the theoretical grounds that a full test of proposition six requires attention to the countervailing unfavorable definitions an individual also holds, not merely the presence of favorable ones (Matsueda & Heimer, 1987). Ross Matsueda and Karen Heimer’s structural equation modeling of definitions as a mediating variable between social structural position, differential association, and delinquent behavior represented a methodologically significant advance, demonstrating that definitions favorable to delinquency substantially mediated the relationship between delinquent peer association and self-reported delinquent behavior, providing some of the clearest empirical support for the theory’s core causal chain (Matsueda & Heimer, 1987).
Mixed and Conflicting Definitions in Modern Social Life
Sutherland’s own commentary recognized that modern, culturally heterogeneous societies expose most individuals to genuinely conflicting definitions rather than to a homogeneous normative environment, a condition he traced to the culture-conflict perspective he had absorbed from Thorsten Sellin and that shaped his broader account of why crime rates vary across culturally diverse urban populations (Sutherland, 1947). This recognition of normative heterogeneity distinguishes differential association theory from cruder cultural-transmission models that assume a single, unified criminal culture is simply passed down intact, instead treating individuals as navigating genuinely competing normative claims transmitted by different associates occupying different positions in their social network.
Mark Warr and Mark Stafford’s research on what peers actually do within delinquent friendship groups extended this insight by showing that peer influence operates partly through direct transmission of favorable definitions and partly through peers’ own delinquent behavior, which peers observe and imitate independent of any explicit verbal transmission of definitions, complicating a purely definitional account of peer influence (Warr & Stafford, 1991). This finding suggests that the ratio of definitions, while central to Sutherland’s theoretical architecture, does not exhaust the full causal pathway through which differential association shapes behavior, since direct behavioral modeling appears to operate alongside, rather than solely through, the transmission of favorable and unfavorable definitions.
Empirical Research on Definitions
Survey Measures of Definitions Favorable to Crime
Contemporary survey research typically measures definitions favorable to crime through Likert-scale items asking respondents to rate their agreement with statements expressing approval of, or neutral tolerance toward, various forms of law violation, a methodology that has generated a substantial body of evidence linking definitions favorable to crime with self-reported delinquent and criminal behavior across diverse samples (Sellers, Cochran, & Winfree, 2003). Christine Sellers and colleagues’ work applying this measurement approach to dating violence found that definitions favorable to relationship aggression predicted individual involvement in dating violence independent of exposure to violent peers alone, illustrating the definitional concept’s capacity to explain variance beyond simple measures of associational exposure (Sellers, Cochran, & Winfree, 2003).
These survey-based measures have also been criticized for potential contamination with the outcome variable, since agreement with statements approving of delinquent behavior may be difficult to distinguish empirically from a general willingness to report engaging in that behavior, a measurement concern that has motivated researchers to develop measures of definitions using hypothetical scenarios and vignettes rather than direct approval statements, in an effort to more cleanly separate normative orientation from behavioral self-report. This methodological refinement has generally preserved the core empirical finding that measured definitions favorable to crime predict subsequent offending, even under more conservative measurement strategies designed to minimize definitional-behavioral contamination.
Definitions as Mediators Between Association and Behavior
A substantial body of research, building on Matsueda and Heimer’s structural equation approach, has examined definitions as a mediating mechanism linking differential peer association to individual delinquent behavior, testing whether the effect of delinquent peer association on behavior operates primarily through the definitions those associations transmit, as Sutherland’s theory predicts, or whether peer association exerts an independent direct effect on behavior not fully captured by measured definitions (Matsueda & Heimer, 1987). This mediation research has generally found partial but not complete mediation, meaning that measured definitions account for a substantial portion, though not all, of the relationship between delinquent peer association and delinquent behavior, a pattern consistent with Warr and Stafford’s finding that behavioral modeling operates alongside definitional transmission rather than being fully subsumed by it.
This partial-mediation finding has shaped subsequent theoretical debate about whether Sutherland’s original definitional emphasis requires supplementation with additional mechanisms, such as direct behavioral modeling or differential reinforcement as specified in Akers’s reformulation, in order to fully account for the empirical relationship between association and behavior, a debate examined further in the companion article comparing differential association theory with social learning theory elsewhere in this silo. The persistence of a measurable, if partial, mediating role for definitions nonetheless provides some of the strongest available empirical support for the specific causal mechanism Sutherland proposed in proposition six.
Definitions in Contemporary Theoretical Extensions
Akers’s General and Specific Definitions
Ronald Akers’s social learning theory retained Sutherland’s definitional concept while integrating it more explicitly with operant reinforcement processes, arguing that definitions themselves function partly as discriminative stimuli that signal whether a given behavior is likely to be rewarded or punished within a particular associational context, rather than as freestanding normative beliefs independent of their behavioral consequences (Akers, 1998). This reformulation gave the definitional concept a more precise behavioral function, specifying that definitions favorable to crime increase the probability of criminal behavior partly by signaling anticipated social approval and partly by providing the verbal justification necessary to reduce anticipatory guilt or anxiety associated with anticipated law violation.
Akers’s distinction between general and specific definitions, discussed above, has become a standard feature of contemporary social learning research, generating a substantial body of work testing whether general or specific definitional measures better predict particular offense categories, with findings generally favoring specific definitions for narrowly defined behaviors and general definitions for broader measures of overall criminal involvement (Akers, 2009). This body of work represents a direct empirical and theoretical extension of Sutherland’s original proposition five, refining rather than displacing his core insight that the content of what is learned through association, not merely the fact of association itself, determines behavioral outcome.
Neutralization Theory and the Refinement of Definitional Content
Sykes and Matza’s techniques of neutralization, treated above as a subtype of favorable definition, have generated their own extensive research tradition testing whether specific neutralization techniques predict specific offense categories, with denial of injury and denial of victim proving particularly predictive of property and cyber-enabled offenses in which the harm to identifiable victims may be more psychologically distant than in violent crime (Skinner & Fream, 1997). This research tradition has effectively subdivided Sutherland’s broad definitional concept into a taxonomy of specific justificatory mechanisms, providing considerably greater analytical precision than Sutherland’s own original formulation while remaining fully consistent with his foundational claim that the content of criminal learning includes not merely technique but motive and justification.
The continued vitality of this research tradition, more than seventy years after Sutherland first specified definitions as the central content of criminal learning, testifies to the definitional concept’s productive ambiguity: broad enough to accommodate an expanding range of empirical refinements, from Akers’s discriminative-stimulus reformulation to Sykes and Matza’s neutralization taxonomy, while retaining sufficient conceptual coherence to organize a genuinely cumulative body of criminological research spanning multiple theoretical generations. Michael Benson’s interview-based research on convicted white-collar offenders documented in close empirical detail how such offenders characteristically denied a guilty mind well after conviction, employing specific vocabularies of motive learned within their professional environments to maintain a conventional self-concept despite a documented legal violation, directly illustrating the definitional concept’s continued application to occupational crime (Benson, 1985). Robert Agnew’s subsequent research examined how neutralization techniques function not merely as post hoc justifications but as genuine facilitators of delinquent behavior when invoked prospectively, providing experimental and survey evidence that neutralizing beliefs measurably increase the likelihood of subsequent rule violation among adolescents who hold them (Agnew, 1994).
Conclusion
Definitions favorable and unfavorable to crime supply the specific normative content that differential association theory identifies as the proximate cause of criminal behavior, a content encompassing generalized orientations toward legal legitimacy, offense-specific approvals and disapprovals, and situationally activated neutralizations that together determine whether a given individual’s associational history tips toward or away from law violation. Sutherland’s insistence that these definitions, rather than mere exposure to criminal associates, constitute the theory’s proximate causal mechanism has generated a substantial and still-active empirical research tradition, from Matsueda and Heimer’s mediation modeling to Akers’s discriminative-stimulus reformulation, each seeking to specify with greater precision how definitions are acquired, measured, and translated into behavior. Vold, Bernard, and Snipes’s comprehensive review of theoretical criminology situates this definitional research tradition among the discipline’s most empirically productive, noting that few mid-century theoretical constructs have generated as sustained and methodologically varied a body of confirmatory research as Sutherland’s concept of definitions favorable and unfavorable to crime (Vold, Bernard, & Snipes, 2002).
The persistent difficulty of fully operationalizing the ratio of favorable to unfavorable definitions, combined with mounting evidence that behavioral modeling operates alongside rather than solely through definitional transmission, has motivated substantial theoretical refinement without displacing Sutherland’s core claim that what is learned matters as much as the fact of learning itself. This continued analytical productivity secures the definitional concept’s place as one of differential association theory’s most consequential and most thoroughly tested theoretical contributions.
Related Articles
- The Nine Propositions of Differential Association
- Learning Criminal Behavior Through Social Interaction
- Frequency, Duration, Priority, and Intensity
- Differential Association-Reinforcement Theory
- Differential Association Theory vs. Social Learning Theory
References
- Akers, R. L. (1998). Social learning and social structure: A general theory of crime and deviance. Northeastern University Press.
- Akers, R. L. (2009). Social learning and social structure: A general theory of crime and deviance (2nd ed.). Transaction Publishers.
- Akers, R. L., Krohn, M. D., Lanza-Kaduce, L., & Radosevich, M. (1979). Social learning and deviant behavior: A specific test of a general theory. American Sociological Review, 44(4), 636–655.
- Agnew, R. (1994). The techniques of neutralization and violence. Criminology, 32(4), 555–580.
- Benson, M. L. (1985). Denying the guilty mind: Accounting for involvement in a white-collar crime. Criminology, 23(4), 583–607.
- Cressey, D. R. (1953). Other people’s money: A study in the social psychology of embezzlement. Free Press.
- Krohn, M. D., Skinner, W. F., Massey, J. L., & Akers, R. L. (1985). Social learning theory and adolescent cigarette smoking: A longitudinal study. Social Problems, 32(5), 455–473.
- Klockars, C. B. (1974). The professional fence. Free Press.
- Matsueda, R. L. (1988). The current state of differential association theory. Crime & Delinquency, 34(3), 277–306.
- Matsueda, R. L., & Heimer, K. (1987). Race, family structure, and delinquency: A test of differential association and social control theories. American Sociological Review, 52(6), 826–840.
- Piquero, N. L., Tibbetts, S. G., & Blankenship, M. B. (2005). Examining the role of differential association and techniques of neutralization in explaining corporate crime. Deviant Behavior, 26(2), 159–188.
- Sellers, C. S., Cochran, J. K., & Winfree, L. T. (2003). Social learning theory and courtship violence: An empirical test. In R. L. Akers & G. F. Jensen (Eds.), Social learning theory and the explanation of crime (pp. 109–128). Transaction Publishers.
- Short, J. F. (1957). Differential association and delinquency. Social Problems, 4(3), 233–239.
- Skinner, W. F., & Fream, A. M. (1997). A social learning theory analysis of computer crime among college students. Journal of Research in Crime and Delinquency, 34(4), 495–518.
- Sutherland, E. H. (1947). Principles of criminology (4th ed.). J. B. Lippincott.
- Sutherland, E. H., & Cressey, D. R. (1978). Criminology (10th ed.). J. B. Lippincott.
- Sykes, G. M., & Matza, D. (1957). Techniques of neutralization: A theory of delinquency. American Sociological Review, 22(6), 664–670.
- Tittle, C. R., Burke, M. J., & Jackson, E. F. (1986). Modeling Sutherland’s theory of differential association: Toward an empirical clarification. Social Forces, 65(2), 405–432.
- Vold, G. B., Bernard, T. J., & Snipes, J. B. (2002). Theoretical criminology (5th ed.). Oxford University Press.
- Warr, M., & Stafford, M. (1991). The influence of delinquent peers: What they think or what they do? Criminology, 29(4), 851–866.