Some dychotomous classifications of aggression according to its function

Aggression is a multifaceted phenomenon, with different concepts under the same term. A lack of agreement in the attempts to categorize its different types explains the overabundance of classifications found in the literature, with a consequent poor conceptual clarity. Some authors use the form or mode of the aggressive act as a criterion of classification, and others prefer to focus aggression on its function, purpose, or goal. This manuscript reviews a variety of functional classifications of aggressive behavior, analyzing similarities and differences among them, in an attempt to clarify the nature of diverse types of aggression. A literature review compares various dichotomies of aggression, presenting comparisons between instrumental and hostile aggression, between proactive and reactive aggression, and between premeditative and impulsive aggression. An analysis of empirical data from some self-report instruments applied in our and in other laboratories shows high correlations between hostile, reactive and impulsive aggression, on one side, and between instrumental, proactive and premeditate aggression, on the other. This supports the convenience of unifying some measurement instruments in order to offer a more conceptual clarity and parsimony in the categorization of aggression. Introduction Aggression is a complex phenomenon operating at multiple levels, with a large variety of meanings, and expressed behaviourally in a myriad of ways. Attempts to understand it at a ‘generic level’, therefore, are inadequate. It is essential to make finer discriminations between specific kinds of aggression, defining which category is being analyzed, because each has distinctive determinants and regulatory mechanisms, different functions and antecedents, and separate genetic and neural control mechanisms, being instigated by different external circumstances (Bandura 1986, 1991). Each category describes diverse phenomenological manifestations. The general purpose of this article consists in improving our knowledge of aggression research and the diagnosis, prevention and treatment of the abnormalities of its various types. More specifically, the aim of this review is to point out some limitations in the construct validity of some measurement instruments, and to facilitate its resolution through a more practical, effective and inclusive, yet parsimonious, classification system