I

ideal type. It might be said that to talk about an ideal type is like talking about wet water, for any type, being an abstraction, is ideal and not real in the sense that a given material object is real: there exists this horse and that, but not a horse in general. The difference between an ideal type and a type pure and simple lies not in the abstractness of connotation but in the definiteness of denotation: whereas the types established by biological systematics have referents which fall under them and nowhere else, this is not the case with ideal types. No horse in general ever lived but there are many horses which satisfy perfectly the specifications of ‘horsiness’, whilst nothing like a perfectly rational organization has ever been observed. The idea behind the concept of ideal type is that social phenomena, in virtue of their manifold and fluid nature, can be analysed solely in terms of the extreme forms of their characteristics, which can never be observed in their purity. Pareto pointed out that all concepts of physical sciences are idealizations: that no movement without resistance of the medium has ever been observed (but only surmised in case of celestial bodies), that nothing perfectly straight has ever been found, that vectorial analysis assumes movements which never take place, and that social sciences must proceed likewise. As far as social sciences are concerned, the most useful idealizations can be found in the most mature of them, which is not surprising: the concepts of economic theory, such as perfect competition or static equilibrium, provide the best examples of ideal types.