In his effort to escape from the individualizing andparticularizing approach of German Geisteswissenschaft andhistoricism, Weber developed a key conceptual tool, the notion of theideal type. It will be recalled that Weber argued that noscientific system is ever capable of reproducing all concrete reality,nor can any conceptual apparatus ever do full justice to the infinitediversity of particular phenomena. All science involves selection aswell as abstraction. Yet the social scientist can easily be caught in adilemma when he chooses his conceptual apparatus. When his concepts arevery general--as when he attempts to explain capitalism or Protestantismby subsuming them under the general concepts of economics orreligion--he is likely to leave out what is most distinctive to them. When, on the other hand, he uses the traditional conceptualizations ofthe historian and particularizes the phenomenon under discussion, heallows no room for comparison with related phenomena. The notion ofthe ideal type was meant to provide escape from this dilemma.
An ideal type is an analytical construct that serves theinvestigator as a measuring rod to ascertain similarities as well asdeviations in concrete cases. It provides the basic method forcomparative study. "An ideal type is formed by the one-sidedaccentuation of one or more points of view and by the synthesis of agreat many diffuse, discrete, more or less present and occasionallyabsent concrete individual phenomena, which are arrangedaccording to those one-sidedly emphasized viewpoints into a unified anlayticalconstruct." An ideal type is not meant to refer to moralideals. There can be an ideal type of a brothel or of a chapel. Nor did Weber mean to refer to statistical averages. AverageProtestants in a given region or at a give time may be quite differentfrom ideal typical Protestants. The ideal type involves anaccentuation of typical courses of conduct. Many of Weber's idealtypes refer to collectivities rather than to the social actions ofindividuals, but social relationships within collectivities are alwaysbuilt upon the probability that component actors will engage in expectedsocial actions. An ideal type never corresponds to concretereality but always moves at least one step away from it. It isconstructed out of certain elements of reality and forms a logicallyprecise and coherent whole, which can never be found as such in thatreality. There has never been a full empirical embodiment of theProtestant Ethic, of the "charismatic leader," or of the "exemplaryprophet."
Ideal types enable one to construct hypotheseslinking them with the conditions that brought the phenomenon or eventinto prominence, or with consequences that follow from its emergence. If we wish to study the religious roots of modern capitalism, it may beadvisable to construct an ideal type of Protestant, based on thedistinct features of sectarians as these emerged during the Reformation. We shall then be in a position to determine empirically whether theconcrete conduct of Protestants in, say, seventeenth-century England didin fact approximate the type and in what specific aspects it failed todo so. This type will further allow us to distinguish between theconduct of men who adhered to Catholic or Protestant religious bodies. We can then proceed to correlations and causal imputations as to theconnections between the emergence of Protestantism and that of moderncapitalism--both being conceived in ideal typical terms. As JulienFreund puts it, "Being unreal, the ideal type has the merit of offeringus a conceptual device with which we can measure real development andclarify the most important elements of empirical reality."
Weber'sthree kinds of ideal types are distinguished by their levels ofabstraction. First are the ideal types rooted in historicalparticularities, such as the "western city," "the Protestant Ethic," or"modern capitalism," which refer to phenomena that appear only inspecific historical periods and in particular cultural areas. A secondkind involves abstract elements of social reality--such concepts as"bureaucracy" or "feudalism"--that may be found in a variety ofhistorical and cultural contexts. Finally, there is a third kind of idealtype, which Raymond Aron calls "rationalizing reconstructions of aparticular kind of behavior." According to Weber, all propositions ineconomic theory, for example, fall into this category. They all referto the ways in which men would behave were they actuated by purelyeconomic motives, were they purely economic men.
From Coser, 1977:223-224.