historical generalizationany generalization about the past, which may or may not be true, e.g. ‘the Danes plundered before they settled’. Such generalizations may be tested against the evidence, and may apply universally to the past, or to an aspect of the past. This is an important sense in which historiography and sociology may claim to be 'S cientific’. It should also be noted that, contrary to some suggestions (see EMPATHIC UNDERSTANDING), such testable generalizations may refer to ‘meaningful’ as well as non-meaningful relations.
Historical generalizations have sometimes been seen as very different from scientific generalizations in that they do not assert fully-fledged universal conditional statements (see COVERING-LAW MODEL) which are intended to apply within their respective frames of reference (e.g. ‘that all mammals suckle their young’) utterly universally, i.e. to have an application to present and future events as well as to past events. Thus, scientific generalizations have been contrasted with merely ‘accidental generalizations ’ (e.g. a statement that all the stones in a particular box are black).
However many historical generalizations are not merely ‘accidental generalizations’ in the above sense, since they may refer to significant regularities in the past which have explanations and implications and may also have a universal basis (e.g. the general tendencies associated with PATRIMONIALISM or BUREAUCRACY). That events in history do sometimes have important ‘accidental’ causes (e.g. ‘Cleopatra's nose’) is not in dispute. Nor are historical generalizations usually easily reached, the reasons being:
- the complexity of the variables involved;
- the ability to manipulate the variables is limited to ‘quasi-experimentation’ (see COMPARATIVE METHOD);
- the limitations of available historical data, and uncertainties and disputes in their interpretation.
However, this need not imply that significant historical generalizations cannot be made or that generalizations with a more universal basis are confined only to the physical sciences. In one respect, historical generalizations may be more easily reached than universal generalizations in the sciences in that, by definition, the past is completed and thus not subject further to the vagaries of human choice and future cases.