ADVERTISEMENTS:
There is a debate over proper relationship between explanation and prediction. One view regards them basically the same, while other view finds them fundamentally different. But there cannot be a valid prediction without having a proper explanation.
Despite several hurdles, prediction remains the basic goal of social sciences. In a way, capacity to predict events indicates scientific development and growth of a discipline, including a theory. Prediction enables us to apply knowledge involved in a theory or science to solve practical problems of organised political life. It enables politicos to anticipate and avert negative events that might occur in future.
Structurally, predictions are identical with explanations. They have, like explanations, covering laws and initial conditions with the difference that in explanations the conclusion already occurs, and the explanans are sought, but in predictions the explanans are given and the conclusion is sought. Every adequate explanation is potentially a prediction, and every adequate prediction can be treated as a potential explanation. But there are situations when a scholar is able to explain but not predict, or able to predict but not explain.
Prediction about individual events can be made if the covering law generalisations are universal. But such universal covering law generalisations are rarely available in Political Science. Only limited or partial prediction is feasible in Political Science as it has to be based on statistical generalisations and sets of initial conditions. Statistical generalisations do not allow prediction in individual events. They can permit only probability of an individual event, that too in proportion to the degree of frequency.
When relevant information in the explanans is missing, it is not possible to predict the event or explanandum. Sometimes prediction takes place without being able to explain. The statistical prediction of election-results is far better than explanation of those results. Prediction without explanation may take place when only some initial conditions are known but none of the covering laws are known or specified. But such prediction cannot be considered as informed or scientific prediction. An event can be explained and predicted when covering laws and initial conditions are discovered and made known.
Explanation requires knowledge of necessary and sufficient conditions. But Social Sciences have only limited measure of that knowledge. A successful theory stands on its explanatory power.
There are two criteria of evaluating this explanatory power:
(a) Making of scientific prediction, or
ADVERTISEMENTS:
(b) Providing understanding or verstehen.
Prediction can be scientific only when we are able to know all causative factors or have knowledge of all necessary and sufficient conditions making an event. Till we are able to do it, political scientists cannot make scientific predictions. As such, they cannot talk of ‘control’ over human events. Therefore, we cannot accept Hempel’s proposition to evaluate theories only on the basis of their ‘power of making predictions’.
Dilthey, Wildenband, Weber and others put emphasis on the power of man in determining their environment. Therefore, ‘understanding’ or man must be the basis of adequacy of explanation. They put analysis of data as secondary. Man often thinks in terms of prevailing societal structure or ideology. This makes historical perspective much more important. Often they grapple, in their analysis, with the search of ideals and norms of society. Sometimes they indulge in vague historicism.
Thus, both views suffer from severe limitations. The first one fails to reach its height of scientific prediction; second one, makes its ‘understanding’ subjective, changing and captive of past. Whatever be the ‘scientific’ prediction, man begins to falsify it, as in case of Marxian prediction of Future Society. Man himself is an independent variable. Automation, IT, LPG etc. have increased his power beyond all calculation or prediction.
ADVERTISEMENTS:
Analysis and Explanation do not stop at any destination of making ‘scientific prediction’. Explanations are always problematic, always imperfect, and always uncertain. Their quality depends on logical coherence compatibility with other accepted explanations in the field, experimental evidence, of course, use to achieve the purpose. There are various grounds for criticising explanations, but all of them cannot be properly justified.
Explanations are in some respects analogous to maps. Using conventional symbols and transformation rules that link them to the environment, maps produce a record of particular observations that tells us what to expect when certain landmarks are sighted.
The five basic points for testing maps or explanations are:
(i) Internal consistency,
ADVERTISEMENTS:
(ii) The relation of selection of variables to the stated purpose,
(iii) Accuracy of data incorporated in the structure,
(iv) Compatibility with accepted structures used in analogous situations, and,
(v) The test of use in achieving the purpose applied to maps or to suggested explanations to evaluate their quality.
There are three major forms of testing. It is done by probing into the explanations on the basis of their compatibility (a) with existing knowledge, (b) with historical illustrations, and, (c) use or application in laboratory or natural experiments.