Experiment means controlled observation, that is, observation of the phenomena under artificially created conditions An experiment requires three conditions: the things or phenomena should be of such a nature and dimensions as could be arranged and controlled by the experimenter; secondly, he must be able to exclude extraneous factors and influences from the artificially created conditions, and thirdly, the conditions must be such that they can be repeated at different times and in different places so that the conclusions arrived at may be verified by all subsequent experiments.
Now, these conditions are fulfilled, par excellence, in physical sciences, but not so in social sciences in general and Political Science in particular. The reason is that Political Science deals with such materials as political parties, governments and States, which cannot be confined in a laboratory to be experimented upon. Nevertheless, if by experimental method we mean the trial and error method, or if we mean by it to be guided by past experience and observation, then the experimental method is applicable in Political Science also.
The nature of the society and State is such that they cannot be made an object of artificial experiments. “We cannot do in Politics, writes Lewis, “What the experimenter does in Chemistry. We cannot take a portion of the community in our hands as the King of Brobdignag took Gulliver, view it in different aspects and place it in different positions in order to solve social problems and satisfy our speculative curiosity.”. If a political scientist, for instance, wishes to experiment with democracy, he cannot select a State at will, introduce his democracy in it and wait for determinate results. He will not have them, because, firstly, the people will not permit him to experiment upon them. But even if they do, he will be unable to exclude such influences from his experiment, as famines, commercial crisis, revolts and the like which will affect his conclusions differently on different occasions. All this means that man; society, State and other social and political phenomena are such materials that cannot be experimented upon as we can experiment upon acids, alkalies, electricity or atom
Another difficulty of applying experimental method to politics is that of measurement. Natural phenomena can be quantitatively measured. You can measure the temperature of a hot substance with a thermometer, but you cannot measure the Tieat’ of a heated debate on a bill in a legislative assembly. Opinions and other factors which influence politics are not capable of measurement.
Still another difficulty is that social phenomena do not repeat themselves uniformly as do the natural phenomena. Everything in human society and politics is constantly changing. We cannot repeat our actions again under the same conditions, because they do not reappear again. An ancient Greek Philosopher, Heraclius, said, “you cannot cross a river twice’.
This is the real reason why prophecy or prediction is not possible in politics and other social sciences, and their laws are not exact, uniform and universally valid, as are those of natural sciences. So, even if we make an experiment in Politics and draw conclusions and laws, they may be useless, for we will not be able to apply them on the next occasion. Every event in social sciences is unique and has very little in common with similar events before and after it. Probability and not absolute prediction is possible in political reasoning and laws.
So we conclude with Lowell that “Politics is an observational and not an experimental science”, and that the method of observation is its true method of investigation. “The main laboratory fen: the actual working of political institutions,” he adds, “is not a library but the outside world of political life’. And there the phenomena must be sought and observed at first hand.