Is History a Science?


The great value of the ‘Is history a science?’ debate is the manner in which it helps clarify the nature of history and to delimit what history can, and cannot do. To the ordinary common-sense mortal, the most striking difference between history and natural science is the degree to which  proof can be established of the various contentions made by the scientist and  the historian respectively. I say ‘degree’, though the more self-regarding historians would probably join with history’s severest critics in saying there is little or no similarity between the scientist’s methods and the historian’s ‘intuition,’ between the scientist’s empirical expertise and the historian’s creative flights. Yet neither ‘intuition’ nor ‘creation’ need represent a fundamental divide between history and science. The gifted scientist will usually develop a ‘feel’ for his subject which may not be greatly different from the intuition of which some historians boast. The scientist of course will attempt empirically to demonstrate the validity of any hunch he may have; his ‘feel’ will take him in the direction of  trying one kind of experiment rather than another, not towards stating untested assumptions. But again this is not terribly different from the way the professional historian (as distinct from the inspired charlatan) sets to work; intuition may suggest certain causal connections but the historian will do his best from the material at his disposal to establish at least the probability of such a casual relationship; better still he may be stimulated to seek for entirely new source materials (rather as a scientist might devise an entirely new type of experiment). On the matter of ‘creativity,’ it is surely not to be contested that Einstein’s theory of relatively is one of the great monuments to human creative thinking. Of course most practising scientists are engaged on much more basic tasks; but then a large number of historians are engaged on pretty mundane work as well.


Rather cunningly, a few sentences ago, I introduced the word ‘probability.’ The historian can only show from his sources that it was likely or at most, probable, that something happened in the way he says it did. But natural science today also deals in probabilities rather than in the certainties of nineteenth century days. Many of those who so vehemently deny that history can have any resemblance to a natural science reveal appalling ignorance of the direction natural sciences are currently taking With the ‘Relatively Revolution,’ the Newtonian absolutes were dethroned. The discovery of ‘quanta’ contradicted the conception of the continuity of the infinitesimal calculus. The theory of mutations pointed to change coming through leaps, not by gradual process. Today, scientists can, from time to time, be heard calling for a revision of scientific laws: in October 1968 Professor Fred Hoyle challenged the Royal Astronomical Society with the need for a radical change of the laws of physics; only thus, he said, would it be possible to account for the ‘funny things which are going on’ in the universe. Let us therefore agree that, save on the most banal level, there are no absolutes in the natural world. So when the historian fails to establish conclusive proofs for his version of past events he may not necessarily be exposing himself as thoroughly unscientific.


However, there is a difference, and we all know there is a difference. The physical scientist cannot call for a repeat performance of the past. The scientist, it may be argued further, can preserve an objectivity towards the phenomena he is studying, whereas the historian can never be completely objective. On the whole this distinction to must be allowed to stand, though again as one of degree rather than as an absolute. After all, as has often been pointed out, the man who assembles the apparatus for a particular experiment effectively becomes a part of the experiment: even in physical science the human, subjective element can never be entirely excluded.


The other central problems in the ‘Is history a science?’ debate are these: the ultimate (often remote) objective in scientific exploration is the formulation of a scientific law, but there are no general laws in history (attempts in this direction having commended themselves neither to historians nor to non-historians); scientific knowledge provides the power of prediction; the historian cannot predict. The latter point is in some ways a bit of a red herring: the historian’s concern, by definition, is with the past; he may well, as a result of his expertise, make some intelligent predictions about the present and future, but that is not strictly his business. E.H. Carr has given an example of the kind of prediction the historian might indulge in:


“People do not expect the historian to predict that revolution will break out in Ruritania next month. The kind of conclusion which they will seek to draw, partly from specific knowledge of Ruritanian affairs and partly from a study of history, is that conditions in Ruritania are such that a revolution is likely to occur in the near future if somebody touches it off, or unless somebody on the government side does something to stop it; and this conclusion might be accompanied by estimates, based partly in the analogy of other revolutions, of the attitude which different sectors of the population may be expect to adopt. The prediction, if such it can be called, can be realised only through the occurrence of unique events, which cannot themselves be predicted; but this does not mean that inferences drawn from history about the future are worthless, or that they do not possess a conditional validity which serves both as a guide to action and a key to our understanding of how things happen.”


An overwhelming majority of historians in fact will probably have no knowledge whatever of Ruritania and little knowledge of revolutions. Non- historians do have a right to expect intelligent and informed comment on current events from their historical friends; statesmen, civil servants, television commentators may reasonably be presumed to react more intelligently to current crises if they have a historical training. But this is really to move away for the kind of ‘prediction’ continuously practised by the professional historian in the normal line of business, a type of ‘prediction’ which in a small way is analogous to the prediction of the physical scientist. This comes about when the historian, using the evidence he has painfully accumulated, together with the feel for the way things happen in certain circumstances which he has developed over the years, makes an inference about something for which he does not in fact have full and sufficient evidence. He is ‘predicting’ what will be seen to have happened once the full evidence is forthcoming. This, in the end, is certainly not the same as the scientist’s prediction (indeed the special word ‘retrodiction’ has been coined to capture the difference) but again, one might argue that it is a difference of degree rather than of kind.

Much time, most fruitfully, has been spent over the issue of general laws. It has now become something of a platitude that ... whenever the historian speaks of a ‘war,’ of a ‘revolution,’ of ‘feudalism,’ of ‘representative government,’ he is using a generalisation. In each of these cases he is suggesting that there are certain general features which characterise a ‘war,’ “feudalism,’ etc. However, the use of such generalisation still leaves us a long way from the formulation of general laws. Given all the qualifications that must now be made about the nature of scientific laws, that they are working hypotheses, expressions of tendencies and probabilities, not iron-hard certainties, it still does seem that in the physical sciences there are laws which differ in scale from any generalisation which the historian might feel competent to make. Again the distinction is not an absolute one; after all, the formulation of scientific laws only exercises the physical scientist operating at the highest level: most scientists are just as immersed in detail as are most historians.


There is one further point which is sometimes thrown into the argument, notably by Trevelyan. This is the one about science having use, while history, of course, is ‘useless.’ ‘What is meant, of course is immediate tangible use: television sets, pasteurised beer, nuclear bombs. The natural scientist working as a scientist (and not as an industrial chemist or scientific adviser to some government or corporation) would however deny that his researches are directed towards such utilitarian products. The scientist seeks knowledge of the phenomena of the physical universe as the historian seeks knowledge of the human past. If the scientist is anything more than crusty misanthropist he will believe that somewhere sometime his discoveries in ‘pure’ science will have practical application; that belief is not fundamentally different from that of historians.


All this would suggest that while there is no fundamental distinction between the main aims and methods of the historian and of the physical scientist, nonetheless there are good reasons for the common-sense assumption the differences do exist. The final point which highlights this sense of difference springs from the manner in which, in one form or another, history becomes implicated in the making of value judgments. Most historians would accept Professor Knowlses’s neat statement: ‘The historian is not judge, still less a hanging judge.’ But they also rejoice at the delicate comeuppance which the late Professor Alfred Cobban administered to Professor Michael Oakeshott’s pleas for complete moral neutrality (in paper read to the second and fourth Irish Conferences of Historians):


“It is admittedly difficult,” says Professor Oakeshott, “to avoid ‘the description of conduct in, generally speaking, moral terms.’ This I take to mean that, for example, we cannot help describing the September massacres as massacres. The important thing is to avoid any suggestion that massacres are a bad thing, because this would be a moral judgment and therefore non-historical.”


The historian cannot help but moral judgments, if only by implication or by virtue of his selection of the facts: these judgments are of a type not encountered in the natural sciences.


Finally, to recall a point already made if the historian’s activities truly are necessary to society, he must communicate the fruits of his labours to that society. There falls upon the historian a duty to write serviceable prose which does not fall on the scientist, whose labours may best be summed up in a few pages of equations.


The most apposite words of all are those of Professor E.E. Evans-Pritchard, the anthropologist: ‘When will people get it into their heads that the conscientious historian is no less systematic, exacting and critical in his research thana chemist or biologist, that it is not in method that social science differs from physical science but in the nature of the phenomena they study.’ Here surely is the crucial point: the historian is concerned with a different kind of material, human experience in the past, from that with which the natural scientist is concerned.

by Arthur Marwick ‘The Nature of History’ Macmillan Press, 1970.