[In the following excerpt, Collingwood analyzes some of the mental predispositions of Greek historians that influenced their view of history, historical writing, and the role of the historian.]

… I should like to point out how remarkable a thing is [the] creation of scientific history by Herodotus, for he was an ancient Greek, and ancient Greek thought as a whole has a very definite prevailing tendency not only uncongenial to the growth of historical thought but actually based, one might say, on a rigorously anti-historical metaphysics. History is a science of human action: what the historian puts before himself is things that men have done in the past, and these belong to a world of change, a world where things come to be and cease to be. Such things, according to the prevalent Greek metaphysical view, ought not to be knowable, and therefore history ought to be impossible.

For the Greeks, the same difficulty arose with the world of nature since it too was a world of this kind. If everything in the world changes, they asked, what is there in such a world for the mind to grasp? They were quite sure that anything which can be an object of genuine knowledge must be permanent; for it must have some definite character of its own, and therefore cannot contain in itself the seeds of its own destruction. If it is to be knowable it must be determinate; if it is determinate, it must be so completely and exclusively what it is that no internal change and no external force can ever set about making it into something else. Greek thought achieved its first triumph when it discovered in the objects of mathematical knowledge something that satisfied these conditions. A straight bar of iron may be bent into a curve, a flat surface of water may be broken into waves, but the straight line and the plane surface, as the mathematician thinks of them, are eternal objects that cannot change their characteristics.

Following the line of argument thus opened up, Greek thought worked out a distinction between two types of thought, knowledge proper … and what we translate by 'opinion'.… Opinion is the empirical semi-knowledge we have of matters of fact, which are always changing. It is our fleeting acquaintance with the fleeting actualities of the world; it thus only holds good for its own proper duration, for the here and now; and it is immediate, ungrounded in reasons, incapable of demonstration. True knowledge, on the contrary, holds good not only here and now but everywhere and always, and it is based on demonstrative reasoning and thus capable of meeting and overthrowing error by the weapon of dialectical criticism.

Thus, for the Greeks, process could be known only so far as it was perceived, and the knowledge of it could never be demonstrative. An exaggerated statement of this view, as we get it in the Eleatics, would misuse the weapon of dialectic, which is really valid only against error in the sphere of knowledge strictly so called, to prove that change does not exist and that the 'opinions' we have about the changing are really not even opinions but sheer illusions. Plato rejects that doctrine and sees in the world of change something not indeed intelligible but real to the extent of being perceptible, something intermediate between the nullity with which the Eleatics had identified it and the complete reality and intelligibility of the eternal. On such a theory, history ought to be impossible. For history must have these two characteristics: first it must be about what is transitory, and secondly it must be scientific or demonstrative. But on this theory what is transitory cannot be demonstratively known; it cannot be the object of science; it can only be a matter of … perception, whereby human sensibility catches the fleeting moment as it flies. And it is essential to the Greek point of view that this momentary sensuous perception of momentary changing things cannot be a science or the basis of a science.

The ardour with which the Greeks pursued the ideal of an unchanging and eternal object of knowledge might easily mislead us as to their historical interests. It might, if we read them carelessly, make us think them uninterested in history, somewhat as Plato's attack on the poets might make an unintelligent reader fancy that Plato cared little for poetry. In order to interpret such things correctly we must remember that no competent thinker or writer wastes his time attacking a man of straw. An intense polemic against a certain doctrine is an infallible sign that the doctrine in question figures largely in the writer's environment and even has a strong attraction for himself. The Greek pursuit of the eternal was as eager as it was, precisely because the Greeks themselves had an unusually vivid sense of the temporal. They lived in a time when history was moving with extraordinary rapidity, and in a country where earthquake and erosion change the face of the land with a violence hardly to be seen elsewhere. They saw all nature as a spectacle of incessant change, and human life as changing more violently than anything else. Unlike the Chinese, or the medieval civilization of Europe, whose conception of human society was anchored in the hope of retaining the chief features of its structure unchanged, they made it their first aim to face and reconcile themselves to the fact that such permanence is impossible. This recognition of the necessity of change in human affairs gave to the Greeks a peculiar sensitiveness to history.

Knowing that nothing in life can persist unchanged, they came habitually to ask themselves what exactly the changes had been which, they knew, must have come about in order to bring the present into existence. Their historical consciousness was thus not a consciousness of agelong tradition moulding the life of one generation after another into a uniform pattern; it was a consciousness of … catastrophic changes from one state of things to its opposite, from smallness to greatness, from pride to abasement, from happiness to misery. This was how they interpreted the general character of human life in their dramas, and this was how they narrated the particular parts of it in their history. The only thing that a shrewd and critical Greek like Herodotus would say about the divine power that ordains the course of history is that it … rejoices in upsetting and disturbing things. He was only repeating in (i. 32) what every Greek knew: that the power of Zeus is manifested in the thunderbolt, that of Poseidon in the earthquake, that of Apollo in the pestilence, and that of Aphrodite in the passion that destroyed at once the pride of Phaedra and the chastity of Hippolytus.

It is true that these catastrophic changes in the condition of human life, which to the Greeks were the proper theme of history, were unintelligible. There could be … no demonstrative scientific knowledge [of them].

But all the same history had for the Greeks a definite value. Plato himself laid it down [in Meno] that right opinion (which is the sort of pseudo-knowledge that perception gives us of what changes) was no less useful for the conduct of life than scientific knowledge, and the poets maintained their traditional place in Greek life as the teachers of sound principles by showing that in the general pattern of these changes certain antecedents normally led to certain consequents. Notably, an excess in any one direction led to a violent change into its own opposite. Why this was so they could not tell; but they thought it a matter of observation that it was so; that people who became extremely rich or extremely powerful were thereby brought into special danger of being reduced to a condition of extreme poverty or weakness. There is here no theory of causation; the thought does not resemble that of seventeenth-century inductive science with its metaphysical basis in the axiom of cause and effect; the riches of Croesus are not the cause of his downfall, they are merely a symptom, to the intelligent observer, that something is happening in the rhythm of his life which is likely to lead to a downfall. Still less is the downfall a punishment for anything that, in an intelligible moral sense, could be called wrongdoing. When Amasis in Herodotus (iii.43) broke off his alliance with Polycrates, he did it simply on the ground that Polycrates was too prosperous: the pendulum had swung too far one way and was likely to swing as far in the other. Such examples have their value to the person who can make use of them; for he can use his own will to arrest these rhythms in his life before they reach the danger-point, and check the thirst for power and wealth instead of allowing it to drive him to excess. Thus history has a value; its teachings are useful for human life; simply because the rhythm of its changes is likely to repeat itself, similar antecedents leading to similar consequents; the history of notable events is worth remembering in order to serve as a basis for prognostic judgements, not demonstrable but probable, laying down not what will happen but what is likely to happen, indicating the points of danger in rhythms now going on.

This conception of history was the very opposite of deterministic, because the Greeks regarded the course of history as flexible and open to salutary modification by the well-instructed human will. Nothing that happens is inevitable. The person who is about to be involved in a tragedy is actually overwhelmed by it only because he is too blind to see his danger. If he saw it, he could guard against it. Thus the Greeks had a lively and indeed a naïve sense of the power of man to control his own destiny, and thought of this power as limited only by the limitations of his knowledge. The fate that broods over human life is, from this Greek point of view, a destructive power only because man is blind to its workings. Granted that he cannot understand these workings, he can yet have right opinions about them, and in so far as he acquires such opinions he becomes able to put himself in a position where the blows of fate will miss him.

On the other hand, valuable as the teachings of history are, their value is limited by the unintelligibility of its subject-matter; and that is why Aristotle said [in Politics] that poetry is more scientific than history, for history is a mere collection of empirical facts, whereas poetry extracts from such facts a universal judgment. History tells us that Croesus fell and that Polycrates fell; poetry, according to Aristotle's idea of it, makes not these singular judgements but the universal judgement that very rich men, as such, fall. Even this is, in Aristotle's view, only a partially scientific judgement, for no one can see why rich men should fall; the universal cannot be syllogistically demonstrated; but it approaches the status of a true universal because we can use it as the major premiss for a new syllogism applying this generalization to fresh cases. Thus poetry is for Aristotle the distilled essence of the teaching of history. In poetry the lessons of history do not become any more intelligible and they remain undemonstrated and therefore merely probable, but they become more compendious and therefore more useful.

Such was the way in which the Greeks conceived the nature and value of history. They could not, consistently with their general philosophical attitude, regard it as scientific. They had to consider it as, at bottom, not a science but a mere aggregate of perceptions. What, then, was their conception of historical evidence? The answer is that, conformably with this view, they identified historical evidence with the reports of facts given by eyewitnesses of those facts. Evidence consists of eyewitnesses' narratives, and historical method consists of eliciting these.

Quite clearly, it was in this way that Herodotus conceived of evidence and method. This does not mean that he uncritically believed whatever eyewitnesses told him. On the contrary, he is in practice highly critical of their narratives. And here again he is typically Greek. The Greeks as a whole were skilled in the practice of the law courts, and a Greek would find no difficulty in applying to historical testimony the same kind of criticism which he was accustomed to direct upon witnesses in court. The work of Herodotus or Thucydides depends in the main on the testimony of eyewitnesses with whom the historian had personal contact. And his skill as a researcher consisted in the fact that he must have cross-questioned an eyewitness of past events until he had called up in the informant's own mind an historical picture of those events far fuller and more coherent than any he could have volunteered for himself. The result of this process was to create in the informant's mind for the first time a genuine knowledge of the past events which he had perceived.…

This conception of the way in which a Greek historian collected his material makes it a very different thing from the way in which a modern historian may use printed memoirs. Instead of the easy-going belief on the informant's part that his prima facie recollection was adequate to the facts, there could grow up in his mind a chastened and criticized recollection which had stood the fire of such questions as 'Are you quite sure that you remember it just like that? Have you not now contradicted what you were saying yesterday? How do you reconcile your account of that event with the very different account given by so-and-so?' This method of using the testimony of eyewitnesses is undoubtedly the method which underlies the extraordinary solidity and consistency of the narratives which Herodotus and Thucydides finally wrote about fifth-century Greece.

No other method deserving the name scientific was available to the fifth-century historians, but it had three limitations:

First, it inevitably imposed on its users a shortness of historical perspective. The modern historian knows that if only he had the capacity he could become the interpreter of the whole past of mankind; but whatever Greek historians might have thought of Plato's description of the philosopher as the spectator of all time, they would never have ventured to claim Plato's words as a description of themselves. Their method tied them on a tether whose length was the length of living memory: the only source they could criticize was an eyewitness with whom they could converse face to face. It is true that they relate events from a remoter past, but as soon as Greek historical writing tries to go beyond its tether, it becomes a far weaker and more precarious thing. For instance, we must not deceive ourselves into thinking that any scientific value attaches to what Herodotus tells us about the sixth century or to what Thucydides tells us about events before the Pentecontaetia. From our twentieth-century point of view, these early stories in Herodotus and Thucydides are very interesting, but they are mere logography and not scientific. They are traditions which the author who hands them down to us has not been able to raise to the level of history because he has not been able to pass them through the crucible of the only critical method he knew. Nevertheless, this contrast in Herodotus and Thucydides between the unreliability of everything farther back than living memory and the critical precision of what comes within living memory is a mark not of the failure of fifth-century historiography but of its success. The point about Herodotus and Thucydides is not that the remote past is for them still outside the scope of scientific history but that the recent past is within that scope. Scientific history has been invented. Its field is still narrow; but within that field it is secure. Moreover, this narrowness of field did not matter much to the Greeks, because the extreme rapidity with which their own civilization was developing and changing afforded plenty of first-class historical material within the confines set by their method, and for the same reason they could produce first-rate historical work without developing what in fact they never did develop, any lively curiosity concerning the remote past.

Secondly, the Greek historian's method precludes him from choosing his subject. He cannot, like Gibbon, begin by wishing to write a great historical work and go on to ask himself what he shall write about. The only thing he can write about is the events which have happened within living memory to people with whom he can have personal contact. Instead of the historian choosing the subject, the subject chooses the historian; I mean that history is written only because memorable things have happened which call for a chronicler among the contemporaries of the people who have seen them. One might almost say that in ancient Greece there were no historians in the sense in which there were artists and philosophers; there were no people who devoted their lives to the study of history; the historian was only the autobiographer of his generation and autobiography is not a profession.

Thirdly, Greek historical method made it impossible for the various particular histories to be gathered up into one all-embracing history. Nowadays we think of monographs on various subjects as ideally forming parts of a universal history, so that if their subjects are carefully chosen and their scale and treatment carefully controlled they might serve as chapters in a single historical work; and this is the way in which a writer like Grote actually treated Herodotus' account of the Persian War and Thucydides' of the Peloponnesian. But if any given history is the autobiography of a generation, it cannot be rewritten when that generation has passed away because the evidence on which it was based will have perished. The work that a contemporary based on that evidence can thus never be improved upon or criticized, and it can never be absorbed into a large whole, because it is like a work of art, something having the uniqueness and individuality of a statue or a poem.… [The works of Herodotus and Thucydides were] written to rescue glorious deeds from the oblivion of time, precisely because when their generation was dead and gone the work could never be done again. The rewriting of their histories, or their incorporation into the history of a period, would have seemed to them an absurdity. To the Greek historians, therefore, there could never be any such thing as a history of Greece. There could be a history of a fairly extensive complex of events, like the Persian War or the Peloponnesian War; but only on two conditions. First, this complex of events must be complete in itself: it must have a beginning, a middle, and an end, like the plot of an Aristotelian tragedy. Secondly, it must be … like an Aristotelian city-state. As Aristotle thought that no community of civilized men under a single government could exceed in size the number of citizens that could be within earshot of a single herald, the dimensions of the political organism being thus limited by a purely physical fact, so the Greek theory of history implies that no historical narrative could exceed in length the years of a man's lifetime, within which alone the critical methods at its disposal could be applied.