Thucydides as a ‘Scientific’ Historian?

[There are still some formatting and character issues to be resolved in the following text]

That the question which is the subject of this discussion was posed at all, primarily reflects the concerns of the modern and only secondarily those of Thucydides. The possibility of a ‘scientific’ history while certainly suggestive in Thucydides is very much a modern construct embodying several elliptically formulated assumptions that need to be brought out and disambiguated.

Any suggestion that the question can be comprehensively answered within the scope of this discussion is to hold a superficial view of the epistemic problems involved. The question assumes that:

a) we already have a characterisation of historical explanation, that

b) we also have specified the nature of scientific explanation and that

c) there is consensus in understanding the relationship (the continuities and discontinuities) between these fields of enquiry.

The problem is further compounded by filtering Thucydides uncritically through these assumptions; we need to retrieve Thucydides from what modern scholars “assume” to be Thucydides (de Ste. Croix 1972: 3). Of course we should not “judge” (Finley 1972: 29) Thucydides solely from a modern perspective but it would be incoherent not to engage in rational reconstruction. The problem lies in commentators failing to “specify” the nature of modern inquiry yet blithely thinking it appropriate to evaluate Thucydides on these terms. Thus unavoidably I highlight the issues of a), b) and c) on which my two pronged analysis depends: an analysis of the modern Thucydidean scholarship coextensive with an analysis of Thucydides. Thucydides’ distinctiveness as a classical historian lies in his rudimentary appreciation of the epistemological dimension of historical inquiry, and it is this aspect that has continuing resonance for the philosophically minded modern historian.

The question of Thucydides’ prima facie “scientific method” divides into two possibilities. Firstly, does Thucydides employ the kinds of explanatory models that scientists now use? For example, the deductive-nomological model (Hempel 1962), stressing the need for reference to universal laws in our explanatory accounts. Perhaps Thucydides intuitively believed that some such laws could be discovered if we consider the following well known extract (I: 22) from his History of the Peloponnesian War: “It will be enough for me, however, if these words of mine are judged useful by those who want to understand the events which happened in the past and which (human nature being what it is) will, at some time or other and in much the same ways, be repeated in the future . . . my work is not a piece of writing designed to meet the taste of an immediate public, but was done to last forever”. This excerpt is by the strongest implication suggestive of an appeal to the criteria of scientific character, i.e. law-like, premised on the immutability of human nature. The second question that presents itself is whether Thucydides avoids bias and “subjectivity”, and tests his evidence, to the extent that we now require of (and expect from) a professional historian, in light of his limited discussion of methodology as laid out in I: 20-22? In I: 21 Thucydides does not claim to offer conclusive evidence, rather ‘better’ or ‘reasonably accurate’ evidence and it is against these claims that we need evaluate the question of bias and subjectivity. We have to consider the deeper claim that historical enquiry as such (regardless of who conducts it – Greek or modern) is infected with subjectivity to such a degree that to talk of historical ‘objectivity’ has no place in the discussion. With this strategy in mind the structure of the discussion is as follows: Section II: The continuities and discontinuities between historical and scientific explanation; Section III: Evaluates Thucydides’ “subjectivity” in light of his own concerns and in light of Section I; Section IV: I offer some concluding remarks.


 I don’t think it is helpful, nor indeed possible to offer an essentialist definition of “scientific”, however we can identify typical features. Using Lindley (1978: 97) as my guide I list the following features of scientific features: deductively drawing conclusions from hypotheses, explaining events or classes of events, confirming theories by observation and experiment, disconfirming, (or disproving) theories by observation and experiment, generalizing from observation and experiment. I think that non-scientists would recognise items two and five as the more typical characteristics of scientific activity. So what does a modern scientific explanation look like? Lindley (1978: 102-9) identifies four requirements for a scientific explanation: deductive connection; truth; empirical content; reference to universal law. Appealing to universal laws we might get a Hempelian explanation of this form:

1. Always (or necessarily) if conditions C1 . . . CN obtain, then an x-type event occurs

2. Conditions C1 . . . CN obtained/ will obtain


3. An x-type event occurred/ will occur.

This is the deductive-nomological model; nomological because it contains a lawlike statement, ‘Always (or necessarily)’ and deductive because 3. follows logically from 1. and 2. We should notice a problem raised by JB Bury. His point is that Hempel-type laws or probabilistic generalisations – of which Bury himself is quite ready to accept exist – are liable to ‘cross’ in non-lawlike and non-probabilistic ways (cf. Hempel 1962 on probabilistic-statistical explanation). To illustrate this point Bury (1930: 62) offers the following example: “The appearance of the microbes which caused the plague . . . causing among other effects the death of Pericles . . . had its own definite chain of causes, but this sequence had nothing to do with the sequence of events which led to the war . . . the microbe had no interest in the war . . . the political effects of the plague were a contingency.” It must, however, be admitted that the deductive-nomological model is an ideal that for the most part not even science can meet. More often in practice, a scientific theory is an inference to the best explanation. The Hempel model offers in a bold and clear manner, with conceptual economy and richness, a model that best accommodates the idea of ‘laws’ and ‘evidence’ and as such it should be viewed as an ideal. To apply the deductive-nomological model to historical explanation we need laws of history. Marx’s laws of history were overall trends or ‘patterns of behaviour’ (de Ste. Croix 1981: 27) rather than Hempelian ‘Always (or necessarily)’ laws. It is only the vulgar Marxist (or Marxist detractor) that makes claims for ‘iron laws of history’, supposedly derived from Thucyides’ holding a fixed view of human nature, something we’ll look at in greater detail in section III. We should note a problem concerning deductive validity as applied to historical inquiry. Goh (1970: 408-9) marks this problem in making the distinction between supporting and covering laws. Essentially, one needs a covering law connecting the conditions with the event, otherwise the deuction has a missing link. With history being inferential activity, there’s no necessary connection between the laws and the conditions; so they can’t yield a conclusion. So even were we able to describe the conditions leading to the outbreak of the Peloponnesian war: the growing strength of Athens (I: 23: 6); the Spartans feeling threatened by this growth; the Megarian decree forbidding trade; the personal qualities of Pericles, and so on, I would still be no nearer to knowing what the relevant supporting or covering laws would look like. Is it a matter of not happening to know the laws which in principle could be found, or are there deeper reasons why no such laws are available in the social sciences?“’

Suppose then, that we relax the nomological element into the modified form:

Probably, if conditions C1 . . . CN, an x-type event occurs. This reformulation replaces strict, exceptionless laws with probabilistic generalisations. It also cancels the deductive aspect of the model. So we are back to history as an inferential inquiry.


There are two particular and inextricably linked epistemological problems (among others) that confront historical explanation: the non-existence of the past and the uniqueness of historical events Past events happened, but the past does not exist. Thucydides does not deny that there was a series of events that comprised what is called the Peloponnesian War, but the events existed in space and time which creates a problem for knowledge and truth. But in the case of the past, the ‘evidence’ is in principle uncheckable against what it is supposed to be evidence of. That’s because what it’s supposed to be evidence of, no longer exists – namely, the past. Thucydides was well aware of this problem. In I: 1 he says, ‘. . . I have found it impossible because of its remoteness in time, to acquire a really precise knowledge of the distant past or even of the history preceding our own period. . . ‘ Thucydides understood that historical perspective was limited by the fallibility of memory and that beyond the confines of living memory, historical accuracy was bound to become increasingly tenuous. If we cannot in principle check the past as evidence of itself, or in other words create a repeatable experiment to verify our hypothesis, then the second problem that then arises is the de facto uniqueness of historical events, for in the absence of testing how else could we prove otherwise? How can the required generalisations be derived to offer the sort of explanatory accounts we set out to find?


I have to be fair in saying that I doubt whether so-called “scientific” historians are making the claims of scientific activity as we’ve discussed; it was important to examine what strictly speaking constitutes scientific activity so that we are not blithly making unsubstantiated claims. There are two dominant views each with their respective strengths and weaknesses that have given rise to the polarised discussion of whether history is sui generis (Berlin 1980: 131), the view most often held by the idealists, or whether history is a branch of science with its own special procedures which enable it to pass beyond common-sense in the search for regularities and uniformities, perhaps best termed as Wissenschaft. I doubt whether there is a single procedure properly called scientific method, something self-consciously applied by all and only scientists. Each science has its own methods. The idealist claim that when a historian invokes ‘scientific’ history all one really means is “accurate” history, has no force (Thucydides’ makes the weaker claim to offer a ‘more’ accurate history I: 21); what competent historian would seek to refute such a self-evident claim. Dover says (1994: 257) that even when he was at school the question of whether science was a history or art was a weary topic.


There are three sources that together have encouraged the view of Thucydides as ‘scientific’ which I will deal with, beginning with the least contentious. The Sophists, who leaving aside Plato’s pejorative view of them, did create an intellectual climate fostering the exercise of man’s rationatory powers through argument and persuasion, is reflected in Thucydides’ attempt at offering competing accounts and claims of the protagonists in his History. De Ste. Croix (1972: 13) is concerned that the uniform style of the speeches with their distinctly sophistic polish would have been beyond the understanding of the [insert Greek]. The second ‘scientific’ influence on Thucydides and most often pointed to as conconclusive, were the existence of his contemporaries of the Hippocratic ‘school’. They offered an empirical approach to events. The detailed account of the disease which plagued Athens reflected Thucydides’ concern to document the situation so that the reader might recognise the symptoms. But as de Ste. Croix (1972: 31) points out, alerting the reader to recognising the disease is not to suggest that the disease can therefore be anticipated and avoided. It is more a case of accumulating wisdom and experience in a general way needed by the reader who was most likely to be participating in the day to day running of the [insert Greek], an idea that Polybius was later to take up in his [insert Latin]. The medical analogy of diagnosis and of prognosis should therefore not be taken too literally by the modern commentator. It is Thucydides’ view of human nature that has most contributed to the idea of Thucydides as a ‘scientific’ historian. From Collingwood onwards it has become the accepted widom, an orthodoxy that has uncritically permeated the work of others: a view that I’m most anxious to dispel. That a writer makes assumptions about human nature should not invite the ascribtion of scientific. Liberals and Marxists do just that – their political philosophies run on the supposed immutability or mutability of human nature. Thucydides’ view of the immutability of human nature, [insert Greek], does not entail a vulgar notion of scientific determinism but is best viewed in a looser sense of ‘the same or similar to’ (de Ste. Croix 1972: 32). Only the vulgar Marxist, de Ste. Croix not being one, would make the fatuous claims for ‘iron laws of history’. This scientism manifests itself in Collingwood’s notion of psychological history which he attributes to Thucydides – ‘the father of psychological history’ (1946: 29-30). He writes: “Now what is psychological history? It is not history at all, but natural science of a special kind. It does not narrate facts for the sake of narrating facts. Its chief purpose is to affirm laws, psychological laws. A psychological law is not an event nor yet a complex of events: it is an unchanging rule which governs the relations between events.” Cornford (1907: 59; 64) many years earlier had already surmised that a typical feature of ancient historians was to ascribe causal explanation solely in terms of psychological causes. Collingwood’s and Dilthey’s process of re-enacment by which the historian empathetically enters into the state of mind of the person whose action he wants to explain might seem to lend support to Thucydides’ intention: for example, Thucydides’ avowed intention to capture an idea of ‘national character’ as in III: 82 (the Athenians) and in I: 68-72; 80-85 (the Spartan character). De Ste. Croix (1972: 5-6; 31-32; 1981: 27) is absolutely correct in suggesting that it is ‘patterns of behaviour’ or trends as we mentioned in section I, that Thucydides is concerned with. What de Ste. Croix fails to do is offer a convincing refutation of this quasi notion of ‘law’, a task I now undertake. There is a tension in Collingwood’s idea. In trying to understand Pericles’ action I want to understand this particular person’s specific action; I am interested in the unique, not in the general. I gain no help, and seek no assistance from appeals to probabilistic generalisations about the tendency of this type of person to do this type of action in this type of situation. It does not alter the fact that Pericles wasn’t a ‘Periclean’ type person who fought in a ‘Potidaean’ type battle. McClelland (1975: 90-91) asks, then how do we gain knowledge of the dispositions of others and knowledge of how people with those dispositions are likely to behave?: . . . that the answer would seem to involve the perception of similarities and the organisation of those similarities into generalisations of varying degrees of specificity. If the historian grants this latter point, he denies the basic methodological precept of both Dilthey and Collingwood, that one can pass directly ‘from the facts to an understanding of those facts with no reliance upon generalisations’. Collingwood’s Thucydides is a nomological Thucydides, a view that is untenable (cf. Dray’s ‘rational explanation’).


To be objective is to ostensibly be scientific and could not be coherently discussed had we not already specified the paradigmatic nature of scientific knowlege and its relationship to historical knowledge. Charges of historical subjectivity – of the subjectivity of historical inquiry – usually proceed in two stages (Martin 1979: 49). First, an attempt is made to show that historical inquiry necessarily carries a value component – that all historical judgments are necessarily determined by value judgments. Secondly, value judgments are held to be incapable of justification since there is no evidence in virtue of which one judgment of that kind is more justified, that is, more worthy of belief, than some other judgment of that kind which is incompatible with it. Consequently, the charge of historical subjectivity will be defeated if we can show either that: ç¤ it is not the case that all historical judgments are necessarily evaluative, or that evaluative judgments are not necessarily subjective For the purposes and constraints of this paper, we’ll confine ourselves to examining the idea that all historical judgments are necessarily evaluative. This is certainly a major idea that people have in mind when they talk about subjectivity and we’ll look at it in relation to Thucydides. Critics of historical objectivity claim to find three evaluative elements in historical inquiry (Dray 1957: 250): ç¤ selection of items for historical investigation evaluation in causal explanation ç¤ evaluation in action description In I: 21 Thucydides offers an explicit and unambiguous statement of the distinction he makes between mythology and ‘factual’ history, a thinly veiled attack on Herodotus. However, we find at the close of Book III Thucydides devotes a significant amount of space to the Homeric Hymn to Apollo when Thucydides explicitly has said that mythology is not a ‘truth-bearer’, that is the kinds of things that can be true or false?, the candidates being ‘what is said’; beliefs; and knowledge, which by definition cannot be false. He says: ” . . . I do not think that one will be far wrong in accepting the conclusions I have reached from the evidence which I have put forward. It is better evidence than that of the poets, who exaggerate the importance of their themes, or of the prose chroniclers, who are less interested in telling the truth than in catching the attention of their public, whose authorities cannot be checked, and whose subject-matter, owing to the passage of time, is mostly lost in the unreliable streams of mythology.“’ Thucydides’ intention cannot be faulted even from a modern perspective (Oakeshott 1933: 74): “An imagined past, a past of myth, may, for some purposes, be valuable and significant; it may for example, have an immense and beneficial control over belief and activity; but it is a past alien from that of history.” I would argue that it is a failing both on Thucydides’ part and on the part of the modern commentator (Collingwood 1946: 28-34; cf. de Ste. Croix 1972: 5-6) that Herodotus’ project is misconstrued. Given that Thucydides is using Herodotus as a methodological foil we need to briefly examine their points of convergence and divergence. Herodotus and Thucydides both share a concern for the fallibility of conscious memory (the act of recalling) for the transmission of past events but for different reasons. In the opening sentence of Herodotus’ Histories he says: ‘ . . . (his) Researches are set down to preserve the memory of the past by putting on record the astonishing achievments . . . ‘ For Herodotus, the qualititive dimension of memory was its breadth, not its criterion of truth. For Thucydides and the modern to evaluate Herodotus’ project even on the loosest definition of historiography is to completely misunderstand the nature of the Herodotean project. Herodotus’ recorded stories may or may not be true but his project was to record stories that formed the fabric of an inherented cultural tradition. He does not naãÈvely believe all he hears (II: 123; VII: 152) and it is not clear to me whether it is the veracity of the stories themselves that is doubted, whether it is Herodotus’ recording of the stories without his embellishing them that is doubted, or whether or not he shouldn’t be in that line of business at all (Burns 1972: 28). Thucydides was implicitly critical of Herodotus for downgrading fact in preference for emphasis on artifice or entertainment. That would be a cogent criticism were Herodotus claiming to be writing ‘factual’ history, a claim he never makes. Perhaps viewing Herodotus as an anthropologist of sorts (Burns 1972: 10) best captures the nature of his project. I use the term anthropologist in its loosest sense akin to explorers and travellers such as Marco Polo, discovering the diversity of mankind, even if their ‘discoveries’ were not entirely empirical. This view is consistent with Herodotus’ assertion in II: 123, repeated again in VII: 152 (cf. 8: 8; 3: 116; 4: 25): “’ Anyone may believe these Egyptian tales, if he is sufficiently credulous: as for myself, I keep to the general plan of this book, which is to record the traditions of the various nations just as I heard them related to me.“’ Keeping these thoughts in mind we now turn to Thucydides’ speeches which have generated a great deal of scholarly dispute. In I: 22 he sets out his famous ‘method’: ” . . . I have made use of set speeches some of which were delivered just before and others during the war. I have found it difficult to remember the precise words used in the speeches which I listened to myself and my informants have experienced the same difficulty; so my method has been, while keeping as closely as possible to the general sense of the words that were actually used, to make the speakers say what, in my opinion, was called for by each situation. He goes onto explain that with so many contradictory accounts through impartiality and the natural limitations of human memory, the truth was very difficult to divine. He ends I: 22 with the famous appeal for his work to be of relevance in perpetuity, premissed on the notion of a fixed human nature, [insert Greek], which we looked at earlier. It is the ambiguity of the phrase ‘keeping as close as possible’ that has been the source of so much dispute concerning Thucydides’ ‘proclaimed’ objectivity. It is not my place to adjudicate the correct rendering of [insert Greek] (judgement, opinion, intention in Liddell & Scott 1993; purport, sense, meaning: von Fritz; upshot: Walbank both cited in de Ste. Croix 1972: 8-9) or of [insert Greek]. Many commentators find it difficult to resist the conclusion that the judge of ‘what was appropriate’ was Thucydides himself and that this is a devastating blow against Thucydides’ claims for factual history. Surely as long as the ‘main thesis’ (de Ste. Croix 1972: 10) or the ‘upshot or gist’ (Walbank 1965) is communicated Thucydides’ intention will have been met. Jebb (1913: 253) suggests that the first thought that would occur to Thucydides would be, ‘What were the best arguments available? rather than, ‘What would have the agent himself have used’ subject to qualification on how well the speaker was known to Thucydides. Jebb (1913: 256) does however suggest that ‘there is every probability that he had heard most or all of the important discussions which took place in the Ecclesia between 433 and 424 B. C.’ and that Thucydides would have had access to a full and accurate account of the speeches he may have not witnessed for himself. Pompa (1993: 14) echoes this point that given the public nature of an event such as a war and the public nature of the actors speeches etc., there is little reason to doubt the veracity of events that took place in the public domain. 


It is my contention that the key to a sound understanding of Thucydides in particular and historical explanation in general lies in accomodating the insights of both the idealists and the so-called scientific historians but not allowing these insights to contrive a fictitious Thucydides. We cannot escape that: history is on one level a distinct discourse, primarily because history is what the evidence obliges us to believe, not what actually happened’ (Oakeshott 1933: 112) and thus in the absence of another history Thucydides’ history is the history of the Peloponnesian War. As for the claim that Thucydides is most vulnerable to charges of ‘subjectivity’, this can be countered in two ways:

First, it’s more of a case of commentators not being sensitised to the epistemological problems of historical explanation, being easily being drawn in by Thucydides’ suggestive ‘scientific’ language, then, unfairly setting him up as a straw man.

Secondly, as for the charge levelled against Thucydides that he is merely a conduit for Periclean ideology, that can be met by taking the Marxist line that any intellectual enterprise, is subject to and functional to the dominant ideology of that particular society, more so in Thucydides’ case given Pericles’ unparalleled dominance within Athenian society. To vulgarly apply the assumptions of science to history is to be as bad a ‘historiographer’ as it is to be a bad ‘scientist’, and ends up doing neither. Thucydides, then, should be exonerated from the fallacious charge of being unscientific or indeed of being scientific, a view that has arisen from a vulgar notion of scientific/ historical explanation. Thucydides’ profound and lasting insight was his identification of the epistemic problem of the tenuousness of distant history. Seen thus, Thucydides in the context of his time was as good a historian as any good modern historian in the context of our time.


1 Patrick Gardiner writes in his The Nature of Historical Explanation (1961: xi): ‘. . . (the question) “Is history a science?” is a question dangerous in its simplicity, particularly so because to call some branch of study scientific is frequently a means to giving it an air, if not of glamour, then at any rate of respectability. And all too often bold attempts to give a straight yes or no answer to the question, supported by reasons, end up in verbal disputes amounting to little more than alternative linguistic recommendations.’ Gardiner goes on to offer the corollary: ‘The question: “What is the nature of historical explanation? . . . implies that, provided a careful enough search is conducted, a clear and distinct idea of what historical explanation really is may somewhere be found, and, with labour, brought to light. This thought no doubt originates from Thucydides’ famous statement in I: 23: 6 about the origins of war. De Ste. Croix (1972: 5-6) is equally critical of modern (English) Thucydidean scholarship for failing to clarify and locate Thucydides’ History within the broader problems of historiography, for example, in attributing to Thucydides a search for ‘laws’ in the vulgar sense rather than his having an interest in historical events.

2. Though de Ste. Croix is characteristically provocative, his diagnosis is spot on. He fails, however, to offer arguments that match the force of his diagnosis.

3. Popper held that we should aim, not to confirm or corroborate, but to refute of falsify hypotheses; a counter-instance carrying more weight than a confirmatory instance.

4. see The Philosophy of Social Science, ed. A Ryan, Oxford (1973) particularly Ryan’s introduction pp. 4-5.

5. Several of the great philosophers of history were idealists – Croce, Collingwood, Dilthey but I will take Oakeshott as the preeminent representative. Suffice to say, that as an idealist, the nature of truth has to be coherence (the coherence theory of truth not to be confused with the coherence theory of knowledge) and each ‘mode of experience’, i. e. the historical, the practical and the scientific are intellectual constructs, each capable of “’generating conclusions to itself“’. Therefore history would be considered sui generis as would science (see Oakeshott 1933). For a detailed and balanced account of the coherence theory of truth see Ralph Walker’s The Coherence Theory of Truth, London & New York (1989).

6. de Ste. Croix (1972: 6 note 4) cites MI Finley among others.

7.  In his “’Preface To A Critique Of Political Economy“’, Peking (1976), Marx talks of the Ancient world as being one of the epochs comprising the process of dialectical materialism. For the most sustained modern treatment see GA Cohen’s Karl Marx’s Theory of History: A Defence (1978), Oxford and his ‘Functional Explanation, Consequence Explanation, and Marxism’ Inquiry 25.


Berlin I (1980) “The Concept of Scientific History” in Concepts and Categories, ed. H Hardy, Chicago

Burns AR (1972) ‘Introduction’ to Herodotus’ Histories, tr. A de Selincourt, Harmondsworth

Bury JB (1930) ‘Cleopatra’s Nose’ in Selected Essays, ed. H Temperley, Cambridge

Collingwood RG (1946) The Idea of History, reprinted 1970, Oxford

Cornford FM (1907) Thucydides Mythistoricus, London

Dover K (1994) Marginal Comment: a memoir, London

Dray WH (1957) Laws and Explanation in History, Oxford

Finley MI (1968) Aspects Of Antiquity, London

Finley MI (1972) ‘Introduction’ to Thucydides’ History of the Pelopponnesian War, tr. Rex Warner (1954), Harmondsworth

Froude JA (1864) ‘The Science of History’ in Short Studies on Great Subjects, London

Gardiner P (1961) The Nature of Historical Explanation, Oxford

Gaselee S (1939) ‘Ancient Greek Scientific Ideas’ Proceedings of the Classical Association Vol XXXVI

Goh ST (1970) ‘Some Observations on the Deductive-Nomological Theory’, Mind

Gomme (1945) A Historical Commentary On Thucydides, Oxford

Hempel CG (1962) ‘Explanation in Science and History’ in Frontiers of Science and Philosophy, ed. RG Colodny, London, reprinted in Explanation (1993), ed. DH Ruben, Oxford

Hornblower S (1987) Thucydides, London

Jebb RC (1913) ‘The Speeches Of Thucydides’ in Hellenica, ed. E Abbott, London

Lindley R (1978) ‘Scientific Explanation’ in What Philosophy Does, London

Martin R (1979) ‘History and Subjectivity’ Ratio 21

McClelland PD (1975) Causal Explanation and Model Building in History, Economics, and the New Economic History, Ithaca NY

Momigliano A (1958) ‘The Place of Herodotus in the History of Historiography’, History XIII

Momigliano A (1971) ‘Popular Religious Beliefs and the Late Roman Historians’  Studies in Church History 8, reprinted in Essays in Ancient and Modern Historiography (1977), Oxford

Momigliano A (1972) ‘Tradition and the Classical Historian’ History and Theory 11, reprinted in (1977), Oxford

Momigliano A (1974) ‘Historicism Revisited’ Mededelingen der Koninklijke Nederlandse Akademie van Wetenschappen, Afd. Letterkunde, NR 37: 3, reprinted in (1977), Oxford

Momigliano A (1979) ‘The Origins of Universal History’, presented at the University of Chicago and in the Christian Gauss Seminars in Criticism at Princeton University, reprinted in On Pagans, Jews, and Christians (1987), Middletown Conn.

Momigliano A (1981) ‘History and Biography’ in The Legacy Of Greece, ed. MI Finley, Oxford

M (1933) Experience And Its Modes, Cambridge

Oakeshott M (1983) On History And Other Essays, Oxford

Pompa L (1993) ‘The Possibility Of Historical Knowledge’ Aristotelian Society, Supplementary Vol LXVII

Ryan A (ed.) (1973) The Philosophy of Social Science, Oxford

de Ste. Croix GEM (1972) The Origins Of The Peloponnesian War, London

de Ste. Croix GEM (1983) The Class Struggle in the Ancient Greek World, London

Walbank FW (1985) ‘Speeches In Greek Historians’ in Selected Papers: studies in Greek and Roman Historiography, Cambridge