How a Historian Establishes “What Happened” when “we only have the words of the text”

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

by Neil Godfrey

If all we have is an ancient historical or biographical narrative that we cannot verify by independent evidence (and keeping in mind that, as we saw in the previous post, external claims also need to be capable of verification) then how can a historian go about deciding how much of the narrative is likely to be true?

Pericles’ Funeral Oration (Perikles hält die Leichenrede) by Philipp Foltz (1852) — Wikimedia

Continuing with Peter Kosso’s argument we come to his fourth method of verification, an examination of the internal features of our document. Kosso is using Thucydides as a case study.

There is the fact, abhorrent to modern historians, that he never tells his sources [at best he only says he garnered information from (anonymous) eyewitnesses and his own experiences — my note], and that he never justifies his opinions.” There are no arguments in Thucydides, and no footnotes. These silences force the judge of his credibility to use internal methods, since they eliminate the easiest way of finding other, independent sources of information. (p. 9)

Take the long speeches he puts into the mouths of key actors. Thucydides explains that it was obviously impossible to report these accurately but he attempted to reproduce what he believed would have been the general sense of what each person said. Thus,

With his own words Thucydides makes us uneasy over his veracity and he plants the worry that the message of the speeches may be as much a report on his own opinion as on the facts of the matter. (p. 9)

Internal features compatible with accuracy and objectivity

  • Vivid and full of detail

The writing is exceptionally vivid and full of detail, “participatory” in the sense that the reader is drawn in to relive the events. This is reminiscent of Hume’s suggestion that the products of imagination are less vivid than the products of observation. (p. 10)

That sounds fine at first blush, but of course a moment’s reflection will warn us of the catch.

But of course a good novel can be vivid and participatory, and many works of fiction are livelier and more real-seeming than The Peloponnesian War. Attention to detail and realistic style, in other words, are not necessarily indicative of truth. (p. 10)

  • Expressing divergent opinions

Thucydides gives us two sides of the story when he sets out his speeches. He will allow a figure to present the Spartan point of view as well as another to give the Athenian one. That he does so suggests to us that he is trying to be fair and even-handed.

Presentation of all sides is of course possible to do in fiction as well, but it is perhaps less likely, since good fiction intends to make a point. Thus Thucydides’ reporting from a variety of perspectives would be a symptom of his objectivity if it remained evenhanded and no discernible opinion, no favored perspective of the events, emerged in the narrative. (p. 10)

But a close reading of Thucydides will reveal another, far less objective or historical, purpose for the presentation of these diverse viewpoints. At this point I leave Peter Kosso’s article for a moment and turn to a closer look at another article that Kosso cites:

Wallace, W. P. 1964. “Thucydides.” Phoenix 18 (4): 251–61. https://doi.org/10.2307/1086359.

Opposing the view that Thucydides is the first “scientific historian” is another perspective that believes he is more akin to writing as a philosopher or tragedian:

The means by which Thucydides brings his readers to see the facts as he saw them were made the subject a few years ago of the careful and interesting book by Mme de Romilly [link is to the more recent English translation] which was mentioned above. Her dissection has laid bare Thucydides’ method. The intent, the plan, the purpose of every series of events — and this is equally true whether the events in question compose a skirmish, a battle, or the war itself — the underlying idea is suggested beforehand in a speech, in a remark about what naturally would or usually does happen under such circumstances, in some aside about how men always act. The reader has thus been prepared beforehand, the probable explanation has been suggested to him; there has been no blunt and partisan statement of the private opinion of Thucydides, but some reference to the invariable character of human behaviour, some cold and almost statistical generalisation about the usual result in such cases, thrown in as it were in an aside, has prepared his mind without arousing his opposition. So when he comes to the simple and apparently objective narrative of bare events, he knows at once what he thinks about them. And in case he has forgotten, in case his attention has lagged, the key words of the previous explanation, the very phrases which were used before, arc quietly repeated in the exposition of events so carefully selected and so subtly coloured that every reader comes to the same conclusion about them, and comes convinced that he thought of it for himself. No totalitarian meeting of voters assembled to elect a single slate of candidates has ever been more unanimous than the readers of Thucydides in assessing the issues of the Peloponnesian War. The effect depends to a considerable extent upon what one may almost call subliminal persuasion, upon careful repetitions and echoes of words and phrases. It is probable that most of this does not reach the level of any reader’s consciousness, but analysis of the text reveals the method, and its effectiveness is proved by the unanimity it has produced. Such a method of predigesting facts, such careful presentation of only the most palatable and nourishing provender, produces happy readers. Not for them the knitted brow, the puzzled mind. The drama hurries them along; they are in the grip of fear and pity; it would seem irrelevant to ask if that is really exactly how it happened. (pp. 257f)

What is going on here? Thucydides writes a vivid and detailed drama in a way that compels all readers to share his viewpoint. An examination of themes and descriptors and choices of stories has suggested to a number of historians that Thucydides has no interest in “what actually happened” from a historian’s point of view — such a pursuit he and his philosophical peers would consider to be trivial and unintellectual. Rather, Thucydides’ views human affairs as governed by grand laws or principles that he wants the readers to learn. He is attempting to inform readers of a deeper reality that governs human destiny.

It is a reality which, I think, he never found. A will o’ the wisp which led him, as it has led other great historians, into the bog of pseudo-explanation, the kind of explanation which has made some sec in history gold, silver, bronze, and iron age cycles, or preparation for the return of the Messiah, or the patterned rise and fall of civilisations. It is to Thucydides’ credit that he never made explicit the explanation towards which he obviously felt that he was making his way. In the humble and honest manner of the Hippokratic doctors, but also with their implicit faith that what they wrote would some day be understood, he set forth the particulars of the case he was describing, disentangled its symptoms, and underlined its tendencies. He never doubted that, as a careful and accurate case history, his work would take its place in the ultimate Principia Medica of mankind. He thought, to paraphrase his own words, that while the tales of Herodotos were well calculated to fill an idle moment pleasantly, his own history would be a permanent possession of mankind — permanent surely because it would be part of the Science of Man. (pp. 260f)

To sum up in Kosso’s words:

Wallace  . . .  sees Thucydides’ narrative as selective in a way that stresses the enduring and ideal features of humanity and society, in exclusion of the ephemeral and common. Thucydides, in other words, pursues a Platonic point and reports only those details that serve his purpose. Hunter makes a similar assessment of the lack of objectivity but describes a philosophical directive that is counter to [above] view of the determinism in Thucydides. It is more about free will and the ability of people to intervene and take control of the activities around them.

Thus the apparent epistemic virtue of having presented both sides of each story is balanced by a pervasive epistemic fault of directing the emphasis and selectivity of the narrative to suit a philosophical cause rather than simply to report the facts. In the tension between telling the speeches as they were or as they ought to have been, the tension between historical accuracy and philosophical conviction, there is reason to think that philosophy gets the upper hand. (p. 10)

Historians have increasingly come to learn through other new evidence (especially archaeological inscriptions)  that Thucydides completely ignores what must have been front and centre of the leaders and populations of the period — in particular economic pressures, the bankruptcy facing Athens, the economic pressures that led to revolts, massacres that were related to pressures of economic destitution. Such facts must surely present a very different type of history from the one Thucydides has given us. Also instructive is Thucydides’ neglect to give us any character sketches of generals or political leaders. Even vital details of military operations are found critically lacking. (Wallace, p. 255)

Kosso calls upon another historian who even suggests that Thucydides’ “point-counterpoint presentation of speeches” is really a sign of “contrivance”, “too well staged to be real”. The situations of the rival speeches are artificial. The speakers often were not in a position to know what others were thinking and to therefore respond with comparable rhetorical arguments. The speeches, moreover, are “excessively obedient to the rules of rhetoric . . . beyond belief.” The speeches can therefore undermine our confidence in the historical accuracy of what we are reading rather than bolster it.

Internal features helping us decide what not to believe

Instead of optimistically pulling out “criteria of authenticity” (a long time staple of the study of the gospels by biblical scholars) Kosso pessimistically informs us that

Virtually all internal features of the text that are epistemologically significant are eliminative, that is, useful in helping us decide what not to believe. (p. 11)

One such feature is inconsistencies in the narrative. So when Thucydides reports Pericles delivering a speech endorsing the strategy of fortifying headlands near the enemy’s coast, but then indicates that one key such area was not fortified, the reader is left in a quandary. Was Pericles “quoted” accurately? Or is Thucydides mistaken in thinking one particular headland was not fortified? Or was Pericles himself inconsistent in his words and actions? Without any explanatory comment by Thucydides we are left wondering.

Another “eliminative” feature is Thucydides’ “implicit reliance on seemingly inaccessible information, particularly other people’s ideas and motives.” Thucydides does not even qualify his statements with “supposedly” or “must have thought”. Yet he comes across as quite certain when he says, for instance, that a certain general “knew he was becoming unpopular” and was “now thoroughly scared.” In the works of historical Jesus scholars one often encounters a sentiment such as “there is no reason to doubt (what the text says)”. But compare Kosso:

There is no reason to regard these descriptions as anything but the author’s invention since Thucydides was not, as we piece together from his own account, close to Kleon. Hornblower, underscoring doubt about this detail of Thucydides’ history, says of the historian: “He is not likely to have been in Kleon’s confidence.”

What the historian can do when examining the text in the search for features that might help us establish its reliability is be alert to those features that “eliminate” our confidence.

One important generalization presents itself. In the closely internal constraint of the text itself, epistemic features tend to be eliminative but not positive. Features of the text more readily indicate what is not likely, than what is likely to be true. (p. 12)

But what about the positive features we find in the text, the ones we noticed at the beginning of this post?

But any positive features of the text, such as consistency, vivacity, or attention to detail could be the result of polished fiction. These may be features that are necessary to an accurate report but they are by no means sufficient, even collectively, to indicate likelihood of truth. (p. 12)

So we come back to the scientific method of falsification. If we want to justify the historical account of a text where we have no recourse to external corroboration then the only certain way is to look “for signs of falsehood”. By looking for “signs of truth” we are prone to find as many as we might like.

The process of justification consists not of looking for signs of truth, but for signs of falsehood. Look for things that could be wrong but aren’t. This is the only strategy for dealing with the predicament of having only eliminative epistemic criteria. (p. 12)


In this post I have focused on the problem facing historians when they lack any significant external checks. We covered what those independent tools look like in the previous post.

Kosso concludes by reminding readers that the best historians can do is to arrive at “indicators of what could not be true rather than what is likely to be true.” Yet even when we put everything together, confidence in a narrative will increase to a level of relative certainty only as the narrative continues to pass tests posed by new evidence and information over time.

Or in his own words,

This narrow internalism [studying internal features of a text for historical reliability – my note], while best avoided when more independent information is available, has been informative of the nature of justification. In the closely internal attempts at accounting, individual epistemic features are only eliminative, indicators of what could not be true rather than of what is likely to be true. This suggests a similar status for broader forms of internalism as well. Even the widest consistency throughout a system of claims about the past, and extensive coherence in terms of explanatory connections and relevance of details, is possible for a thoroughly false system whether it is false by contrivance or by mistake. Coherence itself does not indicate a likelihood of truth, but continued coherence in a process of accumulating claims (more observations, more data, more texts read) adds credibility by surviving the increasing chances of failure. The credit is in passing the potentially eliminative tests, that is, of avoiding incoherence with the addition of new data. This requirement is not part of the coherence; it is in addition to coherence. Coherence in the system must be maintained under the challenging influence of new observations. The system of historical claims that is coherent by contrivance or fortuitous mistake is less likely to remain coherent when forced to account for a steady input of new evidence than is the system that is coherent because its claims are true.

Making effective use of the eliminative epistemic features in an internal process of justification requires the dynamic kind of coherence. Coherence in a static system of claims is not a hallmark of justification. The maintenance of coherence, that is, the avoidance of eliminative features such as inconsistency, is the epistemic virtue. In this light, internal coherence justification is seen as an eliminative process not unlike the Popperian suggestion of falsification, and an epistemology of coherence finds a clearly supportive example in the epistemology of history. (p. 13)

This post and its companion have not applied Kosso’s arguments to an evaluation of the methods most common among biblical historians. But both posts, along with similar ones, will serve as a point of reference in future posts addressing biblical studies. Meanwhile, interested persons are more than welcome to suggest their own applications to the study of Christian origins.

Kosso, Peter. 1993. “Historical Evidence and Epistemic Justification: Thucydides as a Case Study.” History and Theory 32 (1): 1–13.

Wallace, W. P. 1964. “Thucydides.” Phoenix 18 (4): 251–61.



The following two tabs change content below.

Neil Godfrey

Neil is the author of this post. To read more about Neil, see our About page.

Latest posts by Neil Godfrey (see all)

If you enjoyed this post, please consider donating to Vridar. Thanks!

2 thoughts on “How a Historian Establishes “What Happened” when “we only have the words of the text””

Leave a Comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Discover more from Vridar

Subscribe now to keep reading and get access to the full archive.

Continue reading