Ruarigh raised an important point in a comment, and it deserves a longer consideration than it has had so far.
The issue is translation.
Now, decent translations of many ancient texts are available at reasonable prices in English, and these are what I’ve been using. So, Caesar, Tacitus, Seutonius and, more recently Herodotus grace my bookshelves.
However, there are problems in using them. The Penguin translations, for example, have an irritating habit of translating ‘cohort’ to ‘battalion’ and ‘legion’ to brigade, let alone ‘pilum’ to ‘spear’. Piecing this back to what is actually meant by the author is a frustrating business, but I don’t read Latin or Greek, or anything else except English (and some might argue I don’t do that very well!).
There is a problem here, outlined by the US philosopher Willard Quine.
Consider a sentence, S, in a language system, call it L. S only has meaning in L by virtue of other sentences also in L. Therefore, the meaning of S is not fixed by S, but by the whole language system L. The effect of this is to show that translation is always indeterminate. A translation from L into L’ is always underdetermined by the data and thus open to question.
For Quine, then, everything is up for grabs because everything is under-determined by the data and we can only choose, for example, which translation we prefer.
In my view, it gets slightly worse than this. Consider an ancient language system G. G exists, and thus gives its sentences S meaning in a specific culture, C. If C is not understood, then some aspects of G will remain unclear, as the translation is not just from one language to another, but from C to another culture.
Thus, for example, we see the Greeks rushing off to Delphi at the slightest opportunity, to consult the oracle. Why? The oracle was a powerful religious and cultural force in the culture, but unless we know this part of the culture (and we never can fully know what it meant for the Greeks), the full meaning of the language, will elude us.
Therefore, we seem to be facing a double whammy in terms of translation. There is Quine’s indeterminacy, but there is the further problem of translating from an ancient culture, and if C determines L and L determines the meaning of S, then C, to some extent, determines S.
To put this back into English, the culture of the time determines, to some extent, the meaning of the sentences we read, especially when translated. We can understand the words, and even the sentences, but the cultural block can remain.
The upshot of this for wargaming is tricky. According to our sources, I should be making sure that the players send off to Delphi before any major decision, sacrifice sheep before going into battle and undertaking all the cultural paraphernalia that can be derived from our sources that the Greeks did.
But there is a further problem.
Our sources can only give a partial account of what happened, even if those sources are concluded to be accurate. So even if we’ve got good sources (which is rare in the ancient world), we still only have a partial picture of what happened, and an even worse view of what might have happened.
Furthermore, in many cases we cannot check our sources. For early Greek history, there is only one source, Herodotus. We could dismiss Herodotus as being inaccurate. Certainly, he is not all that accurate in places where he can be checked. But if we do dismiss him, we cut off the branch upon which we are sitting. We can then say nothing else about the period except a few scraps of archaeology.
Given all this, it is a wonder that we can get anywhere at all. But we can, slowly and carefully, if we engage with our sources (albeit, in my case, in translation) and with secondary sources in books and journals which try to interpret those sources alongside others and with the archaeology of the period and places of interest.
As I’ve observed before, the danger is lifting a piece of our source material as good evidence for something, but then finding it is downgraded by different interpretations or translations of the same thing. The trick is to try to work holistically, trying to grasp the culture as well as the battles, but that is hard work, and very, very slow.
But it does explain, at least to my satisfaction, why I’m reading the whole of Herodotus, not just the paragraph or two about Marathon.