Wikipedia talk:Primary sources in classics

Comment

@Psychastes, HLHJ, Caeciliusinhorto, and NebY: Comments welcome on this essay. Ifly6 (talk) 12:48, 27 August 2024 (UTC)

Nice work, and I do like your scholarly, sourced and technical examples in What is the big deal with primary sources? I do wonder if it could be more persuasive if those were preceded by some more commonplace ones. There are ways we all do - or at least should - now treat ancient historians as unreliable; a few, off the top of my head:

troop numbers, casualties, accounts of engagements

numbers in general

numbers that stand for "a lot" or "a helluva lot"

precision for verisimilitude

resonant numbers - 10-year siege, but of Veii; 300 fighting to the last, but the Fabii and only one survived, not two

speeches (hardly novel - cf Dionysius on Thucydides, Polybius on Timaeus, Diodorus on everyone)

bias, tainting not only tone and description, but the very inclusion and exclusion of material (Xenophon, Tacitus ...)

emulative falsehoods and dramah (Sallust taking details of battles from Thucydides, because that's how you write battles, Cassius Dio adding colour and horror).

Of course, even when we accept that we must read ancient historians sceptically, it's sadly tempting to treat them as reliable if filleted, like treating the Gospels as factual so long as you skip the god bits. Still, opening with such content might serve along the lines of "we know they're unreliable, and it's even worse than you might think."

Meanwhile I've been idly watching a discussion about primary sources in articles about very recent events (Wikipedia talk:No original research#New articles based on primary sources) which in part relates to the shifting definitions of "primary" - eyewitness accounts and court documents are more primary than reports of the day's proceedings, which are more primary than newspaper feature articles, which would be primary if published 100 years ago but may now be close to secondary as we can get. At millennium scales, the "secondary" status of historian shifts too. In some ways, Wikipedia:Reliable sources/Perennial sources is independent of PST. Its main categories are

Generally reliable in its areas of expertise

No consensus, unclear, or additional considerations apply

Generally unreliable

Deprecated

Blacklisted

I wonder in which category the community would put Herodotus? NebY (talk) 19:39, 28 August 2024 (UTC)

Ngl, I laughed aloud at numbers that stand for "a lot" or "a helluva lot". I'm intending to write a short paragraph inserted before "Primary sources are wrong" touching on the points you made. I had in a previous draft included a heading for such a section but it seems to have disappeared so I never bothered writing it. Outline-first has its issues. Ifly6 (talk) 00:32, 29 August 2024 (UTC)

As to Hdt, I would expect him to fall into additional considerations apply except on his gold-digging-ants-in-Afghanistan and the-Persians-attacked-with-100-billion-men stuff. Ifly6 (talk) 00:35, 29 August 2024 (UTC)

Having checked, the number is exactly 5,283,220 per Hdt 7.816. Beyond the fact that France in 1914 had total mobilisation across the entire empire of 3.78 million men making Hdt's figure just absolutely hilarious, the precision of Hdt's figure is similarly laughable. Like actually hilarious. He says he knows down to the tens place how many men Xerxes brought into Greece. Not 5,283,210 or 5,283,230 men, Hdt counted that extra tent over there. Ifly6 (talk) 15:47, 29 August 2024 (UTC)

I'm sure he had a very reliable source, quite as good as the Egyptian priests who told him that twice in the preceding 11,340 years the sun rose in the west, and twice it set in the east, but it never made any difference. (Hdt 2.142)

Maybe, if you've not seen it before, you'll enjoy a Polybius quote about Timaeus, as an example of historians arguing about how to compose speeches.

For he has not written what was said nor the true sense of what was said, but instead offers what he thinks should have been said, and enumerates in all these speeches the concomitant details like someone at school trying his hand at a set theme; as if he were making a display of his own ability but not offering an account of what was truly spoken.

— Polybius 12.25a

NebY (talk) 16:16, 29 August 2024 (UTC)

Honestly it's like Polyb. 12.25a = ancient Wikipedia talk page (DD MM YYYY BC). Ifly6 (talk) 17:19, 29 August 2024 (UTC)

Fair! So I've put it on my user page.

I'm quite fond of Herodotus's math - so many ships, so many per ship, infantry counted 10,000 at a time by squeezing them into a special enclosure, and so on. The Aegean islanders contribute 17 ships and ships have 30 marines each, making 510, so there's a 10 in the total number of fighting men, 2,641,610. Then he suggests that there'd probably be more camp-followers, grain transport boat crew, and suchlike than there were fighting men but even assuming it was merely 1:1, that gives 5,283,220. It's like he's heard that rounding too soon introduces errors. Then he works out their rations and reveals he's not counted the women and eunuchs. NebY (talk) 20:49, 29 August 2024 (UTC)

Surely once he counts the women and eunuchs he has to incorporate that wagons will be needed to supply them and wagon drivers and their support staff. With this kind of logic, the Greeks aren't fighting Persians anymore. They're fighting the Zerg. Ifly6 (talk) 21:19, 29 August 2024 (UTC)

And then he mentions the dogs need feeding too. NebY (talk) 23:19, 30 August 2024 (UTC)

In one sense, Ancient literary sources ... are considered primary sources for the purposes of classical studies and All literary, epigraphic, and archaeological evidence from the ancient world are primary sources for the purposes of classical studies are unnecessarily restrictive; you could simply say "for Wikipedia", or add "and for Wikipedia". Wikipedia:No original research#Primary has a long footnote, currently [d], Further examples of primary sources include: archeological artifacts; ... medieval and ancient works, even if they cite earlier known or lost writings; tomb plaques and gravestones. (The stricture on ancient works has been in WP:OR since at least 2011, when archeological artifacts were already included.) Worth mentioning? NebY (talk) 23:41, 30 August 2024 (UTC)

@UndercoverClassicist: Also wanted to ping you! Ifly6 (talk) 19:31, 28 August 2024 (UTC)

Good essay -- I think it does a good job of fleshing out why WP:PRIMARY is as it is, particularly for sources in our field. I think your point about secondary sources which simply rehash primary ones is also important; I note that you tactfully stopped short of naming names there. If I may make a couple of suggestions, all very minor:

• I think it is worth being explicit that primary sources are only useful for evidence that can be verified from the source itself (WP:PRIMARY) -- in practical terms, this means for their own contents. In other words, a sentence cited to (e.g.) the de bello Gallico almost always has to begin with "in the de bello Gallico, Caesar writes that..."
• You have slightly misrepresented Bret Devereaux, who goes on to point out that this "rule" is actually pretty murky, and doesn't hold up well for the end of the Western Roman Empire (where we have big and important disconnects between what the literary sources seem to say and what the archaeology seems to show, and neither resolving those or ignoring them is an easy business), or for certain sub-fields like demography. The basic principle is sound, but even as an archaeologist by training, I'd stop short of saying "archaeology always beats text" -- as ever, it's more complicated than that.
• Likewise, I think you have been a bit heavy-handed on Cambridge Scholars Publishing, which sometimes gets a bad name from authors -- but that is largely a factor of its aggressive marketing tactics, and the perception that they reach out to (e.g.) newly-minted PhDs and sweet-talk them into accepting an offer when they might have got a better one had they taken the manuscript to an academic press and been prepared to do more work meeting their process. Their status is tricky, but very few people would go beyond "some issues, maybe handle with care" to "this is vanity press". See for instance a breakdown here. They do feature in Beall's list, but that listing is specifically about journals, which they no longer publish. The point isn't wrong, but the phrasing should, in my view, be softened.

Again, nice work on an essay that I hope will be read where it is needed. UndercoverClassicist ^T·C 21:03, 28 August 2024 (UTC)

These points are well taken. I've weakened the claims in the latter two, though I am also open to suggestions as to how they could (or should) be rephrased from here as well. As to the first point, I think this is addressed already (actually marginally more broadly) in the example with Cicero and the one-day consulship of 45 along with the following note that Reasons to use primary sources alone may include ... use of a primary source to support a claim that the source makes that claim. Ifly6 (talk) 00:29, 29 August 2024 (UTC)

Perhaps, but I wouldn't necessarily agree with the statement in the essay that citing the letters of Cicero to support the claim that Cicero was displeased by the election of Gaius Caninius Rebilus -- Cicero could have been being ironic, or lying, or using some cultural idiom or code that is not obvious today, or the text may be corrupt, or there may be other sources that cast doubt on it... In this context, I would only consider those letters good enough to support a claim that Cicero wrote that he was displeased, and then only with a supporting secondary source. This is a small distinction, admittedly -- as long as readers understand that WP:PRIMARY is the guideline and that this explains it rather than replacing it, there should be no real problems. UndercoverClassicist ^T·C 09:17, 29 August 2024 (UTC)

I've substituted expressed displeasure, which is what I think I was trying to say in the first place. Ifly6 (talk) 13:14, 29 August 2024 (UTC)

Totally offtopic here, but I find it credible that a Mediterranean lagoon would be drained by a tsunami, often called a tidal wave because it acts like a tide (even in the 21st century, with seismometer networks and tsunami warning systems, people stand around in harbours going "Wow! Why has the tide suddenly gone so far out? Let's go out and look!" -- and then, sometimes, they die; if the harbour drains, run for the hills along the local tsunami evacuation route). A tidal wave might well not be readily distinguishable from the effects of tides and waves, but of course it isn't 24-hour periodic, or just-the-north-lagoon hyperlocal, and it might well leave geological traces which presumably are not evident. Ancient observation is often more reliable than ancient theory. Herodotus's famous insistence that any report of the sun being to the north while rounding the south tip of Africa is an obvious lie...

The predominance of expensive sources in the classics is a problem for professional classicists, too. And society more generally. If any classicists reading this want to start a platinum open-access journal, Wikiversity will host it for you, for free.

Do the second and sixth bullet point imply that primary archeological papers are secondary sources here?

Could we clarify that it is entirely permissible to quote ancient sources (and PD translations), at reasonable length, to give context to the modern scholarship on what they mean? Classical scholars tend to assume that anyone interested in their work has all the relevant source texts (which may amount to a few hundred words altogether) memorized, which the average Wikipedia reader will not. So a quote often makes things make much more sense, like the extensive examples in this essay.

A quibble, but I find some of the sentence structures a bit hard to follow. I've done a bit of minor rephrasing and can do more if it isn't unwelcome. HLHJ (talk) 01:47, 31 August 2024 (UTC)

The story of the capture of New Carthage in Polybius is that Scipio waited for the tides to drain the lagoon north of the city, then marched his troops over the now drained lagoon, to captured the northern sea walls. They sacked the city that evening. Livy's story is the wind did it even though there's nowhere for the wind to have built up from. Most older treatments are of the view that it was both. I think this is methodologically untenable: why posit one impossible event when you can posit two probability-independent impossible events at the same time?

There are multiple levels of archaeological report. The ones that we get to read in stuff like the Journal of Roman Archaeology are many times something which tries to incorporate lots of different strands of evidence to argue for some narrative. I wouldn't call those primary sources in the same way a field report or daily dig log is a primary source.

Where would you want to put the translation element? Any such quote would have to be contextualised just like the citation.

No issues with your copyedits. Ifly6 (talk) 22:45, 1 September 2024 (UTC)

References

Move to WP space

I'd like to move this article to WP space, probably with the title Wikipedia:Primary sources in classics or Wikipedia:Ancient primary sources. Are there any objections? Ifly6 (talk) 16:34, 30 August 2024 (UTC)

Transmission, editing, and translation of primary sources

I meant to look over this when it was first written back in August, and again when it was moved to mainspace in December, so I'm only 9 months behind schedule in making this comment! I generally agree with everything in the essay: it's definitely a worthwhile thing to be able to point to.

I wonder if it might be worth briefly discussing the transmission of ancient literary sources – when e.g. Plutarch tells us in the Life of Pericles that Aspasia was prosecuted for impiety, we need to remember that Plutarch has taken this from an earlier ancient source (in this case we have a reasonable guess as to what that source might have been; in some cases we do not know), then there have been multiple layers of manuscript transmission, redaction, edition, and translation to create the "primary source" in the form that most Wikipedia editors and readers are reading it. In this case there's no reason to doubt that this is in fact what Plutarch did write, but that's not always the case. Sometimes (especially if it is written in verse) we can be pretty sure that the primary text as we have it is wrong (it's e.g. unmetrical or dialectically implausible) even if we can't be sure what the right text would be.

As an example of how this has consequences for the interpretation of a source, Sappho's Ode to Aphrodite first appeared in a printed edition in the 16th century in a very corrupt form; it wasn't until 1835 that the reading "εθελοισα" in line 24 was first proposed (making Sappho's beloved a woman for the first time) and this didn't become the standard reading until the 1960s. See also translations of Catullus which historically have often softened or omitted the more obscene parts, such as in Catullus 16. Caeciliusinhorto-public (talk) 12:20, 28 March 2025 (UTC)

I agree that transmission is also something worth while to touch on as an added reason to disbelieve what some internet version of Plutarch's Life of ___ says. I'm not sure where to put it in the current structure though. Have you any ideas? Ifly6 (talk) 19:31, 28 March 2025 (UTC)

I added a section on this. I'm also familiar with a dispute about Marius' family in Vell which I added. If you have a source for the claim about late emendation for Sappho I'd appreciate it if you added it. Thanks! Ifly6 (talk) 10:35, 6 April 2025 (UTC)

Some small mention of forgery?

It may not be a high-level problem, but I guess some editors might chance on ancient, medieval or modern forgeries of primary texts and think to use them. Relying on the letters of Plato would be bad, not recognising that Aristotle didn't write various attributed works less so, finding some of the forged 13% of Galenic works less likely still, but I've seen Clementine Homilies cited on Wikipedia as if Clement himself (as a source for the sexual activities of Greek gods), so I suppose anything might happen. Links might include pseudepigrapha, focused on religious works, and Outline of forgery#Examples of forgery. NebY (talk) 12:41, 13 May 2025 (UTC)

My thoughts are of the Ps-Sallust and Ps-Cicero which were probably declamation exercises. Even in old sources these are marked as attributed. But in either case (misattributions or forgeries) would not a rule that simply says "don't use it unless it is used by a reliable secondary source", somewhat like Wikipedia:USEDBYOTHERS, suffice? Ifly6 (talk) 14:34, 13 May 2025 (UTC)

Oh yes, the same "don't" applies. Prompted by the RSN discussion and reading Ehrman's Forgery and Conterforgery, I looked at the four categories in Wikipedia:Primary sources in classics#What is the big deal with ancient primary sources? and couldn't think quite which one applied, though "5. ancient primary sources are sometimes fake" might be too weighty. Maybe some mention under "Ancient primary sources are wrong a lot" that they may even be ancient, medieval or modern fakes? NebY (talk) 15:06, 13 May 2025 (UTC)

I'm not sure exactly where to put it. I agree that it should probably go somewhere. Maybe in the preface to the whole section? It also depends too as to what the source is used for. The invectives attributed to Sallust and Cicero are still useful for shedding light on the ancient world and the ancients' mentalities even though they are "fake". (Indeed, I have cited Beard SPQR citing those two invectives.) Even a fake from the mediaeval period could be used to give evidence on reception. But I would think that the core issue is that editors cannot presume to "just know" what is fake and what is not – I say this because some people attempt to confuse truth for verifiability – so should defer to HQRS on that matter. Ifly6 (talk) 15:38, 13 May 2025 (UTC)

With the promotion of transmission to "they might not say what they seem to say", I think that's probably a decent place essentially to add on "or be fake altogether". I expect that Ehrman's Forgery and counterforgery is related to patristic texts? There was a passage in How Jesus became God that related to how orthodox scribes essentially rewrote early church fathers to match the orthodoxy in their time and accused the heretics of doing essentially the same. Ifly6 (talk) 04:32, 17 May 2025 (UTC)

Yes, that seems a good place too, and you'll have a better idea of the whole structure of the page. Forgery and counterforgery considers all early Christian literature (canonical, heretical, and patristic) with a particular interest in the polemical purposes of the authors. His dramatic opening statement includes that only eight of the thirty works from the period of the New Testament "go under the name of their original author, and seven of these derive from the pen of one man." (Of course there's a theological view that the integrity of the Bible is not to be doubted, and many prefer to speak of "pseudepigrapha" rather than "forgeries".) NebY (talk) 15:57, 17 May 2025 (UTC)

I don't have the book. Do you think you could contribute the relevant passages with citations? Ifly6 (talk) 19:13, 17 May 2025 (UTC)

I've inserted a subsection, which you might like to trim. So far it only has one citation, though I'm inordinately pleased with that one. I can add Forgery and Counterforgery for the Galenic 13% and Ehrman's use of "forgery"; are others needed or are the wikilinks enough? NebY (talk) 20:55, 22 May 2025 (UTC)

I promoted it up and added the example of the pseudo-Sallustian invective against Cicero with citation. Some more citation would probably be warranted for the statistics about Galen and concerns with Aristotle. I also made some minor changes to the text, splitting the classical paragraph from the biblical one and inserting some language explaining who Bart Ehrman is. Ifly6 (talk) 20:56, 22 May 2025 (UTC)

Nice work. I've added a Galen ref and switched to a more direct statement re Aristotle (I think our summary in Aristotelian corpus is sound but we want a single ref that I've actually read). Promotion surprised me, I hadn't thought of this as a starting point, but the speeches make it work. (Re-reading Thucydides recently, I was particularly struck by how much work those speeches do, and remembered accepting them long ago as reasonable reconstructions.) Now I must read the pseudo-Sallust/pseudo-Cicero paper - that's all new to me! NebY (talk) 01:13, 23 May 2025 (UTC)

Scope of synthesis

Bringing this up here because it's probably relevant for explaining what synthesis is. Do you think something like:

The ancient sources report that the lagoon was drained. Polybius and Appian report that the tide went out, lowering the water level. Livy says a strong wind blew the water out to sea. After the drainage, Scipio's forces attacked over the now-fordable waters, reaching the north side of the city...

Putting aside the fact that many reliable sources essentially shrug and say "eh, both happened" is there an implication that both wind and tide happened? If so, is saying that the two happened together – which is not what is in either Polybius/Appian or Livy, both of which say only their thing happened, – a synthesis? Ifly6 (talk) 04:41, 17 May 2025 (UTC)

@Michael Aurel, Caeciliusinhorto, and Psychastes: correcting template ({{u}} → {{ping}}) to have correct pings. Ifly6 (talk) 04:43, 17 May 2025 (UTC)

Hmm. As written, my first impression of that passage would be that those are competing explanations, so I do think it's probably a misuse of the sources to say both happened, such as The ancient sources report that the lagoon was drained when the tide went out during a strong wind. (because neither source verifies that), but I'm not sure if it's WP:SYNTH specifically unless you said something like the combination of these two together led to an unusually low water level, allowing Scipio's forces to attack over the now-fordable waters which reaches a conclusion implied by neither source. Psychastes (talk) 05:25, 17 May 2025 (UTC)

That's fair, I think, since the prefacing statement indicates that something happened and then gives two separate causes. However, consider this: Polybius and Appian report that the tide in the lagoon then went out; Livy also reports that a strong wind blew its water out to sea. Scipio's forces then attacked over the now-fordable waters, reaching the north side of the city... This feels still like a straightforward presentation of the fact that P/A and L say these things and even that they say them separately. But to me it implies, however, an intersection of P/A and L's narratives which would be synthesis. Ifly6 (talk) 06:06, 17 May 2025 (UTC)

Yeah, I think "Livy also ..." is dubious here unless Livy in fact says "the tide went out and strong winds blew the water out to sea". I think your initial example paragraph is fine on SYNTH grounds (though if two different ancient sources give different potentially contradictory explanations for something, we should certainly not be relying on the primary sources and should instead base our framing on how secondary sources discuss this!) Caeciliusinhorto (talk) 13:02, 18 May 2025 (UTC)

It's probably a bit of an edge case, though I'd tend to lean in the direction of working to clarify passages which dangle close to the SYNTH line. Juxtaposing claims from ancient sources is of course fine in terms of SYNTH when those claims are obviously contradictory (eg., different sources giving different parentages), but in cases where they aren't it probably doesn't hurt to clarify things. Adding a word such as "instead", "alternatively", "whereas", etc., might be enough to do the trick. As to explicitly saying that the two happened together, yes, I would says that's SYNTH. – Michael Aurel (talk) 06:28, 17 May 2025 (UTC)