Primary Sources

[Nathan Ross]

A source with direct knowledge of an event is not necessarily "better" than one without, but its usually easier to assess its perspective and limits.

Yes. And I think the 'perspectives and limits' are an important aspect of this. The Vindolanda tablets come to us (as far as I know) in as close to a raw state as any ancient texts get - they're also highly opaque and often completely obscure. They are a primary source, but without secondary sources (including ancient ones) to give them context, they have only curiosity value.

[Sean Manning]

it seems to me that arguing about definitions is not helpful.

I don't think it's a question of arguing about definitions. It's a question of knowing the conventions, so that, if someone speaks of primary or secondary sources, we know what they're talking about.

Since I started my graduate education, I have come to doubt that there is a convention! I have found educated people who say "primary source" when I would say "ancient source" and other people who say "primary source" when I would say "main source." The primary source/secondary source dichotomy may have become so abused that it is best to avoid the terms, just like "western/eastern." (Here is someone with a PhD in 16th century English history who finds it confusing). I can only say that I thought that the original definition of "primary source," and the only definition which has no other short name, was "a source with direct knowledge" or "a contemporary source."

David Schaps' "Handbook for Classical Research" pp. 39 and 40 has some thoughtful but revealing comments; "primary sources are those with direct knowledge of the event or thing under discussion ... since, however, vast amounts of ancient literature have been lost, we usually treat as primary any source beyond which we cannot go, because its own sources are inaccessible to us ... what source is primary and what secondary also depends on the problem with which we are dealing." But of course when we study the fourth century BCE, having a Livy is fundamentally different from having the chronicle of one of the big temples in Rome, and calling him "as close to a primary source as we can get" does not help. If we want real sources from fourth-century-BCE Italy, we must work with coins and tombs and inscriptions even though they won't give us a tidy narrative.

[Nathan Ross]

The primary source/secondary source dichotomy may have become so abused that it is best to avoid the terms

I can see the reasoning, but wouldn't that risk losing the sense of the terms as well? Which, as you've said, is quite important and useful in helping us to better understand what we're looking at?

"Handbook for Classical Research"... we usually treat as primary any source beyond which we cannot go, because its own sources are inaccessible to us ... what source is primary and what secondary also depends on the problem with which we are dealing."

Thanks for the quote - that does cover one interpretation very neatly (and quite a common one, going by some of the responses here).

But I'm still not sure what's gained by treating earliest available sources as primary sources by default. There's still an important difference between a genuine primary source document and the work of an historian writing much later - surely calling Livy an ancient secondary source is both more accurate and preserves the sense of the term 'primary' in this context?

even though they won't give us a tidy narrative.

I wonder if 'tidy narratives' are the exclusive preserve of secondary literature!

[Sean Manning]

Thanks for the quote - that does cover one interpretation very neatly (and quite a common one, going by some of the responses here).

But I'm still not sure what's gained by treating earliest available sources as primary sources by default. There's still an important difference between a genuine primary source document and the work of an historian writing much later - surely calling Livy an ancient secondary source is both more accurate and preserves the sense of the term 'primary' in this context?

Well, for what its worth, I agree with you. But I have to chose my battles, and right now I would rather put my energy into arguing that we need to find a new way to think about the Achaemenid army, and maybe that a rule of thumb from the nineteenth century is not a universal law for all combat. I don't know if I will convince anyone of either, but I think that my chances are a bit better.

I wonder if 'tidy narratives' are the exclusive preserve of secondary literature!

Probably!

[Justin I]

This is just my standard, but for the Second Punic War I would consider Livy a secondary source. He may not have been a historian by modern standards, but he was a historian nonetheless, and his account was based on other sources, some primary and some secondary.
I would consider Quintus Fabius Pictor, Silenos of Kaleakte, Sosylus of Sparta, et al the primary sources for that war, along with any inscriptions or documents preserved from the time.
I think of Polybius as a secondary source for the Second Punic War too, because he was born in 200 B.C., one year after that war ended. That would be like me -- born in the 1980s -- writing a book about the Vietnam War. I would be a secondary source. But Polybius is a primary source for lots of other events.
I guess my point is that for some of these ancient events, we don't necessarily have primary sources. Obviously, my view is completely at odds with another expressed earlier in this thread, that "primary" or "secondary" depends one where our source is in the chain of research.
But then again, I'm a journalist and not a historian. I guess it comes down to who's defining it.