Among the many arguments I gave for this conclusion was one in particular about the history of history itself:
Graves' scholarship is obsolete, having been vastly improved upon by new methods, materials, discoveries, and textual criticism in the century since he worked. In fact, almost every historical work written before 1950 is regarded as outdated and untrustworthy by historians today.I was subsequently asked in feedback what I meant by that. Not, that is, in reference just to Graves, but in apparently condemning the whole field of history even up to the middle of the 20th century. As I have made the same point in many other contexts, I gave a detailed reply to this question years ago. I now realize this is well worth publishing here, since it applies far beyond the case of Graves and relates a point I will continue to make again and again.
So here it is, with some minor editing:
The first day I arrived in the office of my graduate advisor at Columbia University, Professor William V. Harris, a very distinguished scholar of ancient history, one of the first things he said to me is (paraphrasing, since I can't recall his exact words--this was now about ten years ago), "Don't rely on anything written before 1950 or so unless you can confirm what it says from primary evidence or more recent scholarship." Point blank.
I have since found that his advice was quite apt. That doesn't mean we don't consult such texts (many crucial references were produced in the 1920's and 1940's that have never been revised) but generally we only use them as a "guide" back to the primary evidence or to check against later scholarship, etc. Hence the biggest exceptions are works that do little but present primary evidence (e.g. collections of inscriptions, critical textual apparati, etc.). Though there are a few exceptions in historical scholarship--but very few. For example, Ronald Syme's Roman Revolution (1939) is practically required reading on the Roman Civil Wars. Yet even then no historian would ever use Syme as a source without backing up whatever claim he is relying on with primary evidence or more recent scholarship. As there has been a lot challenging even Syme on various points. And again, Syme is exceptional.
Generally, the reasons for our attitude toward such early historical work are many, but here are four major ones:
- Historians were often, for some reason, more textually naive before the 1950's, trusting what historical texts and primary sources said too much, and trying too hard to make the evidence fit them. The situation has reversed since then, and archaeology is now more important, and multiple corroboration, and other methodological approaches are required (e.g. showing how a claim fits general cultural knowledge, and conceding uncertainty more often than previous historians did, etc.). As a result, a lot of what was argued before 1950 has been refuted or heavily qualified or modified. So you have to check and see in any given case if a claim still stands.
- A lot more evidence has come to light. For example: new Arabic texts relating to the history of science; papyrological finds pertaining to the Hellenistic period in Greece, the origins of Christianity, Roman history and economics, Egyptian government and society under Greek and Roman rule, etc.; plus documents recovered from Pompeii and Herculaneum and the Dead Sea and Nag Hammadi, and so on (heck, we're still recovering texts from Oxyrhynchus); etc. This evidence has often changed, sometimes radically, the findings of earlier decades. Since you cannot know in advance what has been revised in light of new evidence, you simply can't assume old works stand as written, and must check more recent work to confirm any conclusions.
- Social and cultural history were largely (though not completely) neglected before the 1950's, and when addressed, were approached with the less sophisticated tools of the time. Since then significant advances have been made in sociology, psychology, economics, and anthropology, which have changed the way we understand and study other cultures. This has made a significant impact on the study of the history of religions, of ancient economics, social relations and interactions, background assumptions and worldview studies, and so on. And since these things connect with and affect every historical event in some way, improvements in our understanding of culture and its various facets entail changes in the way we understand and interpret historical texts and events.
- Finally (though this list is not comprehensive) the methodology of historians has become more scientific after WWII. That is, historians have become more method-conscious, and more concerned about distinguishing opinion from fact, and causal theory from chronological sequence, and much more concerned with thorough documentation, relying as much as possible on primary evidence, and being very critical when forced to rely on scholarship instead. Citation of sources is more meticulous. For example, before the 1950's you will find a lot of historians making claims to fact that are really the opinions or theories of earlier historians--and often they won't even tell you that.
A really good example of these factors can be gleaned from reading my dissertation advisor's rather famous book (in our field, that is): William Harris, War and Imperialism in Republican Rome: 327-70 BC (1979, reprinted with a new preface in 1991). He basically shows how earlier historians were hugely wrong on this subject--which is really something, since political and military history was the major thing before the 1950's, so you would think that if they could get anything right, it would be that. But no--and precisely because the religious and cultural and economic contexts, for example, were not properly understood or properly taken into account.
Of course, Harris also refutes a lot of historians from the 60's and 70's, too, but how he does so is also what distinguishes this work (and most works of professional history today) from most pre-50's material: enormously copious (and meticulous) citations, references, and reliance on primary evidence, a careful distinction between fact and theory, and taking into account new discoveries and scholarship, especially (but not only) in the ancillary fields that study human nature and the nature of cultures and societies (psychology, sociology, economics, anthropology, etc.). Harris deploys the same superior methodology (and again refutes common conclusions of earlier eras) in his equally-definitive works Ancient Literacy (1989) and Restraining Rage: The Ideology of Anger Control in Classical Antiquity (2004).
And yet I still recall in high school, in the late 80's, being taught a theory of Roman imperialism that Harris had so soundly refuted in 1979 that his findings became the gold standard and remain unchallenged. And that wasn't the only thing I was taught in high school history classes that I later found out was not only false, but soundly refuted, by work well-predating my high school years. It seems that high school textbooks, and teachers, are still relying on obsolete historical scholarship. And that's a problem. But the solution is caution. Always double-check a claim or conclusion against more recent scholarship.