I understand why some complain about Ramsay MacMullen’s writing style, which can be dense and convoluted. But at his finest he’s something like a great impressionist painter, or even a pointillist, daubing multiple points of color on his canvas, out of which a vivid yet maddeningly indistinct atmosphere emerges.
More importantly, his style also reflects, I think, his perception of the historian’s craft, which is not to simplify or create grand patterns, but rather just the opposite: to try to recreate the myriad complexities and contradictions of long-dead societies that are now hidden from us by bland marble statues, scattered facts and unsupported assumptions.
Two examples may suffice, both from Prof. MacMullen’s fine Paganism in the Roman Empire, to illustrate his insistence that an honest attempt to capture the past must avoid the impulse to generalize. The first follows a review of specific examples testing "the vitality of paganism":
This is the kind of mixed bag of facts that renders general statements about religious life so hard to frame or so easily criticized if they are framed too narrowly. What, for instance, can one make of the assertion that oracles, “it is true, enjoyed a recovery in popularity in the second century” [quoting from a 1929 work by Belgian archaeologist and historian Franz Cumont] – for which a single inscription is cited, recording help sought by a city in Sardinia from Apollo in Claros? Such characterizing of the feelings and thoughts of fifty million people on any day out of thirty-six thousand has something ludicrous about it, as if one were to measure the pulse of the western world on the basis of a single headline in the St. Albans Sentinel. Worse than that, perhaps: since religious feelings are not something to talk about in public, in some of their aspects, they must prove all the harder to assess from the outside. The more need for care.
Another quote of poor M. Cumont – “In the third century, the misery of the times was the cause of such great suffering . . . that people sought asylum in the expectation of a better life” – draws a second outburst. After questioning the evidence supporting the assertion, Prof. MacMullen expresses broader misgivings about whether it is ever possible, or desirable, to make such broad characterizations:
[W]hat sense does it make to assign a single character to so long an era? – as if one were to say, “in Italy, Switzerland, the Low Countries, Britain, France, and Spain between about 1400 and 1600, people were tense and worried.” The statement denies the very change and complexity which it is the job of historians to discover . . .. As if a century, let alone two or three, could be “an age” – that is, a stretch of time in which just about everybody acted in significantly different ways from other human beings before and after, but with something close to a characteristic uniformity among themselves!
The human tendency to arrange facts into patterns and the paucity of facts available concerning the Classical world – certainly in comparison to those available to a historian of, say, the Twentieth Century – suggest that the ancient historian must be particularly wary:
It can only be the observer’s ignorance that would make all the life of a vast area for so long a span, in the mind’s eye, blur, shrink, stop. Such ignorance is the natural condition of the ancient historian, paradoxically inviting him to arrange his few scattered facts into grand patters – the fewer, the grander. Where the temptations and the hazards are so pressing, perhaps it should be a rule that no one may generalize about ancient history until he has served an old-fashioned, seven-year apprenticeship in the teaching, or at least the formal study, of modern history!