This post is probably best seen as a remix of some of the arguments in the last two posts, prompted by some of the very useful comments Ive had on those. One of Chris arguments was that archaeologists were now in a position to start producing their own grand narratives rather than simply relying on those of historians. Which prompts the question: what makes a good grand narrative?
In my view, a good/useful grand narrative (in the sense of an outline story of what happened in a period) has three characteristics. One, obviously is that its grand: it covers a wide area and/or a long period of time. Two, is that its memorable: these grand narratives are intended to provide a framework into which you can organise other material you read/see. And thirdly, following on from that, you need to be able to attach new ideas onto it; it needs to have metaphorical hook points.
How do these factors influence the choice of document-based or archaeological-based grand narratives? (Im assuming here cases where there is a reasonable amount of both archaeological and textual evidence available: if youve only really got one kind of evidence, youre stuck with it). On grandness, an archaeological synthesis can potentially cover a wider geographical range, at least until the twelfth century, if not beyond: you cant do the textual equivalent of what Chris does in Framing the Middle Ages and look across the whole post-Roman Mediterranean, because the sources either arent there or are of such disparate genres as to be impossible to synthesise.
My instinct, however, after hearing Chris version of an archaeological grand narrative of Palestine and Syria was to say automatically that this was less memorable than a historical one. But Phils comments on how he needed to visualise an area before reading its history effectively got me re-thinking this. How much does the memorability of a grand narrative depend both on ones previous knowledge and also on what your own memory stores more easily? I have a relatively good memory for dates, but my visual memory and spatial sense is much less good: I find it hard to remember where places are on a map even when Ive seen it a number of times. I suspect for some archaeologists the reverse may be true. If you have acquired a feel for the difference between an early Saxon and a mid-Saxon settlement and you have images of them in your minds eye, an archaeological grand narrative may be just as effective, or more so, than an historical one.
Where I think archaeology is more likely to fall down is in providing a narrative framework into which other events can be inserted. The big problem here is accuracy of dating. For a historian, any source you can only date to within a century or two is almost more trouble than its worth; you just cant connect it definitely to any other sources. (This doesnt just apply to archaeological evidence, as Beowulf shows). And as far as I know, the number of periods when you can get relatively hard dates for archaeology isnt that many.
All this means that I dont think that historians are going to be enthusiastic about archaeologically based grand narratives in most circumstances. Chris (and Richard Hodges before him) show the possibility of using them very profitably for the Fall of the Roman Empire problem, where you do have widespread and very well-dated archaeology telling a relatively coherent story. And while archaeology wont solve the Year 1000/Feudal Whatever problem, if we did get a more definite answer to when the economic and demographic takeoff of Western Europe took place, that would be a major advance. But I cant think of many other occasions when I personally would want to swap my text-based chronology for archaeological typologies.