I read a lot of books that most readers would consider dry. But this one makes a sponge that has sat in a closet for ten years seem like a source of living water.

The idea is that, by looking at the way a manuscript is organized and presented, one can learn something about the attitudes of the compilers -- and, from that, you can learn something about what they thought about the contents. The first point I grant, with reservations. The second... well, it's possible, but it's going to take a lot more than this book to demonstrate the point.

Author Evans takes fifteen manuscripts which contain multiple Middle English romances, and starts classifying elements -- illustrations, illuminations, various sorts of framing devices, scribal comments, etc. Then he starts cranking out statistics. Then he analyzes the results.

There are a lot of problems here. First, fifteen examples is not a sample. Fifteen examples is a set of anecdotes that can only be confirmed if you have a sample many times that large. The statistics -- which are incredibly dull, as well as being poorly explained (those two facts are probably related) -- have no meaning, because there isn't enough base data to mean anything. Also, the whole thing depends on an arbitrary classification system -- how do you decide what is a "title" and what a "long title"? (Evans has a rule, which is fine, but is there any reason to think that a scribe sat down and said, "Oh, gee, I'm supposed to give this book a Long Title and this title is only three words long; better expand it"? Of course not.)

Having given us all those statistics about layout features, Evans basically throws it all overboard and starts analyzing the way individual writings are organized in books. Here, everything is sui generis -- there are no manuscripts with identical contents, so Evans starts inventing groups of romances for the seeming purpose of making everything an exception to a rule he just invented! This too was too dense for me to come away with anything except a strong sense that Evans has a monomania. And, after all that, his conclusion is just that we should read romances in their manuscript context. I doubt anyone would argue. After all, someone chose to put those pieces together.

The amount of effort that went into this book was clearly immense. There might even be some value in there, somewhere, if it had been made a little more digestible. But I frankly doubt it. I rather suspect that whichever editor passed this book was so desperate for something that sounded scholarly that (s)he took this without having any idea what (s)he was publishing. Yes, it's immensely detailed and carefully illustrated and documented. But so, very often, was a manual of phrenology. ( )