I’m in the final stages of corrections and updates to Terry Crowley’s Introduction to Historical Linguistics. It’ll be the fourth edition of this book, which first appeared in 1981. That was the year I started preschool. The first Oxford University press edition was 1992, and the third edition was reprinted three times.
This edition has quite a few changes. I have overhauled to the reading list at the end of each chapter, I have added some new data sets, and added new sections and expanded existing ones on morphological and syntactic change. I have largely kept the focus on Austronesian and Australian languages while also taking examples from languages elsewhere in the world. This was the textbook I used when I was an undergrad first starting on historical linguistics so I have a soft spot for it, but some parts were rather dated.
I did not think this would be a particularly easy job, but it has turned out to have been difficult in unexpected ways. I thought the most difficult part would be to write the new sections. Historical linguistics can be a technical field and students who are not familiar with thinking about things from the diachronic point of view often find it very confusing.
That turned out to be the easy part. I just pretend I’m lecturing, I remove the hesitations and repetitions, and Bob is your proverbial uncle. This turns out to be especially easy with voice recognition.
The difficult parts have been in two areas. The first is that Terry’s and my writing styles are rather different from one another. He uses much more clefting than I do, more modifiers, and his way of introducing an argument is a little different. Therefore there has been a bit of a temptation to rewrite, which I shouldn’t do without a good reason. After all, this is a new edition of an existing book and not a knock off new one. I think I really underestimated how much of an issue this would be when the writing by the coauthors isn’t happening simultaneously.
That brings me to the second issue, changing content. There are some things in this book that I fundamentally disagree with. Some of them are pretty minor, but others are pretty important to linguistics at a basic level. One of the minor things is the discrepancy between the data I can find on some of the languages quoted and what’s in the book. It’s not always possible to go back to the original source that Terry used (since the source seems to have been his brain, and it was a very good brain). There are more serious things. For example, I have argued in several places that areal linguistics does not constitute the failure of the comparative method. On the contrary: the comparative method allows us to identify areal phenomena and to study them. I also disagree about the place of the family tree model as a European construct. I see it as essentially a representational question, with a bunch of languages which fall outside that type of representation. But saying that creoles can’t be modelled on a family tree says nothing more than that: that some representations are appropriate for some types of descent relationship. It doesn’t say anything about the politics of these languages, of the worth of studying them, or about the appropriateness of family trees for languages which appear to be modelled well by them.
I suspect that this view is rather more recent than the first edition of this book. There may well have been reasons to argue over the family tree model in the way it is presented here. It may be an accident of the presentation. Terry was an extremely smart guy with an incredible intuition. Not being able to discuss the changes has turned out to cause more angst on my part than any of the other issues. Which is ironic, because if I could discuss them with him I would not be preparing this new edition at all.