|Perl: the Markov chain saw|
That looks like it might be very interesting--I like playing with parsing, so I typically read all nodes that talk about it. Unfortunately, I'm not conversant with functional programming, nor familiar with the Hutton paper. So I can't really dig into it unless I'm willing to do a bit of digging.
If you want people to take a closer look at it you might need to add a couple of links to your post to the relevent background information. A sentence or two describing why and how we would use such a thing would be even better.
Having said that, I find the first block of code to be mostly clear and easy to read. If I were going to try to work on it, it doesn't look like it would be too difficult to maintain. There are a few odd variable names, but the way you're using other variable names, I'm guessing that they're relatively obvious contractions to someone familiar with the problem domain.
The only criticisms I can offer at this time are:
That's the best I can do. Without enough background, I'm afraid the second code block is rather difficult for me to read. I can decipher bits of the syntax, but I can't see why or how you're tying it all together.
I hope you find this somewhat helpful.
When your only tool is a hammer, all problems look like your thumb.