This is two python programs I wrote in connection with a presentation at the Yale Linguistics department -- improvements on a wordbreaker program I had written over the years, based on a simplified version of Carl de Marcken's ideas. The analyze program was new, and probably important -- it gives us insight into what the heck is actionally going on. I'd like to use these programs to show linguists what this kind of learning suggests. But what are we learning about? It's not literally what people do, but it shows a sort of proof of concept of a dynamic approach to learning, or theorizing about learning.