Documentation as debugging

This may come as no surprise to some readers, but I thought it appropriate to note anyway. During the blog writing about the rhythmic analysis – part 1 , I noticed I would tighten up the definition of terms, and also the actual implementation of the algorithm significantly. I would start writing something, seeing what I had just written and thinking “this just does not make sense” or “this implementation must be off, or just plainly wrong”.  Then, to be able to write sensibly about the ideas, I went back and tidied up my own mess. In the process, making a much more reliable method for analysing the features I wanted to extract. What was surprising was not that this happened but the degree to which it happened.  In the context of artistic research and the reflection embedded in that process, similar events may occur.