Here I am a week later, contemplating the use of statistical methods (some of which, such as regression, I am aware that I never even read about for this class, but may be useful) to assess the influence of Haydn and Mozart on later composers (Beethoven, Schubert perhaps). My plan involves collecting data from piano sonata slow movements (excluding theme-and-variations) such as relative key, melodic contour of the opening phrase, length, and tempo marking to compare "apples to apples" from Mozart's work to Beethoven's.
Perhaps this is the kind of academic minutia which real scholars dismiss as bean-counting, and I might be better off writing a paper or a piece of my own. I certainly do not know enough about statistics yet to use multi-variate analysis or any of the techniques I have read about recently in books on politics (Why Iowa?: How Caucuses and Sequential Elections Improve the Presidential Nominating Process
) or broad social science (Natural Experiments of History
- excellent book!). But down the road, my little database might provide me with an interesting sandbox for experiments!
