Now that you have worked through several extracts from Mozart's sonatas in quite a lot of detail, I hope that you agree that making reductions enables you to understand better the way that the harmony works.
There are two questions that you might naturally have been asking yourself. First, what is the point of making reductions of Mozart's music when the original is so much more beautiful and satisfying than the simple structure with which we end up? And second, did Mozart himself really think of music as composed of a series of levels in this way?
The answer to the first question should be clear by now. We are not interested in making reductions simply to make Mozart's wonderful harmony look simpler. Rather, we are trying to show how the surface of Mozart's music is generated by a series of elaborations of a deeper-level linear logic. The process is somewhat akin to grammatical analysis of a sentence.
The answer to the second question is rather more complex. It is almost certain that Mozart would not literally have worked in this way, conceiving of the ‘skeleton’ first and then elaborating on it in a series of steps. But it is equally certain that outlines such as the one shown in Example 17a were a standard feature of the eighteenth-century style, and indeed of eighteenth-century textbooks. Mozart was certainly familiar with the processes of elaboration and variation that underpin this type of analysis. A certain progression would have sounded satisfying to Mozart just as it does to us, because he had internalised the principles of the underlying processes that give it coherence.
You have now covered most of the basic ideas behind voice-leading analysis. Chief among these is the notion that Mozart's style (like all tonal styles from Bach to Brahms) is rooted in a contrapuntal relationship between treble and bass, where surface gestures are controlled by large-scale linear movement at a deeper level. When this overall linear logic is absent (as in my pastiche, Example 2, with which we started) the music sounds unconvincing and out of style. Developing from this idea is the concept of a system of hierarchical levels, similar to the structures of grammar, where a simple deep-level line may be elaborated rather like a series of ‘variations’ to produce the musical surface. Each level is transformed into the next by a process of elaboration (sometimes called diminution), using four main types of transformation: passing note, neighbour note, suspension and arpeggiation.
It is worth pointing out that this type of analysis has produced mixed reactions among writers on music. Some have embraced the system with almost evangelical zeal, others have dismissed the entire theory as pointless or as unmusical. Obviously, I believe that there are great insights into the workings of harmony that one can express only by using this kind of approach. You may well come across opinions much more critical of analysis in general and voice-leading analysis in particular elsewhere.
As a checklist of the things that voice-leading analysis can achieve which other forms of writing about music cannot, I suggest the following.
Voice-leading analysis enables us to understand at least one aspect of our aesthetic response to music we consider ‘well-written’. In other words, it can explain why one passage sounds ‘logical’ and another ‘illogical’.
Voice-leading analysis allows us to establish some of the workings of Mozart's musical syntax. It gives us insights into his style which we could not achieve by other sorts of analysis. For instance, we saw with Examples 1, 2 and 3 at the beginning of the course how simple chordal analysis alone cannot explain what makes Mozart sound like Mozart.
Simple reductions of the musical surface can sometimes reveal hidden motivic connections (for instance, the replication of motive X in Example 14).
However, it has to be admitted that the analytical method will work equally well, and give almost identical results, with the music of other composers contemporary with Mozart. Because the shapes at the deepest level of analysis are so generalised, and because the rules of counterpoint and diminution are general procedures of the time, the method cannot easily distinguish between the styles of, say, Mozart and Haydn. Defining a composer's individual style is a thorny issue; voice-leading analysis can only ever be one small part of a larger study of the Viennese Classical style. However, I hope that you have found that working through this analytical course has complemented and deepened your musical understanding and that getting to grips with analysis remains rewarding as you continue to learn about and appreciate music.
Once you have completed this course, you can move on to the second of the three courses on harmonic analysis, Voice-leading analysis of music 2: the middleground .