节约地使用音乐动机？
读勋伯格的《作曲基本原理》会产生这样的印象：动机（motive）总是或者应该是音乐发展的最小单元，整首音乐作品可以分解到很有限的几个音乐动机，通过变异和组合这些音乐动机来生成更大的音乐结构，乃至整首作品。勋伯格特别推崇的作品中，音乐动机的经济使用往往是一个很明显的特点。
但是当我具体分析音乐作品的时候，情况并不总是这么“干净”，总是会有“不经济”的音乐材料出现。
然后我读到了 William E. Caplin 的《Analyzing Classical Form》，他的理论要朴实很多，远没有勋伯格的那么“强硬”。他会明确指出某些音乐作品中不经济的音乐材料。（顺便一提，这本书是 textbook 的典范，用户体验非常友好，理论很朴实，或者说没有不必要的抽象。可是豆瓣上读的人很少。）
我们可以用“变异-组合”的视角来思考音乐生成的过程：小的音乐结构被变异，与原有的结构组合，生成更大的结构，直到生成整首作品。比如乐段（period）的后句（consequent）可以看成是前句（antecedent）的变异，两者组合成乐段；乐句（sentence）中 continuation phrase 中的 fragment 是 presentation phrase 中动机的变异；更大的尺度上，某主题的复现是该主题的变异，两者加上对比段构成整首音乐。
但是在这样的视角中，被变异和组合的最小结构并不一定是动机，可能是 phrase，或者更大的结构。同时，也不是所有的材料都来源于原有材料的变异，可以有新的材料。
以上讨论的背后是一些更深层次的问题：经济地使用音乐材料有什么样的美学意义？
我的理解中，勋伯格有点像寻求理论公理化的数学家。如果是这样的话，其它简约的音乐生成机制也有可能吸引他。而这背后其实是对认知简约（cognitive simplicity）的追求。认知简约让人很舒适，但也有误导人的危险，因为它并不总意味着“emotional simplicity”。
关于经济地使用动机的问题，我也询问了一位作曲背景的网友，他的回答更有料：
You have to be careful with Schoenberg. Despite it's title, _Fundamentals_ is best read when you've already got a whack of music theory behind you. His intellectual rigour can be very off-putting. What's more, he makes motivic development prescriptive--the _sine qua non_ of "great" compositions--which, when one looks at the whole of the classical repertoire, just isn't the case. He cherry picks his composers and examples. The same is true of Heinrich Schenker, a much-lauded musicologist who developed a unique system of music analysis in the 19th century. Be leery of his work until you're well-versed in conventional theory.
Lastly, bear in mind that Messiaen, in his mature works, deliberately avoids all notion of motivic development. Which means that from a Schoenbergian perspective, _Des canyons aux étoiles_ and _Vingt regards_ aren't worthy of serious attention.
简单地总结：勋伯格的理论有明显的个人倾向。有的作曲家，比如梅西安，并不走这个路线。
而这位网友个人比较喜欢使用动机发展（motivic development）来作曲：
I am steeped in the notion of motivic development. To me, there doesn't seem to be much point in writing a work of any length that doesn't involve thorough development of the ideas. It's why I compose. I want to turn things over, look at them backwards, speed them up, slow them down, reharmonize them, transform them, combine them contrapuntally... you get the picture. Sometimes, I like to begin with a tiny motif, just a few notes, and see how far I can stretch it (like the piano sonata). Sometimes, I like to take a long melody and pick it apart, motif by motif, exploring them, exhausting them.
That said, I don't believe for a moment that it's the only way to go at composing, and I certainly don't do it all the time. If I'm honest, the reason for favouring strong motivic development is that I'm lazy. I tend to think of motifs as "note generators", i.e. when I don't know exactly what to do next, I look at ways of developing existing material, which provides me with notes to work with AND--very important--a virtual guarantee that the "new" material coheres with what's gone before.
