Analyses of tunings have often been carried out independently of pieces in which they are actually realized. Whereas tunings are prima facie relevant to pieces in which they occur, to what extent is this so? And does such a relationship hold in both directions? That is, are analyses of pieces relevant to analyses of their tunings? Both sorts of analysis involve methodological problems and, at least in principle, both sorts of analysis should mesh. Germane to the present discussion are instances

more »

... of such analytical problems that arise in Central Javanese pélog tunings and in 'skeletal melodies' (balungans) of multi-section pieces (gendhings) that employ these tunings. The present account identifies such problems and proposes solutions that attempt to coordinate both sorts of analysis. With regard to tuning per se, relationships among acoustical spectra, pitch determinacy, interval categorization, and 'errant tones' are considered. Concerning individual pieces, both jointly and severally, longstanding notions about 'exchange,' 'alternate,' or 'substitute' tones (sorogan), modal identity (pathet), and cadential (gong) tones are addressed. Linking both kinds of analysis—and shared by both—is an amplified formulation of Wertheimer's Gestalt Grouping Principle of Similarity. Introduced from post-tonal analysis of European-derived music are concepts of common tones, 'well-formed' (WF) scales, and interval vectors.