source file: mills2.txt Date: Fri, 7 Jun 1996 16:41:00 -0700 Subject: Post from McLaren From: John Chalmers From: mclaren Subject: Several more worthwhile papers -- The journal "Music Perception" , Winter 1984, 2(2), pp. 131-165, carried the article "Tonal Schemata in the Perception of Music in Bali and in the West," It's an interesting paper, but not nearly as useful as you'd imagine from the title--alas. The authors confined themselves to extremely low-level questions as: How well can Balinese listeners identify the place of a missing western pitch in a western scale as opposed to how well a western listenre can identify the place of a missing gamelan pitch in a gamelan scale? The Balinese audience memebrs tested were reportedly members of a remote village with no previous exposure to western music. This sounds unlikely on the face of it, and one is inclined to question the study on that basis. Nonetheless it's worth reading. >From Music Perception, Winter 1984 , 2 (2), pp. 245-264, the article "Studies in Musical Cognition: Comments from a Music Theorist" offer some enlightening criticisms of current higher-level experiments designed to elucidate elements of human melodic and harmonic audition. "A great many of the difficulties in psychology experiments have come from an ignorance of or naivete wth regard to tonality. Four examples will be given. "First, length was assumed until recently to be an important factor in both the `memorable-ness' and the `tonal-ness' of tunes; longer tunes carried more information than short ones. This has been shown to be a false assumption." You'd think this was obvious. Yet it seems to have come as a revelation to music cognition researchers. C'mon, guys! Obviously Mozart's 41st symphony 1st movement theme, while *long*, is *easy* to remember, whereas the much *shorter* theme of the first movement of Webern's symphony is obviously vastly *harder* to remember. Has anyone heard of the words "implied harmonic cues" and "tonal landmarks"? Jeez. "Second, so called `tonal' tunes have thus far for the most part been defined simply as those consisting of pitches drawn from some major- mode diatonic collection, whereas `atonal' tunes are obtained by drawing random pitches from the 12-tone chromatic collection. The notions that tunes could be diatonic but only weakly tonal or chromatic but also tonal have only begun to be articulated and refined." Yow. You'd think this would be obvious. Listening--that's the key, folks. Ya gotta *listen* to those psychoacoustic test melodies before you use 'em... "Third, in music the relationship between frequency (pitch) distance and function distance is a complex one..." Again, a very clearly obvious point--pitches a 12-TET semitone apart represent opposing harmonies even though the notes are melodically close, while pitches 7 12-TET semitones apart represent harmonies close together even though the notes are melodically distant. Duh. Yet researchers are only now catching onto this...? "Fourth, tunes in the psychological literature, whether tonal or atonal, are still all too often characterized as sequences of signed note-to-note intervals. Few notions of melodic structure (certainly none based on a well-founded theory of implicit harmonic structure) have emerged." [Hantz, E., "Studies in Musical Cognition: Comments from a Music Theorist," Music Perception, Winter 1984, 2(2), pp. 245-264.] These objections seem to explain very clearly why much of the current psychoacoustic and psychological and music cognition research in microtonality has failed so miserably to produce results as revelatory or enlightening as simply LISTENING TO NON-12 MUSIC. In study after study, otherwise reliable reserachers like Carol Krumhansl come to obviously wacky conclusions about such tunings as 48-TET, 19-TET, 36-TET, etc. The reason for these obviously wrong concusions is that the researchers assume [1] that melodies in other tunings must exhibit the same melodic structure as 12-TET; and [2] melodies in other tunings msut exhibit the same implied harmonic structure as 12-TET, with identical modes, etc. etc. Ivor Darreg pointed out long ago, as did Easley Blackwood, that other tunings often turn the rules of western harmony *upside-down.* Some tunings require that major thirds be treated as unstable dissonances which resolve DOWN until major seconds (17-TET) or UP into perfect fourths (Pythagorean ji). Some tunings require that the fifth be treated as a dissonance (13-TET, 18-TET, 23-TET, many of Erv Wilson's higher-order CPSs), while other tunings require that the fifth be treated as a consonance (22-TET, standard-limit ji arrays). In some tunings root progression by fifths produce the most powerful and convincing cadences (19-TET, 7-limit ji) while in other tunings root progressions by other intervals produce the most powerful and convincing cadences (some of Erv Wilson's higher- level CPSs, 15-TET, varius non-just non-equal tunings like the metal tube scale or the free-free metal bar scale). This explains why Krumhansl and other respected researchers keep "discovering" that 12-TET is "the best tuning" for melodies. When you measure everything by a 12-TET yardstick, 12 looks best. You'd think that none of this would be controversial or astonishing, yet these ideas seem to wallop the psychoacoustics community with the force of divine revelation. These folks are only now slowwwwwwwwwwwwwwwwwwwly awakening to the idea that you can't usefully measure non-12 melodic or harmonic properties according to 12-TET melodic and harmonic standards. Wow. What a surprise, eh? Incidentally, these same considerations explain why Markov-chain analysis produces junk when you use it to analyze a piece of music and then compose an algorithmic piece. Markov chains are blind to the the melodic contour, the implied harmonies, the sense of balance and unbalance, of parry and riposte in a melody. In this regard Larry Polansky's morphological metrics represent a major advance--so, naturally, no psychoacoustical researcher has yet employed Larry's breakthrough concept in any tests of microtonal melodic perception. --mclaren Received: from eartha.mills.edu [144.91.3.20] by vbv40.ezh.nl with SMTP-OpenVMS via TCP/IP; Sat, 8 Jun 1996 19:50 +0100 Received: from by eartha.mills.edu via SMTP (940816.SGI.8.6.9/930416.SGI) for id KAA11435; Sat, 8 Jun 1996 10:50:50 -0700 Date: Sat, 8 Jun 1996 10:50:50 -0700 Message-Id: <960608134834_322162012@emout13.mail.aol.com> Errors-To: madole@ella.mills.edu Reply-To: tuning@eartha.mills.edu Originator: tuning@eartha.mills.edu Sender: tuning@eartha.mills.edu