source file: mills2.txt Date: Wed, 4 Oct 1995 10:18:47 -0700 From: "John H. Chalmers" From: mclaren Subject: Tuning & Psychoacoustics - Post 11 of 25 --- MYTH: "I KNOW WHAT MY EARS HEAR AND I KNOW WHAT A 'PURE' OCTAVE, A 'PURE' FIFTH, AND A 'PURE' THIRD IS." FACT: Because of the phenomenon of categorical perception, none of us know we actually hear--as opposed to what our ear/brain system brainwashes us into *believing* we hear. The only way to actually *determine* what you're hearing (rather than what you *think* you're hearing) is to use double-blind psychoacoustic tests. Until 1969, such tests were seldom used. Computers were unheard-of; and prior to computer-generated psychoacoustic test tones, the only available readily controllable test tones were those generated by analog circuits whose frequency drifted by significant amounts as the temperature of the sound-generating circuit changed. As a result, ignorance of the ear/brain system's behaviour was near-absolute prior to Max Mathews' creation of the acoustic compiler in 1959. As a result of Mathews' innovation, many surprising properties were discovered in the ear/brain system. One of the most surprising of these properties is known as "categorical perception." --- The phenomenon of categorical perception is familiar to linguists. Everyone pronounces phonemes, vowels, and consonants slightly differently-- and in different regional dialects the sound of a word may be entirely transformed. In New England, "pahk my cah," in the MidWest, "park my car," down south, "purk m' cuhr." The ear/brain system has a learned mechanism for dealing with these differences--different sounds are heard as the same semantic unit. This ear/brain system of learned categorization is known as categorical perception, and it operates so efficiently that people in a given region of the country cannot even "hear" their own accent. As far as they can tell, they're speaking "standard English"--everyone else is slurring or pinching or warping their words "with some strange kind of accent." Categorical perception has been proven to operate in the perception of musical sounds, and it gives rise to many of the same distortions of the auditory system. In the paper "Categorical Perception--Phenomenon or Epiphenomenon: Evidence from experiments in the perception of melodic musical intervals," by E. M. Burns and W. D. Ward [JASA, vol. 63, No. 2, 1978, pp. 456-468], the authors point out: "An experiment on the perception of melodic intervals by musically untrained observers showed no evidence for the existence of "natural" categories for musical intervals." The authors also found that for trained musicians "musical intervals are also rather unique in that musicians are able to perfectly identify more than 30 categories of musical intervals..." These results strongly contradict both the standard 12-tone dogma that only the intervals of the 12-TET scale are recognizable or musically significant; and the standard "natural interval" dogma which holds that some musical intervals are [fill in your own preferred propaganda] "natural," "pure," "preferred," "rational," etc. Among other interesting conclusions, Ward and Burns found that "The average difference limen (based on the 75% correct points from the psychometric functions) for three subjects at the physical octave was 16 cents. The DL's at other ratios in the civicinity of the octave were not significantly different. A DL of 16 cents is in good agreement with the DL estimated from the from the standard deviation of repeated adjustments of sequential octaves (about 10 cents) in the same frequency region found by Ward (1954). (...) As in Moran and Pratt's experiment, large differences were found for DL's at different ratios, but the range of DL's (14-25 cents) was in good agreement with their results." [Burns, E. M. and Ward, W.D., JASA, 63(2), Feb. 1978, pg. 456] This preference for stretched as opposed to purportedly "natural" intervals is not a new discovery. As will be seen in the post after this one, the preference for stretched vertical intervals--and for significantly *wider* melodic than vertical intervals--was discovered by the very first researchers who investigated the operation of the ear/brain system. What are the implications of these particular psychoacoustic data for tuning and music? First, these data explain clearly and convincingly why there are so many different tunings systems and timbres used in the various musics of cultures throughout the world. Because of the influence of categorical perception and the implied importance of learned response on the ear/brain system, any system of pitches can be learned as "preferred" by the ear. Thus a Mongolian Buddhist using the r-gynd-stad tuning can with equal justification claim that the pitches of his musical system enjoy a privileged status in the ear/brain sytem as can Javanese gamelan performer. According to the psychoacoustic results adduced above, both musicians are correct--because the cultures in which their pitch preferences were formed characterize those particular pitches as "special." And because of the known effects of learned response and categorical perception, a wide variety of pitches can equally be perceived as "special" or "uniquely privileged." These psychoacoustic data would also tend to support current Western musical practices, at least to the extent that the Western 12-tone tuning system is acculturated into Wesern musicians and composers, and to which Western performers and audiences perceive departures from those pitches as falling within the range of variability which (as Moran and Pratt point out) characterize all pitches. Categorical perception strongly favors all three classes of tuning-- just intonation, equal temperament and non-just non-equal tunings-- since once the pitches are learned and perceived as "special" or "privileged" both audience and performers strongly tend to perceive departures from those pitches as "ornamental," if indeed the departures are heard as different pitches at all. If these psychoacoustic results are accurate, all tuning systems are self-reinforcing feedback systems, with "errors" heard as slight variations of base pitches (as in the case of Jaipongan, where slendro or pelog are used as base scales for ornamental extra-scalar variations, or as in the various sruti of East Indian practice, where the remaining 22 pitches are used as ornamental extra-modal variational pitches, or as in the vocal inflexions of Sinead O'Conner, Louis Armstrong or Ella Fitzgerlad, all of whom consistently range microtonally outside the 12-tone equal tempered scale in which their song purports to reside). The next post will consider the effects of possible interactions between the various ear/brain processes discussed to date, and the evidence for such interactions, along with the implications for tuning and music. --mclaren Received: from eartha.mills.edu [144.91.3.20] by vbv40.ezh.nl with SMTP-OpenVMS via TCP/IP; Thu, 5 Oct 1995 21:58 +0100 Received: from by eartha.mills.edu via SMTP (940816.SGI.8.6.9/930416.SGI) for id MAA22353; Thu, 5 Oct 1995 12:58:04 -0700 Date: Thu, 5 Oct 1995 12:58:04 -0700 Message-Id: <9510051256.aa03198@cyber.cyber.net> Errors-To: madole@ella.mills.edu Reply-To: tuning@eartha.mills.edu Originator: tuning@eartha.mills.edu Sender: tuning@eartha.mills.edu