Subject: Re: Soft/loud grouping patterns From: Neil Todd <TODD(at)FS4.PSY.MAN.AC.UK> Date: Wed, 14 Jul 1999 14:46:25 GMTDear Al and List > >> >> >>According to Handel, Fraisse explained this in his 1956 book as an >> >>effect of grouping: >> >>1. the louder events group with following quieter events .... [big snip] >> >> >>Is there a general underlying principle operating here? Why would >> >>louder events group with following quieter events rather than preceding >> >>quieter events? > >Perhaps a bias in favor of grouping sounds in the order "strong-weak" >might be useful in everyday auditory perception. First it is worth noting >that the onset of a sound is usually its most intense part (especially in >the case of impacts, plucked strings, and released-stop syllables such as >/da/ in speech) with weaker continuing vibrations and echoes following. A >bias towards grouping in the order "strong-weak" would tend to ensure that >the onset of a sound (which may be somewhat different from the rest of the >sound because of onset transients and spectral spread) would group with >the remaining parts (including echoes) of the SAME sound rather than with >the tail end of the previous sound. -Syl- I I I -Rhyme- I I | On Nu Coda Surely in the case of syllables the most intense part is usually the syllable nucleus (Nu)? The nucleus, usually a vowel, is also considerably longer. Thus syllable onsets (on) generally form a "weak-strong" relationship with the nucleus. This is an example of 'backward' grouping, certainly in the case of isolated syllables. Again I would argue that a sensory memory model provides a good account of this (e.g. see Todd, N.P.McAngus and Brown, G.J. (1996) Visualization of rhythm, time and metre. Artificial Intelligence Review 10, 253-273. ) Cheers Neil