Dear Brian, There is no single natural "best" way to compare songs at any structural level, and you don't want to rely on a single measure of similarity in any case. Further, by specifying the attribute(s) of song you want to compare, rather than seeking to "compare songs", you will arrive at a bunch of candidate techniques more naturally I think. Even for the example you gave, one could use many varied and different measures of similarity (proximity, correlation, asssociation...) based on number of elements, number of element-types, number of shared transitions, etc., or even on measures of variation including information-theoretic measures. In his recent book, Don Kroodsma explains some of these approaches beautifully through examples: Kroodsma, D.E. (2005) The Singing Life of Birds. Houghton Mifflin.
At a lower structural level, you could derive multiple similarity measures derived from characters (quantitative, qualitative...) on single identifiable element-types (e.g., A, J). Parenthetically, in the latter case it wouldn't matter whether the selected element-types appeared in all songs or were even sung by all individuals within a population, as you are comparing populations. In fact, estimation of similarity based on ubiquitous vs. uncommon element-types makes for interesting questions by itself (as would comparisons of similarities in variation or across multiple scales of structural organization). One interesting publication on the topic of scale in communication is by Dave Bain: Bain, D.E. (1992) Multi-scale communication by vertebrates. Pp. 601-629 in J.A. Thomas et al. (eds.), Marine Mammal Sensory Systems. Plenum. Sincerely, Ted Miller -- Dr. Edward H. Miller, Associate Professor Biology Department Memorial University of Newfoundland St. John's NL A1B 3X9 CANADA (phone 709-737-4563; fax 709-737-3018) Quoting "Brian R. Mitchell" <[EMAIL PROTECTED]>: > I'm hoping someone on this list can point me to a technique for > analyzing similarity between bird songs (within a species). Basically, > we have multiple songs recorded from about 20 bobolinks, and we want to > determine if songs recorded from individuals within a local area are > more similar than songs from different areas. Bobolink songs are > complex, but are a sequence of phrases that are put together in > different ways by different individuals. > > My question is whether there is a way to analyze a series of characters > for frequency and placement. For instance we may have the following > three songs, where letters represent phrases: > > 1 AGHEIGHDDS > 2 AGEGDSSHIIJJJ > 3 CTQEUYSEKE > > Songs 1 and 2 share more phrases and start with the same phrases, so I > would classify them as being more similar. Does anyone know of an > analytical technique for dealing with this sort of data? I'm thinking > that the data is probably similar to some types of genetic data, where > letters would refer to alleles, and I'm also wondering if anyone has > applied genetic analysis methods to this type of data. > > Thanks! > > Brian > > ---------------------------------------------------------- > Brian R. Mitchell > Post-Doctoral Research Associate > University of Vermont > The Rubenstein School of Environment and Natural Resources > 81 Carrigan Drive > Burlington, VT 05405-0088 > (802) 656-2496 > [EMAIL PROTECTED] > ---------------------------------------------------------- > >
