Music, Language and Computation Aline Honingh Guest lecture in - - PowerPoint PPT Presentation
Music, Language and Computation Aline Honingh Guest lecture in Logic, Language and Computation Music at the ILLC l Henkjan Honing music cognition l Fleur Bouwer l Gabor Haden l Rens Bod computational musicology l Aline
Guest lecture in Logic, Language and Computation
l Henkjan Honing – music cognition
l Fleur Bouwer l Gabor Haden
l Rens Bod – computational musicology
l Aline Honingh
l Monthly seminar/discussion group on music
l Music and Language l Music and Computation l Research example: automatic classification
l Language: “List the sales of product in 2003” l Music:
l Inherent to all forms of perception: a structuring
l Groups in language form a tree-structure (Wundt
l Grouping structure represents how parts combine
l Very controversial claim
There exists one model that predicts the perceived structure in language, music, vision and other modalities… (cf. Newell 1999)
l Perceived incremental l Alphabet
l A to Z l A to G, #, b
l Syntax
l Language: strong relation between syntax and meaning l Music: three layers
l Scale degrees l Chord structure l Key structure
l Evolution
l Music as language l Language as music (Mithen, 2005)
l Recursion
l Language
l [he dreams] l [he dreams that [he dreams]] l [he dreams that [he dreams that [he
dreams]]]
l Etc.
l Music
l Bach’s Canon per Tonos, a.k.a endless
rising canon (see ``Godel, Escher, Bach’’ by D. Hofstadter)
l Brain
l Commonalities between language and music
are also found in neuroscience (`Music, Language and the Brain’, Patel 2008)
l Computational applications
l Key finding l Pitch spelling l Segmentation l Score following l Automatic analysis l Classification (on basis of genre/composer/..) l …
l Automatic search for key
l Automatically segmenting music in phrases
l Automatic (harmonic) analysis
l Musicology
l Automatic analysis
l Musicology
l Automatic analysis
l Cognitive science
l Understanding of cognitive processes through
modelling
l Musicology
l Automatic analysis
l Cognitive science
l Understanding of cognitive processes through
modelling
l Commercial application
l Search machines for music l Music recommendations l Music notation software (Finale etc.) l Automatic accompaniment
l Musicology
l Automatic analysis
l Cognitive science
l Understanding of cognitive processes through
modelling
l Commercial application
l Search machines for music l Music recommendations l Music notation software (Finale etc.) l Automatic accompaniment
l Two music excerpts:
l A:
l Would you classify C as belonging to A or
l C:
l Two music excerpts:
l A:
l Would you classify C as belonging to A or
l C:
l A: Rock,
l We simplify the music with a model: a new
l We compare representations
l Notes represented
l Intervals have been shown to be more important
l Melody:
l ICn: all pitch class sets that are dominated
1 1 1 2 2 3
1 2 3 4 5 6 7 8 9 10 11
l ICn: all pitch class sets that are dominated
1 1 1 2 2 3
1 2 3 4 5 6 7 8 9 10 11
l Fine-grained level: sequence of categories l Broad level: statistics of categories
Category Percentage of occurrence IC1 6.76 % IC2 5.63 % IC3 25.20 % IC4 22.94 % IC5 36.45 % IC6 3.03 %
IC5 IC3 IC3 IC3 Distribution in Debussy’s Golliwogg’s cakewalk
l Of the 6 categories, we choose 3 as dimensions l Each IC distribution can be represented as a point in
Category Percentage of
1 6.76 % 2 5.63 % 3 25.20 % 4 22.94 % 5 36.45 % 6 3.03 %
Musical period Period composers Middelages 500-1400 various Renaissance 1400-1600 Palestrina Baroque 1600-1750 Bach, Händel, Vivaldi Classical 1750-1830 Haydn, Mozart, Beethoven, Schubert Romantic 1830-1900 Brahms, Mahler, Tchaikovsky, Debussy, Mendelssohn, Sibelius Modern 1900-... Ravel, Stravinsky, Schoenberg, Webern
l Rock versus Jazz
l Classification of music is possible on the basis of a
l What is it in the music that determines it genre? l Does interval information overrule temporal
Alternative method:
Number of correct classified pieces of atonal music Number of correct classified pieces of tonal music Total number of correct classified pieces of music IC method 19 53 72 (94.7 %) Alternative method 14 53 67 (88.2 %) Total number of pieces 20 56 76
category 5 / categorie 1 > α: tonal
for each bar: find key count notes not in this key total number of these notes > β: atonal