About Cantometrics

Cantometrics is an approach to understanding musical variation across cultures and its relationship to social factors. It was designed to privilege all musical systems equally with the assumption that all are equally valid and functional. Cantometrics was developed by musicologists Alan Lomax and Victor Grauer in the 1960s as part of a larger research project on expressive style around the world including dance, conversation, breath intervals in singing, vowel and consonant placement in singing, instruments and ensembles, popular songs and more. (See A. Wood, P. Savage et al, The Global Jukebox, a database of the performing arts. submitted.)

To achieve this, Cantometrics uses cross-cultural comparison. 37 key musical features were selected during months of listening to recordings of music from throughout the world. These standardized variables were assigned ranges so their variations could be coded in a large sample of songs from all areas of world culture. A variety of statistical tests were then used to search for relationships and correlations within the resulting data. Currently nearly 6000 songs were analyzed in this way, but the sample is growing.

In 2021 this data was cleaned, normalized and publicly released for the first time, and key correlations between musical traits and social structure have been rigorously retested and found to be robust. The Cantometrics data set is unique in its coverage and detail, and can contribute to our understanding of musical and cultural evolution. It is designed to complement, inform and dialogue with detailed ethnomusicological studies.

In 2024 Anna Lomax Wood and John Szinger revised the original Cantometric Coding system in order to make it easier to accurately evaluate music using the system and to achieve a higher rate of consensus among coders. For coding lines 5, 8, 16, 23-28, 30, and 33-37, the range of possible values was reduced by combining ambiguous distinctions into a single choice: for example, in Line 24: Tempo, the values “Very slow” and “Quite slow” are now represented by “Very slow tempo”. Wording was also made clearer and more consistent across coding lines. The new system is called the “Revised” system; the original version is called the “Original” system.

All existing Cantometric codings for songs in the Global Jukebox were updated to the new system, which is now presented as the default view.

About this Course

  1. It is based on listening and ear training. It is for everyone, including those who have not had prior musical training.
  2. For those who’ve not had musical training, the lines on Melodic Intervals, Melodic Range, Position of Final Tone, and Rhythm could be challenging. Read about these features in the Coding Guide, do your best, and keep going.
  3. Access the coding instructions for each line by clicking on or read the Guide to Coding Cantometrics in the book, Songs of Earth.
  4. At the beginning of each segment, there are examples of the world extremes of each musical feature. Listen to these extremes several times to hear the range of each musical feature. For example:

Simple Litany (1-2 phrases repeated) <==> Through-Composed (no repetitions)
Or:
Maximal Vocal Blend <==> Individualized
Or:
Very Soft <==> Very loud

Good luck! Good luck! May this journey reward you!

Learning Plans

Lessons are automatically sequenced according to the plan selected. Altogether, to complete the entire course of 37 variables takes about 40 hours. Shorter alternatives are on offer. Click on your choice:

  1. Plan 1. Study the 37 variables in their original sequence.
  2. Plan 2. The 37 variables are grouped according to their musical attributes and functions. Each grouping treats a distinct aspect of musical performance: Social Organization, Musical Organization, Rhythm, Orchestra, Vocal Qualities, Ornament, Articulation.
  3. Plan 3. Abbreviated Study. Focuses on 9 variables; takes six hours or less.
  4. Plan 4. Quick Study. Two alternative Lessons each of which can be covered in one or two class periods. With these it is possible to perceive the main families of song style.
  5. Plan 5. The 37 variables are organized into factors, or latent variables, which are primarily the product of factor analysis, a statistical procedure correlating the variance among multiple variables, and arriving at two or more latent variables (factors) that represent the relationships between the correlated variables. Factor analysis makes it possible to understand and interpret these relationships, and to represent multiple variables with fewer variables. (Based on H. E. A. Tinsley & S. D. Brown (Eds.), 2000, Handbook of Applied Multivariate Statistics and Mathematical Modeling. Academic Press). Put more simply, the factors make sense of the variance of the variables and give them deeper conceptual significance. They help the researcher to understand her data and findings at a higher level.