Photo: from left to right: Puccini, Giacosa en Illica
The multivariate analysis is a useful research tool for questions where multiple variables play a role. This type of statistics is also used in various other disciplines, both in nature as in social sciences, used for modelling systems. The Wageningen researcher Saccenti usually uses the multivariate analysis for synthetic biology. Now he is taking a side path by including his beloved classical music. He is not only a Wageningen researcher, but also member of the Italian Music Critics Association and editor of the online Italian music magazine OperaClick.com.
For years musical experts puzzled on who was responsible for the famous operas of the Italian composer Giacomo Puccini in the early 20th century. Puccini, ‘the greatest composer after Verdi’, wrote important operas such as La bohème, Tosca and Madama Butterfly. But he did not do this alone. A famous duo, consisting of rhyme artist Giuseppe Giacosa and stage artist Luigi Illica, provided important contributions to the well-known operas. The question was: what was each contribution to the end result? The correspondence between the famous composer and the two poets offered no decisive answer. The multivariate analysis did, one hundred years later.
Saccenti solved the linguistic issue with his computer. He took writings of Giacosa and Illica before their successful cooperation and got 34 texts from Illica and 28 writings of Giacosa. A computer converted these into tables, using word choice, word length, and syntax. Additionally, the three shared operas of the poets were cut into nine segments, and they were subjected to a computerised language analysis.
And this led to new insights. In literature research Giacosa was often named to have had the most influence on Puccini, the researcher Saccenti actually found that Illica’s impact on the poetic tone of three libretto’s was much larger than expected. That is why the prevailing conception that Illica was only responsible for the dramatic parts of the songs, must be revised, the researcher concluded in the scientific journal Digital Scholarship in the Humanities.
‘This publication was a way to combine my two main areas of interest – mathematics and music’, Saccenti said. ‘I try to apply the multivariate statistics in humanities more often. I have also analysed the Devine Comedy of Dante.’