Conclusion
Jump to
Table of contents — Introduction — Chapter 1 — Chapter 2 — Chapter 3 — Chapter 4 — Bibliography — The Spreadsheet — Summaries of analysed works — Thesis in PDF
The final answer to the research question, as to what patterns can be found in the relationship between the harmony and the text of secco recitatives composed in France between 1700-1730, is not very clear-cut. As mentioned in chapter 3.1, results were too few for general conclusions to be drawn, so only partial correlations emerged. This research is, therefore, best considered as a demonstration of the potential of such an experiment. Although the results are inconclusive on a larger scale, they certainly prove that some correlation does exist. It cannot be said what exactly a given harmonic progression might mean, but it certainly means something, and having data-driven proof of that is of value. There is no reason not to talk to singers about harmony; even if this research cannot establish how to interpret any given harmony, simply making the effort of interpreting it adds depth to one’s expression.
Since this experiment and this paper were loosely based on the structure of a scientific experiment, it seems important to note the major sources of bias of this research. First, the cataloguing of harmonic passage and associated text pairs was made by selecting harmonic moments of interest and noting their text, rather than the other way around. This means that the conclusions detailed here do not go both ways; though plagal cadences are indeed often associated with words of love and tenderness, no claim can be made from this research as to how often words of love and tenderness are associated to a plagal cadence. There is no guarantee, therefore, that finding a particularly tender text will mean finding many plagal cadences.
Also, as has been briefly touched upon earlier in this paper, the subject of harmony and of harmonic analysis is a slippery one during the Baroque period. Indeed, it was only during the period studied, 1700-1730, that a vertical understanding of music was even suggested. As such, the Roman numeral analysis that was performed in this study, though justified, is nevertheless not indicative of composer’s intentions. Indeed, it is probable that if composers did associate certain colours to certain affects, they created those colours through the use of dissonance and counterpoint, and any resulting correlation to harmony found here was simply a secondary consequence of these choices. It is suggested in chapter 3.3.5 that Rameau and Clérambault used the V/IV chord differently. However, during a discussion with the students of the Early Music Studies #3 class given by Dr. Kathryn Cok, the very valid point was raised that since the two composers used these chords in different inversions, they might not be the same chord at all. It may be that grouping them in the same category boils down to imposing a modern musical notion on music to which it does not apply.
Finally, there is the simple fact that the harmonic analysis was done by a human. Though the analysis process was carefully chosen, and consistency was rigorously applied, it would be near-impossible for a human brain to entirely dissociate itself from the music that is familiar to it and the harmonic customs it has been exposed to since birth. Performed by a different person, then, this entire research might have produced entirely different results, and that is a fascinating subject in itself.
As a final thought, the possibilities of expansion of this research are many and varied. For a start, it would be possible to simply analyze a greater number of secco recitatives from early 18th-century France. It would also be possible to increase the number of composers studied, to see if trends can be extended further. Arias and dances could be added to the dataset, to see how the correlations found in recitatives fit the pattern, or not. The harmonic analysis could also include individual jazz chord names, like G or Gm, as a way to determine whether specific chords correlate to emotion in the text. On a much larger scale, composers from other countries and time periods could be added to the study. Finally, if the resources could be made available, a data analysis software could also be used for faster and potentially more consistent and precise harmonic analysis and subsequent pattern analysis.