Or, Research matrices of the romance writer; Or, Why I play with generalities of research and not particulars.
I am not what you might term a stickler for research. I don’t need every detail of life in 19th-century London to be precise either in what I read or what I write, in the sense of extant shops and clubs, and theatrical productions actually performed at the time I need them to have been performed. As a writer, I am not going to restrict myself to complete accuracy, as long as what I am describing fits the spirit of the age and could have happened.
Thus I will reference a performance of an opera in the spring of 1822 that may or may not actually have been played that season, but could have been based on its London premiere having been in 1820 and wildly popular. I have spent the past 30 minutes trying to track down an appropriate aria for a young woman to sing at a musicale to indicate both her prowess as an amateur and the range and timbre of her voice without tedious descriptions of it which may or may not help the reader “hear” her, anyway.
The requirements were:
- documented to have been performed in London at least as early as 1821
- contains a contralto/lower mezzo-soprano part with at least one notable aria
- contains a contralto/lower mezzo-soprano aria that would appeal to a younger woman to perform
The actual historical performance off such a part would have been an wonderful bonus, but to the best of my knowledge there is not a way to merge the two data matrices of [operas performed in London during 1822] X [operas written before 1822 containing significant contralto/lower mezzo-soprano part]. If there were an easier way than cross-comparing a list of one against a list of the other by hand–be it my own eyes scanning one against a mental list of the other, or creating a unique spreadsheet to let my computer compare the lists side by side–I would have put more care into whether the opera I selected had actually been performed that year.
But this is the kind of detail that it is not worth spending five hours to discover. All that I need for the spirit of historical accuracy is whether the opera could have been performed again that season on the basis of it already having been peformed. For that, Wikipedia was good enough. Contralto entry, notable contralto roles section; specific listings selected for exploration based on my (limited) knowledge of composers. It took half an hour to find the part I needed, and that included reading the voice descriptions for all female parts to be certain I was classifying my heroine’s voice correctly.
Thus I have her singing a piece from Angelina (Cinderella) in La Cenerentola by Rossini, after having seen it performed professionally the year before. That opera probably wasn’t actually peformed in London in 1822…but it could have been. And that is good enough for me.
This is how authors roll, my friends.