But what if the new specimens fall flat? What if the readers and theatergoers who like such things prefer the ones that came into being when the forms were fresh? What if the real reason there have been few epic poems and verse dramas lately is not that the gatekeepers won’t give them a chance, but that creative inclinations have moved elsewhere? Maybe the kind of imaginative vision that generated long, episodic narratives like Orlando Furioso is now generating long, episodic narratives like The Wire. Maybe the people who have a gift for putting music and plot into an evening’s entertainment don’t feel the necessity of acoustic instruments or old-fashioned singing; maybe they want to write Hamilton.

***

What is a successful opera, and when did we start worrying about the supply of them? Let’s start with two simple, arbitrary benchmarks for “success” that seem reasonable in context:

1. presented by at least ten theaters within a decade of premiere
2. presented in at least four theaters outside the country of origin, again within a decade

Most operas never met these requirements, so we’re not casting the net too wide, and yet for well over a century the top few regularly did so, so we’re not being too picky either. We’ll fail to catch operas that didn’t succeed when new but did so later (Les Troyens and Jenufa, for example), but those are exceptions. We’ll also fail to differentiate mere successes from mega-hits (Aida, La bohème and Carmen met the requirements at least ten times over). And we can’t use these criteria to evaluate the Baroque era, when geographic circulation of operas was not yet the norm – circulation started where the standard repertory starts, around the turn from the 18th to the 19th century. But from that point forward, as a decent measure of success- at-the-time, these benchmarks work. They document operas that made opera a going concern.

So, how many operas have been “successful”? The following tally isn’t scientific. It’s based on archives a reasonably diligent person can find online or in a library reference room. I’m sure the numbers aren’t exact, but I’m also sure they don’t misrepresent the situation. Counting in two-decade spans from 1810 to 2009 (and including known plans through 2019 for the ten-year benchmark of those premiered in 2009) I get the following totals:

Already there is a distinct break-point. If we add one tougher criterion, reflecting what 21st-century presenters believe their customers will pay to attend, that point takes on the aspect of a sheer cliff:

3. programmed twice within a five-season span in at least eight cities since 2000

That lonely entry before the list zeroes out, in case you weren’t sure, is Porgy and Bess.

Does this look bad? It’s actually much worse. The period 1810-1929 is full of operas that would still make the cut if you raised the first-decade theater requirement to twenty, or to fifty. In the postwar period, just nudging the foreign-theater benchmark from four up to five would make almost half the list drop out. And don’t even think about comparing box-office receipts.