Interesting idea, but there are problems with the accuracy of calculating P/E and its interpretation. First, while "earnings" is a scalar quantity that can increase infinitely, coverage in encyclopedias is not. No matter how large a composer's "true" value, there is an upper limit to coverage in an encylopedia (as opposed to a biography), at which point there is simply no more information that is important enough to print. (Mozart's favorite flavor of Jamba Juice?)

There are idiosyncratic factors that can limit the amount of material available to researchers, thus limiting encyclopedia coverage under-represent a composer's "worth". Did the composer live a very private life? Did s/he not engage much in correspondance that was saved? Died young? Died in poverty & personal effects were sold or thrown out?

There are factors that can limit the ease of recording a particular work, which would artificially limit its number of recordings or "price" in your model. Some factors that come to the top of my head: extreme technical difficulty (Rachmaninoff), scored for instruments that are largely out of current use (medieval period, perhaps renaissance?).

Even if a perfect calculation of P/E for a composer were calculated, what would it mean? Music posseses intrinsic subjective value, which is unlike stock in a widget-producing firm - which can be bought and sold in a massive marketplace, and worth exactly what other people are willing to pay for it. We listen to music for reasons other than the composer's historical importance. Likewise, we gossip about female celebrities for reasons other than their physical attractiveness.

I use his score in Charles Murray's Human Accomplishment, which measures how much space he is given across a wide variety of music encyclopedias -- how deserving he is

Why is mention of a composer's worth measured by how often he is mentioned in an encyclopedia? Those renaissance composers may be mentioned because they were first, or innovative - people who write encyclopedias like to dwell on questions of origin - rather than because they are enjoyable.

Reminds me of a professor in college who told us the the most important film in history was 'Birth of a Nation' - maybe it was an important film, being the first and all, but nobody would watch that today for pleasure.

Anyway, one could easily make the case that the the numerator and denominator ought to be reversed, that the fundamental value should be decided by Amazon consumers rather than by encyclopedia writers.

Cool.

While it's not orthogonal to "popularity", you may want to consider that different periods of music are performed by different kinds of musical groups -- full vs. chamber orchestra, for example -- which may have different rates of recording music for reasons besides sales prospects. Opera stands out as a genre where there may be few recordings relative to sales, due to the costs of mounting a production.

From the list of 'over-valued' composers you have Puccini, Bizet & Verdi who are primarily well-known for Operas, Tchaikovsky for Ballet, Grieg for the score of a play - maybe these are genres that are not considered as serious and are therefore do not merit as much importance for encyclopedia writers.

This somewhat strengthens my opinion that the fundamental value should be determined by Amazon consumers, since I don't think we live in an age where Opera music is overhyped.

I hardly ever listen to Classical music, so my input here may be completely off the mark.

Interesting, but I see a problem with your metric. It seriously overvalues someone like Schoenberg, not because his music is unlistenable (though most of it is), but because his encyclopedia entries are inflated by the importance of all the words he wrote about music, which really were a more significant contribution than his actual compositions.

Also, did you mean Adriaan Willaert, not Anton?

The main idea is interesting, but the actual comparison may be flawed.

Unlike sales which are set by the general public, the music encyclopedias are written by critics. The latter group has a natural attraction towards obscure composers that general public does not know (like your “most under-valued” group). They are also highly influenced by factors unrelated to the art quality, such as artist’s biography (for the extreme examples of this see “Disumbrationism” or “Ern Malley”).

This is very unlike stocks which, overvalued or not, still require everyone to pay the same monetary price.

I'm worried that this method may overstate the hype for artists with single important works. Vivaldi's Four Seasons is a piece where everyone should own a recording, while Schoenberg managed to revolutionize modern music without ever creating an obvious 'must-own' work. In general, I expect that composers with single great works would be 'overhyped' on this methodology because there is not a lot to spend column inches on in encyclopedias, and people may generally understand that they are not important composers (i.e. Saint-Saens) even as they want to own that one piece.

Opera stands out as a genre where there may be few recordings relative to sales, due to the costs of mounting a production.

Which makes it interesting that several of the "over-hyped" composers (e.g., Puccini) are primarily known for operas.

In any case, it's true that number of recordings offered for sale is only an indirect measure for the interest in a particular composer's works. I think there is muc more direct way to obtain a measure of current popularity of composers: Measure the amount of time people spend listening their works, as measured by the service like Last.fm, which records the number of times a particular track on a particular recording is listened to either online (via streaming from Last.fm) or on iPod or similar device. (For example, one particular recording of Puccini's "Nessun Dorma" has been listened to in its entirety more than 39,000 times in the last week.)

You could then take the number of times a track is listened to by all Last.fm subscribers, multiply by the duration of the track (to get the total time spent listening to the track by all listeners), sum across all tracks attributed to that composer (to get the total amount of time spent listening to that composer's works), and then divide by the total time spent listening to works by all composers (to get the relative time spent listening to this particular composer's works).

The Last.fm data are a bit messy in places, and I don't know how difficult they would be to get in an easy-to-use form, but the project is doable at least in theory. (There are other services like Pandora that could provide similar data.) It could also give a pretty current measure of popularity (at least week-by-week based on the public data) to determine changes in relative popularity of composers over time.

"...music that appeals to more emotional people is over-valued, while music that appeals to more cerebral people is under-valued."

Mozart failed to find his way onto the over-hyped list. Seems odd to me. Don't mean to change topics, but I'm curious as to whether other lovers of Baroque music share my lack of enthusiasm for Mozart.

objective judge the composers' excellence

How does one exactly objectively judge a composers' excellence?

Nah, it's not hype, it's just the difference between scholarly and popular taste (although in this case, "popular" is referring to the top few % of the public in terms of musical appreciation ability).

Many of these names on the under-hyped list are, I would guess, composers from the early days before, say, Vivaldi. Music historians need to write about them in order to tell the story of how we got up to the point where composers were able to compose music that is still popular today, but the public doesn't have to listen to their music. It's like if you were writing a scholarly history of the Beatles, you'd have to write about the skiffle groups of Liverpool in the 1950s that John and Paul listened to as adolescents, but the public doesn't want to listen to them, it wants to listen to the Beatles.

Let me paste in here something from my review in The American Conservative of Murray's "Human Accomplishment" because it will help you comprehend what the book measures and what it doesn't:

Can we trust these data? The scholars upon whom Murray relies have their personal and professional biases, but, ultimately, their need to create coherent narratives explaining who influenced whom means that their books aren’t primarily based on their own opinions but rather on those of their subjects. For example, the best single confirmation of Beethoven’s greatness might be Brahms’s explanation of why he spent decades fussing before finally unveiling his First Symphony: “You have no idea how it feels for someone like me to hear behind him the tramp of a giant like Beethoven.”

In Paul Johnson’s just-published and immensely readable book Art: A New History, you can see how even this most opinionated of historians must adapt himself to the judgments of artists. Much of the book’s entertainment value stems from Johnson’s heresies, such as his grumpy comment on Michelangelo’s Sistine Chapel: “No one ever wished the ceiling larger.” Still, Johnson can’t really break free from conventional art history because he can’t avoid writing about those whom subsequent artists emulated.

For example, Johnson finds Cézanne (who ranks 10th in Murray’s table of 479 significant artists) painfully incompetent at the basics of his craft. Yet, Johnson has to grit his teeth and write about Cézanne at length because he “was in some ways the most influential painter of the late nineteenth century because of his powerful (and to many mysterious) appeal to other painters …” In contrast to Johnson, Murray keeps his artistic opinions upbeat or muted because his goals are scientific.

Just to throw in something else here. I think it'd be really interesting to try to quantify the pleasure people take in different works of art. What's their brain activity look like? What about their biochemistry, their heart rate?

It'd be really interesting if there was a waist-hip ratio equivalent for musical beauty. I know nothing about music theory and for all I know such measures may be out there.

Having now looked at your dataset, I find it quite useful. Evidently, the craft of composition had reached maturity in the second half of the 19th Century. By then, all the tools to appeal to the music-literate public had been created. (Strikingly, the composer universally believed by musicians to be the greatest genius of the second half of the 19th Century, Wagner, is not particularly popular these days.) After about 1900, however, the enormous richness of the 19th Century repertoire weighed down composers, driving them into experimentation, most of which has not proven enduringly popular with the CD-buying public. (Wagner got there first, and was able to create popular interest in his innovations that lasted for a few generations, but has evidently faded.)

Fun exercise.

Just to join in the fun of being picky, can I suggest one tweak? Where you refer to "Western composers," maybe it'd make sense to call 'em "Western composers in the 'classical' tradition," or some such. There are loads and loads of important, influential, and popular composers in the West who haven't worked in the classical tradition -- just in America, Cole Porter, Duke Ellington (who did try a few ambitious, concert-hall experiments), the Brill Building gang.

And of course all those great black musical artists whose genius and influence go a little underdiscussed around these parts, I can't imagine why ...

But very interesting to eyeball your stats and thoughts, tks.

One, possibly trivial, factor to consider is that composers, who are both popular & prolific, will be overrepresented. So, Mozart, with atleast 626 works, enjoys more results at Amazon than Beethoven who has ~200 odd works among the 138 ?opuses officially published on his behalf.

You argue that older music is under-hyped, but wouldn't it be simpler to suppose that music has gotten better with time? It could also be that people prefer more recent music to older forms.

Up until the 20th Century, there's a clear trend in the increase of popularity of music. The reason for the drop off in the 20th Century is likely because of the rapid emergence of other styles of music, which are not depicted in the graph.

Of course Angelina Jolie gets more hits today than Jean Shrimpton. Shrimpton is in her 60s, and has been out of the public eye for 30 years.

As Danny suggests, the numerator and denominator ought to be reversed. The amazon list reflects (imperfectly) what people are buying, which is surely a closer parallel to "earnings" than encyclopedia entries. The entries reflect theory or criticism, which is closer to "price." Music criticism is always questionable, and has gone downhill in a big way in the past couple of generations, culminating the juvenile absurdities of post-modernism. Thought experiment: Imagine making a similar list for architects. The encyclopedia entries would stress Bauhaus, Le Corbusier, etc. The "earnings" list (commissions earned) might place them fairly high as well--but that's because architectural commissions are usually awarded by people concerned about their status among intellectuals and pseudo-intellectuals. The exercise is more meaningful with music precisely because people for the most part buy the music they actually want to hear. Is Schoenberg a better musician than Puccini? I don't think so! More innovative perhaps, but novelty does not equal beauty or excellence. (Most innovations are soon justly forgotten.)

All that said, like Jamie (and many music critics), I prefer Bach to Mozart or Beethoven, but he probably "sells" fewer recordings. There are obviously cases of artistry which is too demanding to find a wide audience, but is nonetheless genuinely excellent, not merely attractive to theorists. However, that sort of thing is hard to quantify. (One might look at the purchases of people of different IQ levels, but that's hard to compile.) For a rough measure, I think this P/E metric is informative, with the numerator and denominator reversed.

I have recordings by all these composers. The ones on the undervalued list are all niche interest. It would be hard for anyone who enjoys music to value Schoenberg any less than he deserves.
How can Brahms be overvalued? Everything he wrote is worth listening to.

Cool! I had seen neither the composers nor the artists posts until this morning.
I agree that 1950 is too recent a cutoff date for the arts (viz Irving Kristol's mot that colleges should not teach any literature written less than a hundred years ago). I have the raw data for recalculating the rankings for a cutoff of 1900, and am tempted to do it, if only I weren't so swamped....