Malthus Und Der Maler

Artists need some idea of what members of their public bring to the encounter with their work. A conventional “canon” of “stuff most educated people have probably encountered most of,” is essential for art to work, because figures and grounds are complements. It’s a set with fuzzy edges (Mozart for sure; “Symphony No. 40,” probably; Die Entführung, well…), but a lot less than everything. If you didn’t have an environment, you couldn’t have a self, and if you don’t subconsciously remember that the first movement of a classical piece modulates to the dominant, you have no way to know that anything has happened when something Romantic doesn’t, or when something newer doesn’t have a key to begin with. If you haven’t read the Odyssey, Ulysses will pass you in the night. The art phenomenon occurs inside the head of the listener/viewer/reader, and the individual work at hand is only the grain of sand around which that viewer assembles an experience that draws on everything else she has ever experienced.

It’s bigger than art, actually. It’s inconsistent with the whole idea of a culture that engagement with specific works not be shared across populations. Paulinho da Viola, in the documentary Meu Tempo é Hoje, remarks about “Carinhoso”: “If you go into any botequim or joint, anywhere in Brazil, and start singing ‘Meu coração, não sei porque…’ everyone in the place will join in.” For decades, every Brazilian singer has covered it.

Closer to home, I used to be able to use the following bit of dialogue to introduce and motivate a section on leadership in a public-management course:

KENT [in disguise]: You have that in your eyes that I would fain call master.
LEAR [in wretched circumstances]: And what is that?
KENT: Authority.

Now I just get blank stares; I can still teach management, but I’ve lost a lot of resources that have no ready substitutes (forget Hal and Falstaff, not even The Godfather works).

That a reader’s, listener’s, or viewer’s experience—not just this or that work—is the core validating phenomenon of art, and that that experience is constructive and active follows directly, but not exclusively, from a truism of psychology attested by a wide variety of evidence: perception is itself an active, interpretive process. Eyewitness recollection is famously labile and approximate. Edwin Land’s “retinex” experiments proved that color perception involves post-retinal interpretation of a whole field.1 You can render a scene with only two wavelengths as full color (which is why a lemon looks yellow by candlelight, on a cloudy day, and under streetlights, even as the wavelengths reflected from it change wildly). An early-music society can use historic instruments and tune down to the A of 1730, but no one will experience what Bach’s audience did, because we have heard Wagner, John Lennon, and an airplane. Everyone who knew anything in the ’30s thought van Meegeren’s forgeries were Vermeer’s; my public policy students pick them out on first exposure with eighty percent accuracy because they see with 21st-century eyes, and van Meegeren could not anticipate what we bring to the encounter.

The art world is much concerned to put new works into the limited attention space of the audience, and there’s nothing wrong with that. What it is feckless about is recognizing, and managing, the result of that displacement on the cultural patrimony already in use. When a new work is hung, especially if the museum is enlarged to accommodate it, the disappearance of something else from engagement is hard to see, the more so if we have the supply-side illusion that everything on the wall is functioning as art. But viewer attention, not wall space, is the binding constraint on consuming art, and taking the experiences of viewers as the fundamental art “good” has all sorts of fairly fell implications for arts policy and management. At the start, it suggests much more attention to that experience than we conventionally allocate: much less observing and talking about the painting on the wall and much more study of people looking at it.

It also demands we think about what is being looked at and maybe shouldn’t be any more, especially with regard to a body of work expected to be broadly accessible in memory. Quantity measures useful elsewhere in the social sciences are hard to find here (the money prices paid to possess works don’t tell us what we need to know). But engagement time is the fundamental resource each of us has to allocate. At least for a rough pass at the question, let’s consider a population of art engagers who manage their time so that, at the margin, each disposable waking minute creates the same value, and each of whom should count the same to society. What are the actual limits on total art engagement, and what does that imply for the size, evolution, and boundaries of the canon of core works? If Thomas Malthus had studied art instead of food and population, he might have said that the stock of art increases exponentially faster than the ability of a society to benefit from it, and he would have been right instead of wrong.2

There was a time when the binding constraint was total art on offer locally. Things have changed: the (1) accessible stock of art (2) worth attending to has exploded both forwards and backwards, as Monteverdi, Tom Jobim, and Arvo Pärt works entered the performed repertoire and Spotify made it all available with a click; the plastic arts’ availability in museums and through the Google Cultural Institute is almost limitless. So in our world, artworks are effectively abundant, and the total number of hours that can be committed to art is the primary constraint.

In the U.S., attendances at the National Endowment for the Arts (NEA)’s “benchmark” activities (museums and galleries, plays, opera, classical concerts, jazz, and ballet) averaged about five per year in 2012 (this number has been declining at least since 2008) for the third of the population who reported any attendances at all.3 If an attendance is four hours (high for a concert or play, maybe low for a museum), average lifetime commitment to these events is about 1,300 hours. Doubling it would be an extraordinary achievement for the arts; but let’s take 2,500 hours as a level of aspiration, and double it again for the more art-attentive members of the population, to 5,000 hours. Divided evenly, a third each, across theater, visual arts and music, that’s about 1,700 hours for visual arts, and at ten minutes per work (the average attention to a painting in a museum is about ten seconds), about 10,000 paintings. The collection of a single major museum exceeds the lifetime attention of this high-art-consuming individual, seeing each work once; twenty such museums would exhaust it with only their works on display, one look each. The music budget is “fractions summing to one” of about 500 operas, 3,400 symphonies/concertos/suites/chamber works, or 34,000 songs.

If we accept that recorded music is music, and that looking at a high-resolution image of a painting is “seeing art,” with one listener/viewer engaging one work at a time, the potential demand for art works is enormous. But a society’s attention capacity for a canon of “the stuff you should know” gets much smaller with several important reality corrections. First, people do not and should not look at an important painting only once; value increases over many viewings/hearings before satiation sets in. Nor should everyone spend all his art engagement time with approved masterpieces. Any idea of a shared cultural life means that many people will engage with any single work. And, finally, the core background experience of art that informs subsequent experience shouldn’t be accumulated over a lifetime, but much faster at the start, like all education and socialization. If we want this to mostly happen by age forty, total hours available for the canon are halved; leaving time for new, fringe, experimental, and not [yet] immortal works, maybe halved again. At four engagements per work, we’re down to: 1500/(4 × 2 x 2) = 94 symphony-sized units of music, and about 625 paintings/sculptures. One every eight years, worldwide, since the pyramids. Gombrich’s The Story of Art includes two thirds of that, omits all but one Vermeer, stops thirty years ago, and is heavily European.4 With subpopulations sharing different sets of favorites, and allowing for individual taste variation, many more, but the important result to this point is the very small number of zeroes left in the total, compared to the potential repertoire of considerable, even great, works that have to share attention time for classic, must-see, works.

There’s one more piece of bad news: as audience members die and more are born, that capacity starts again at zero for the newcomers, while the new generation’s artists are offering stuff relevant to a new world. New works enter the competition for attention time not only by being made fresh but also by being found and resurrected: no one paid any attention to Vermeer for a century and a half, but now he’s unavoidable in print, on museum walls, and even in a movie.

Every work entering the fuzzy canon must displace something already using part of the fixed capacity of attention time. There is no escape by technology: it doesn’t do to listen to music faster, nor to multitask it, Ariadne auf Naxos (and sampling) to the contrary. It buys us nothing (in this arena) to expand museum wall space or release more MP3 files or commission more works or give more concerts, and increasing the average time spent engaged with art only puts off the inevitable incrementally, and can’t go on forever.

We can generally expect new works not to get their appropriate share of attention for several reasons. First, the ticket-buying public has no idea of how much it would benefit from what it has never seen/heard, but it has a fond memory of the last time it met Starry Night. Second, presenting institutions are not very good at marketing (recall the business school bromide that selling is talking, but marketing is listening). Expert commentary (museum labels and program notes) treats every offered work as a masterpiece, and the heartless discipline of denying not just performance hours or museum wall space but engagement time to work A because work B was chosen, is hidden from the public. Nor are directors of institutions alert to the opportunity cost of hanging or performing a particular work; my colleagues and I have interviewed some music directors about balancing contemporary pieces against warhorses and they uniformly display “one-blade-of-the-scissors” thinking about how wonderful and important this or that candidate’s work is, but not what they are inevitably leaving unheard.5 Perhaps as important are supply-side habits like commissioning new works that will never get a second performance.

The operative canon will be pruned somehow, but leaving it to the market (or even to college distribution requirements) is almost certainly not optimal. Leaving aside buyer-seller information asymmetry—“what you think you don’t like, but would really value if you try it enough”—concert ticket sales and applause are much too crude as measures of value created; attendance at a museum or exhibition says nothing about what people are looking at or how. Museum data about works in their collections are extensive, but they are only starting to observe how many engagement hours displayed works are respectively afforded, and without knowing that it is impossible to know what should receive more and what less, let alone how to make that happen outside absurdly broad categories like “Chinese pottery.” Furthermore, and essentially, presenting fine art is not merely a matter of maximizing volume: audience experience data, though indispensable, are a complement to, not a substitute for, connoisseurship and expertise.

It’s time for presenting institutions, curators, and critics to accept their responsibility to affirmatively send work, including work that will break their hearts to cashier from service, into archives designed for scholars (1) with the same critical faculty that they use to discriminate works on offer; and (2) with the same vigor with which they pursue new work whose engagement, not just wall space or program time, will be at the expense of something truly wonderful.


  1. David H. Foster, “Color Constancy,” Vision Research 51/7 (2011).
  2. Thomas Malthus,, An Essay on the Principle of Population, (London: J.Johnson, 1798).
  3. National Endowment for the Arts, A Decade of Arts Engagement: Findings form the Survey of Public Participation in the Arts, 2002-2012,NEA Research Report #58 (Washington DC: 2015).
  4. Ernst Gombrich, The Story of Art [16th Ed] (London: Phaidon Press, 1995)
  5. A. Turrini, M. O’Hare, & F. Borgonovi, “The Border Conflict between the Present and the Past: Programming Classical Music and Opera,” Journal of Arts Management, Law & Society (Spring 2008).


Michael O’Hare

Michael O’Hare is Professor of Public Policy at the University of California, Berkeley. Trained as an architect and engineer, he is an author of Patrons Despite Themselves: Taxpayers and Arts Policy and has worked at the Museum of Fine Arts, Boston, in Massachusetts state government, and held teaching positions at MIT and Harvard University.