Art, science, religion... and economics

In which the hidden connection between the oldest attention economies is laid bare.

“Since I started to deal with information theory I have often meditated upon the conciseness of poems; how can a single line of verse contain far more ‘information’ than a highly concise telegram of the same length. The surprising richness of meaning of literary works seems to be in contradiction with the laws of information theory.

The key to this paradox is, I think, the notion of ‘resonance’. The writer does not merely give us information, but also plays on the strings of the language with such virtuosity, that our mind, and even the subconscious self resonate. A poet can recall chains of ideas, emotions and memories with a well-turned word. In this sense, writing is magic.” — Alfréd Rényi, cited by Gábor Székelyi (1986).

Tribalism is the shared belief in counterfactuals.

I spent the last few newsletters mapping out the role of belief structures in economics and politics, so I can wrap up the topic by applying it to some other domains before I go back to supply chains, production schedules, market mechanisms, and other topics we might consider within the scope of “economics proper”.

This goes back to me leisurely thinking about “what is the economic value of art?” while devoting my actual work hours to the much more pedestrian task of finding a formal expression for the collective adoption of technology standards.

Because, again, there are some obvious commonalities between the two.

My endeavor was mostly motivated by my dissatisfaction with the “textbook” explanation of art as a deliberately wasteful effort sustained by some type of conspicuous consumption by high-status individuals known as patrons.

Not that it is necessarily wrong — patronage as status consumption clearly plays a role not only in the arts — but as an explanation it feels insufficient, in large part because it implicitly posits that art has no redeeming economic value beyond the one-upmanship between rich people.

The corollary of which would be that in times of crisis, it is disposable.

There is certainly a correlation between the social stability and economic prosperity of a commonwealth and the production and consumption of art, especially its elaborateness, and we had multiple junctures in history — the reformation, the transition from baroque to classicism, modern architecture and its disavowal of “ornament” come to mind — where harder times translated into calls for more austere forms of artistic expression.

Casual empiricism seems to back this notion up: the more disposable income we have, the more we shift consumption from necessity goods (things that support our livelihoods) to luxury goods (things that support our wellbeing). Economic downturns have an outsized effect on the market demand for high art (paintings, sculptures) and popular art (concerts, films) alike.

Frances Hodgkins: Solva (Fishing Village in Pembrokeshire), 1936.

But as an explanation for a universal trait that precedes human civilization and might even have played a role in its establishment, it comes across as a bit foreshortened, another case of “economic imperialism” that tries to impose economic rules to domains outside their scope.

Art and the production of deliberate counterfactuals

Instead I will try to explain the role of art using the terms I've already introduced. The attention-seeking actor who rationally directs her attention to the issues with the highest propensity for surprise. Belief clusters built on shared counterfactuals as solutions to intra-organizational goal conflict, and belief propagation as the coordinated updating mechanism.

From these starting points, I can propose that the economic role of art is the production of shared beliefs, or a bit more elaborate: the production of shareable, coalition-forming counterfactuals.

And as it turns out, art is not the only human endeavor with this purpose.

Shared counterfactuals and social cohesion

Stepping away from art for a second, let's recapitulate why we need counterfactuals at all, why we produce them, and why we would relegate the production of counterfactuals to someone else.

Two fundamental problems which economics tries to answer are “how do I get there from here?” and “how do we get there together?”, the latter expressed as “how do we resolve the conflict between moving forward and staying together?” in my newsletter on organization.

The shorthand for tackling these questions introduced by Joe Bain is “structure, conduct, performance”. Or in other words, if I take an action (“conduct”) in a given situation (“structure”), what outcome (“performance”) will it produce? This is the concept that underpins decision theory — can we take multiple decision alternatives and compare them against each other with respect to some defined objective.

But this is also the concept that underpins game theory (and in turn, modern microeconomic theory), where we accept that different people might not only assign different values to the various alternative outcomes, but also different likelihoods that these outcomes are reachable with the proposed actions. These differences in assigned likelihoods is what we know as beliefs.

But any organization requires coherence not only in objectives, but also in beliefs that certain actions yield certain outcomes. This is the fundamental counterfactual required for social cohesion: that we not only agree that certain goals are desirable, but also that we can reach them with the proposed actions.¹

Organization is the concatenation of contingent efforts, meaning we organize when all efforts are needed to produce a desired outcome, and the withdrawal of a single effort by a single contributor can unravel the entire endeavor.

Few gratifications in life are instantaneous, and that's even more true for the kinds of effort that are so intricately interlinked that we need to build organizations around them. Participation, especially voluntary participation in an organized effort requires a shared belief that collective action now leads to a desirable collective future outcome, and that the fruits of the collective effort will be distributed according to the value of the individual contributions — a steep hurdle.

This quandary is encapsulated in the popular coordination game stag hunt. In a group of hunters, each hunter has to decide in secret whether to join an organized effort to hunt a stag, or to hunt a hare on their own. Hunting a hare will always be successful, but hunting the stag will only be successful if everyone joins in. Every hunter would prefer their share of the stag over the hare, but the hare over nothing.

As a game template (and thus as an economic pattern) stag hunt skips over a lot of the complications I mentioned, for example how exactly efforts depend on each other or how contributions and rewards differ between participants.

But it beautifully isolates the underlying trust dilemma: the social reward for joining forces is understood by everyone, but coordinating efforts requires the shared belief that everyone else sees it the same way and acts accordingly. This mutual trust can unravel even faster if efforts, rewards, contingencies, and ultimately beliefs, are asymmetric.²

Some harsh economic realities about art

Coming back to the topic of art, this little excursion should help us answer the two questions: why art is being produced at all (the simpler one), and how it helps the world not only in a spiritual wellbeing sense, but in a hard-nosed, tangible, economic prosperity sense (the more complicated one).

Expressed differently: can we show that not only that prosperity produces art (which would fit the “art as conspicuous waste” theory), but also that art produces prosperity?

For this I need to step into the depths of the harsh economic realities of the art market to highlight a few common features. The first is that collecting art is not just a form of conspicuous consumption, but also a form of conspicuous speculation, and the ability to discover unknown artists and artworks who later become famous is not only a lucrative enterprise, but also a source of personal pride for many collectors.

Second, and this is true both for “high art” and “popular art”, the economics of art production and consumption reveals a regularity that's often summarized as a winner-takes-all dynamic.³ Take the example of popular music, where we can now, thanks to digital music platforms like Spotify, Tidal, or Menudo, quantity just by how much supply outstrips individual demand.

For the sake of expediency, take a music platform with 10 million artists, 100 million songs, and 600 million users.⁴ At an average of 3 to 3½ minutes, this gives us something like 600+ years of listening pleasure without repetition, or some eight to ten lifetimes. But seeing that we have about one million users for each year of music, we should expect that each available piece gets a considerate amount of attention? Apparently not, by some estimates 80% available music remains mostly undiscovered. On the other end of the spectrum, a select group of artist manages to accumulate more than 10 billion streams, equivalent to 10,000 years of listening pleasure, each year.

In economic terms, the modern art market is structured as monopolistic competition. In other words, while there is typically ample supply, even oversupply, of artifacts of a given type (paintings, sculptures, pieces of music), they have different enough characteristics and provenances that they are not easily exchanged. Few people would trade in a Picasso or Rembrandt for a portrait from a painter at a tourist hotspot.

This isn't something that's exclusive to the art market — most of the output of the apparel industry finds no buyer and goes to waste — but we rarely encounter such a stark concentration of attention on a few artists and artworks in other industries, hence the sobriquet winner-takes-all market.

The third and final characteristic about art is that art is in essence qualitative. Of course we can try and extract some quantities out of most artworks — paintings, and sculptures have physical dimensions, songs have durations and beats per minute, and we can apply some mathematical reasoning to color or music theory to uncover underlying harmonies — but we'll find little correlation between these quantitative pointers and the success of a particular piece of art. Modern art has also deliberately challenged the connection between skilled artisanship and art, much to the displeasure of some more conservative-minded patron.

What is left is that between the physical inputs that go into an artwork: skilled labor and materials, and its performance, much of the prediction rests on some unobservable factor we could call (to borrow a term from patent law) “flash of creative genius”.

All of this makes the art market a stark form of an economic market: an almost pure attention economy where the intuitive connection between effort and success is severed, no matter if we think of it in terms of immediate popularity or immortal fame, or popular vs high-minded recognition.

While art follow some obvious economic rules: people trade art and speculate in it, people have often strong preferences about it, we even build a particular type of economic engine: a recommender system, around it to try and turn to the prediction of preferences in art into a science, in many other aspects it seems like it's violating all kinds of economic rules. And that's a pattern it shares with other phenomena, as we'll see.

Counterfactuals and unobservables

Counterfactuals are proposals we hold to be true even if we know they're not, for the simple reason that they allow us to bridge the gap between actions and outcomes outside the observable world. They especially allow us to make comparisons (“what would've happened if…”) even if we can't actually draw from a variety of alternative lived experiences — we just make them up.

Judea Pearl, the Turing Award winner credited with formalizing causal reasoning, offers the example of the Lion Man found in a cave near Ulm, Germany, as the first known human-created artifact that — with a head of a cave lion and a standing human body — is deliberately counterfactual. The argument goes that the figurine, hand-carved with great effort out of a mammoth tusk to combine human and animal features, gave its owners the confidence to team up and confront real cave lions, at a time when they posed a present danger.

In other words, the lion man carvers engaged in an early form of goal-directed myth-making. Maybe the stag hunt game should really be called cave lion hunt.

This explanation is undeniably somewhat speculative, but it's noteworthy that we're still doing the same. Professional sports teams, states, even supposedly rational universities use (often anthropomorphic animal) mascots to appeal to a shared spirit and to instill confidence. The belief that we can summon the strength of the cave lion isn't just a superstition popular with paleolithic cave dwellers.

This all happens with the full understanding that these mythological creatures are outside the realm of physical or biological possibility. The communal leap of faith (or in Kierkegaard's words: the transition from unbeliever to believer) enabled by the artistic effort is the economic contribution: the one little trick that creates a shared belief structure.

Art has advanced since the upper paleolithic, and especially the technologies available to create art have advanced, sometimes pushed by the artists themselves, like al fresco painting, sometimes readily adopted, like the use of fast-drying industrial pigments by the impressionists, sometimes providing the impetus for a major shift in artistic expression, like the emergence of photography ushering in the era of abstract and nonfigurative art (and of course photography itself as an art form).

What has remained unchanged is the role of art in shaping beliefs, especially shared beliefs, and beliefs about what we collectively consider possible. And what also seems to have stayed the same is the protracted process of public adoption of new art movements — a belief propagation process that is eerily similar to the adoption of new technologies.

Indeed artists are often accused of courting controversy for the sake of controversy. But we just have to look back at paradigmatic shifts in art history like the discovery of the laws of perspective, which replaced the established order of giving most prominence to the person with the most worldly or spiritual gravitas to whoever happened to be closest to the observer in a simulated three-dimensional Euclidean space.

As we now know, perspective won out — at least for a few centuries until the “illusion of depth” was challenged by modernist painters — but we get a glimpse of the momentousness of this upheaval once we realize that the otherwise rudimentary architectural details in Leonardo's Last Supper served to reconcile the two views — the use of perspective applied to elevate the most prominent figure in the painting: Jesus.

The unobservable and the unmeasurable world

The discussion so far was almost entirely about art, but I could have written almost exactly the same about science or religion (and maybe I will in the future), in part because they all represent variant forms of an attention economy with its fundamental exchange “surprises for eyeballs”, but even more importantly because they share almost identical adoption patterns geared towards creating belief coalitions (aka tribes) based on shared acceptance and mutual reinforcement of counterfactuals.

Art, religion, and science also have a long, closely intertwined history. We just have to look at famous examples like Stonehenge and other prehistoric megaliths like Göbekli Tepe, the pyramids in Egypt and elsewhere, or even richly decorated medieval cathedrals to recognize that we're dealing with kindred endeavors that often joined forces. Was Stonehenge intended as an astronomical observatory, a marvel of civil engineering, a religious cult center, or simply a piece of art?

Science, of course, fell into the domain of religion until not too long ago. We might think of the Copernican revolution as the moment when religious and scientific inquiry diverged, but Western universities started out as schools to train priests and didn't shed their affiliation until the land grant era.

Both are tasked with making solemn pronouncements about the ultimate unobservable: the future, and the main difference between the scientific method of predicting the future and religious divination is that science requires the pronouncer to reveal their work and to avoid making vague, ambiguous, and opportunistic statements — a high standard academic practice doesn't always uphold.

Since the secularization of scientific research, one could claim that religion has retreated to the domain that is invariably outside the realm of empirical inquiry: the afterlife, or maybe a bit more prosaically, the certification of good conduct in this life that leads to rewards in the afterlife.

This might indeed be the best way to draw boundaries around the domains in which the three produce their shareable counterfactuals.

If the native domain of religion is the unobservable world, the domain of art is the observable but unmeasurable world, and ultimately the domain of science is the quantifiable, measurable and thus, falsifiable world.

Tribalism and shared counterfactuals

I used the term “coalition-forming” in my definition above, but from casual empiricism, the word “tribal” seems to elicit much stronger responses. That might be because it gets one important point across: the purpose of shared counterfactuals is not only to reduce internal goal conflict and enable social cohesion, but also to push the internal conflict towards the boundary between groups, the fault lines between the tectonic plates of social interaction.

This might be the reason why technology standard competition, famously VHS vs Beta or Netscape vs Internet Explorer, and even many academic disputes often feel like they're led with religious fervor rather than with scientific detachment.

At which point we've come full circle.

---

An Economic Pattern Language so far:

1. Stocks, flows, transformations. the cybernetic economy.

2. The attention-seeking rational actor.

3. On organization.

4. Governance mechanisms and economic engines.

5. Value chain integrity.

6. Designing economic mechanisms: the story of Armando Ortega-Reichert.

7. The blurry boundary between economics and operations research.

8. The Bauhaus and economic design.

9. Beliefs, belief propagation and belief clusters.

10. A five-minute political economy.

11. On modeling.

12. Art, science, religion… and economics.

1. There is also the escalation of social cohesion — concurrence seeking as the dominant social goal — in the literature on groupthink instigated by Sidney Janis, which deserves its own newsletter.

2. Unlike the famous prisoner's dilemma (which also deserves its own newsletter), stag hunt, aka the hunters' dilemma, does not offer an individual reward for cheating on the group. The individual decision to hunt the hare comes about if the individual belief that everyone else will join in the stag hunt is low enough to make hunting the hare more worthwhile. Formally, the value of the hare must be as high as the value of one's share in the stag, multiplied by the subjective expectation (the belief) in a joint hunting effort, or in short H ≥ p • S. To distinguish the two patterns, we call the prisoner's dilemma a cooperation problem while stag hunt is a form of a coordination problem.

3. I'm not using the terms “high art” and “popular art” to make a quality distinction here, but an economic distinction. “High art” in my definition is produced to be scarce, yielding distinguishable unique copies that might yield high prices. “Popular art” makes use of mechanical mass reproduction to reach a wide audience. The boundary between the two is blurry of course, and it should come as no surprise that modern artists like Andy Warhol have probed that boundary in their work.

4. The numbers are abstracted from Spotify's data for 2024.