As a writer and someone who teaches an Aesthetics course, the cause of artistic success is a matter that I find rather interesting. When I was an undergraduate I was involved in a faculty-student debate about artificial intelligence. In the course of the debate, I defended free will. The professor on the other side made an interesting point in claiming that I believed in free will because I wanted credit for my success. That remark stuck with me and I found it applied elsewhere, such as matters of luck (that is, chance that turns out favorable or unfavorable).
Since I have been a gamer for quite some time, I am well aware of the role (or roll) of chance in success. However, as the professor noted, I wanted credit for my successes and hence while I acknowledged the role of luck, I tended to minimalize its role. However, after having some modest success with my books and teaching Aesthetics for years, I came to accept the view that luck (that is, favorable chance) has a large role in success. Of course, this was a largely unsupported view. Fortunately, Princeton’s Matthew Salganik decided to investigate the matter of success and had the resources to do so.
In order to determine the role of chance in success Salganik created nine identical online worlds. He then distributed the 30,000 teens he had recruited for his experiment among these virtual worlds. Each group of teens was exposed to the same 48 songs from emerging artists that were unknown to the teens. In return, the teens were able to download the songs they liked best free of charge.
One world was set up as the control world—in this world the teens were isolated from social influence because they could not see what songs their fellows were downloading. In the other eight worlds, they could see which songs were being downloaded—which informed them of what the other teens regarded as worth downloading.
This experiment was certainly well designed: each world is identical at the start and the test subjects (the teenagers) were randomly assigned to the worlds. Given the quality and size of the experiment, the results can be safely regarded as statistically significant.
Given that the same 48 songs were available in each world, if quality was the defining factor for success, then it would seem to follow that each world should be fairly similar in terms of which songs were downloaded the most. However, Salganik found that the worlds varied a great deal. For example, 52 Metro’s song “Lock Down” was first in one world and 40th in another world. Salganik concluded that “small, random initial differences” were magnified by “social influence and cumulative advantage.” In short, chance was the decisive factor in the outcome. As a gamer, I certainly appreciated these findings and could easily visualize modelling this process with some dice and charts—like in games such as Pathfinder and D&D.
Lest it be thought that chance is the sole factor, Salganik found that quality does have some role in success—but much less than one might suspect. Based on additional experiments, he found that succeeding with a work of poor quality is rather hard but that once a certain basic level of quality is achieved, then success is primarily a matter of chance.
In terms of the specific mechanism of artistic success, a group of people will as a matter of random chance decide that a work is good. The attention of this group will attract more attention and this process will continue. Those who are drawn by the attention seem to engage in the reasoning that the work must be good and special because all the other people seem to believe that it is good and special.
Interestingly enough, Leo Tolstoy seemed to have hit on a similar idea—although he obviously lacked the means to run the sort of experiment conducted by Salganik. As Tolstoy said, “a work that pleases a certain circle of people is accepted as art, then a definition of art is devised to cover these productions.” Tolstoy believed that approach failed to distinguish between good and bad art and thus he regarded it as flawed. With a tweak, this can be used to capture Salganik’s findings: “a work that pleases a certain circle of people is accepted as good, then it is believed by others to be special.”
Interestingly enough, the sort of “reasoning” that Salganik’s experiment seems to have shown is the Appeal to Popularity fallacy: this is the “reasoning” that because something is popular, it follows that it is good/correct. It also nicely matches the similar Bandwagon fallacy: that because something is winning, it follows that it is good/correct. Not surprisingly, this is grounded in the cognitive bias known as the Bandwagon Effect: people have a psychological tendency to align their thinking with other people. In the case of Salganik’s experiment, the participants aligned their thinking in terms of their aesthetic preference and thus created a bandwagon effect. The effect is rather like the stereotype of the avalanche: a small, random event can set off a massive tide. Given that the process of selection is essentially not a rational assessment of quality but rather driven by cognitive bias and (perhaps) fallacious reasoning it certainly makes sense that the outcomes would be decided largely by chance. The same, if his experiment extends by analogy, would seem to hold true of the larger world.