How the medium shapes the message: Printing and the rise of the arts and sciences

Communication technologies, from printing to social media, affect our historical records by changing the way ideas are spread and recorded. Yet, finding statistical instruments to address the endogeneity of this relationship has been problematic. Here we use a city’s distance to Mainz as an instrument for the introduction of the printing press in European cities, together with data on nearly 50 thousand biographies, to show that cities that adopted printing earlier were more likely to be the birthplace of a famous scientist or artist in the years after the introduction of printing. At the global scale, we find that the introduction of printing is associated with a significant and discontinuous increase in the number of biographies available from people born after the introduction of printing. We bring these findings to more recent communication technologies by showing that the number of radios and televisions in a country correlates with the number of performing artists and sports players from that country that reached global fame, even after controlling for GDP, population, and including country and year fixed effects. These findings support the hypothesis that the introduction of communication technologies shift historical records in the direction of the content that is best suited for each technology.

💡 Research Summary

The paper investigates how major communication technologies—printing, radio, and television—reshape the composition of historically recorded elites in science, the arts, and sports. Using nearly 50,000 biographical entries from the Pantheon 2.0 and Human Accomplishments datasets, the authors first identify a sharp, discontinuous rise in per‑capita biographies around 1500, coinciding with the spread of Gutenberg’s movable‑type press. They then exploit a classic instrumental‑variable strategy: the geographic distance from each European city to Mainz (the birthplace of the first printing press) serves as an exogenous predictor of the timing of local press adoption. Two‑stage least‑squares regressions show that cities closer to Mainz—hence earlier adopters—produced significantly more scientists and artists born between 1450 and 1550, while the number of political or religious leaders remained unchanged. This pattern suggests that printing catalyzed a new cultural elite rather than merely amplifying existing power structures.

At the global level, changepoint analysis confirms that the printing revolution marks a structural break in the time series of notable births, reinforcing the claim that the technology altered the historical record itself.

To test whether similar dynamics hold for later media, the authors turn to country‑level data on radio and television ownership from the Historical Cross‑Country Technology Adoption (HCCTA) dataset, together with GDP and population controls and country‑year fixed effects. Panel regressions reveal that higher numbers of radios are associated with more globally recognized performing artists (actors, singers, musicians) and, to a lesser extent, sports figures, while the number of scientists shows no significant relationship. Television ownership displays an analogous pattern, with a pronounced increase in the birth of famous athletes after the 1950‑1970 television boom. These findings align with the hypothesis that each medium preferentially amplifies content that best fits its technical affordances: print favors text‑based knowledge, radio favors auditory performance, and television favors audiovisual spectacle and sport.

The study acknowledges several data limitations. The Pantheon 2.0 and Human Accomplishments collections inherit biases from Wikipedia editing communities and from the selection criteria of printed encyclopedias, respectively, meaning the results reflect the records of a literate, digitally empowered elite rather than the entire population. Nevertheless, the use of an exogenous distance‑to‑Mainz instrument and robust fixed‑effects specifications strengthens causal inference.

In sum, the research provides compelling empirical evidence that the introduction of a new communication technology does not merely increase the volume of information but systematically reshapes which types of knowledge and cultural production are recorded and celebrated. By demonstrating this effect across three distinct media epochs, the paper offers a powerful framework for anticipating how contemporary digital platforms—social media, streaming services, and AI‑generated content—might similarly bias the historical record toward the forms of expression they most efficiently disseminate.