Reblog | We have never been capitalist. Big data insights into social macro trends (1800–2000)

We think of modern societies as areligious and oriented more towards money and power. But what makes us sure that this idea is true? How can we know that secularisation, the dominance of politcs, and the primacy of money really are in the DNA of our societies?

In an earlier article published in Technological Forecasting and Social Change, we turned these and similar assumptions into research questions asking how important the different function systems of society (such as politics, economy, science, art, religion, health, education or mass media) actually have been in the 19th and 20th century.

To answer our research questions, we conducted big data research on the digital database created by the Google Books project, which has, as of 2019, digitised over 40 million books and thus created the largest books corpus worldwide. Using the Google Books Ngram Viewer, a free online graphing tool that plots annual word counts as found in the Google Books database, we investigated the English, Spanish, Russian, French, German, and Italian language areas to find out which have been the most important function systems between 1800 and 2000.

Figure 1: Word frequencies of combined religious (orange), political (blue), economic (purple), scientific (red), and mass media-related keywords (green) in the English-language Google Books corpus between 1800 and 2000. Authors Google Books Ngram Viewer query (source: Roth et al., 2017)

The results of this big data research suggested that all language areas are strongly politicised; that most language areas are secularised; that science play a strong role in many cases; and that none of the investigated language areas is dominated by the economy.

As the last result contradicts the common sense that modern societies have been capitalist societies for most of the investigated 200 years, our methodology has been criticised during conference presentations, on the social media, and in research articles.

The most pertinent methodological criticism pertained to the fact that our reliance on the Google Ngram Viewer limited our analysis to the five most frequent keywords per function system as found in the Google Books corpus.

Seeking to advance our approach and to control the risk that keyword strength can be biased due to the neglect of keyword quantity, we used an enhanced research design and a superior plotting software to scrutinise the findings of our earlier study, with the focus being on the English, French, and German language areas this time.

Figure 2: Combined occurrence frequencies of all the “function system”-specific keywords as found in the English language Google Books corpus (1800–2000) (source: Roth et al., 2019).

The results of our replication study, which have recently been published again in Technological Forecasting and Social Change, are largely consistent with those of our earlier study and thus lend credence to our sceptical assessment of the idea that modern societies have been capitalist in the 19th and 20th century.

Thus, our headline is a bit inaccurate only in that it abstracts from a period roughly between World War I and the protests of 1968, where the economy has been the second-most important function system after politics. This period, therefore, corresponds to the idea of a capitalist as a specific form of political economic society. This — temporary — reservation, however, applies only to the English language area, whereas the economy has never been more than the third most important system in the French and German language areas.

We therefore conclude that we might have indeed (almost) never been capitalist and plead for a routinisation of big data-driven checks of the modern social theories and the corresponding definitions of modern societies.


Roth S., Clark C., Trofimov N., Mkrtchyan A., Heidingsfelder M., Appignanesi L., Pérez-Valls M., Berkel J., and Kaivo-oja J. (2017) Futures of a distributed memory. A global brain wave measurement (1800–2000), Technological Forecasting and Social Change, 118, pp. 307–32.

Roth S., Schwede P., Valentinov V., Zazar K., and Kaivo-oja J. (2019), Big data insights into social macro trends (1800–2000): A replication study, Technological Forecasting and Social Change, 149, 119759.

Further reading:

Roth S., Santonen T., Heimstädt M., Clark C., Trofimov N., Kaivo-oja J., Atanesyan A., Laki B., and Sales A. (in press), Multifunctional cabinet portfolio analysis of 201 national governments, Journal of Organizational Change Management, DOI: 10.1108/JOCM-10–2018–0290

Roth S. (2015), Free economy! On 3628800 alternatives of and to capitalism, Journal of Interdisciplinary Economics, Vol. 27 No. 2, 107-128.

Tags: #bigdata #googlenram #socialtheory #systemstheory #functionaldifferentiation

Article first published on 04 November 2019 on

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.