Electronic copy available at: http://ssrn.com/abstract=2798759
* Roth, S., C. Clark, and J. Berkel (in press) The Fashionable Functions Reloaded. An Updated Google Ngram View
of Trends in Functional Differentiation. In: Mesquita, A. (Ed.) Research Paradigms and Contemporary Perspectives
on Human-Technology Interaction. Hershey: IGI-Global, forthcoming.
The Fashionable Functions Reloaded.
An Updated Google Ngram View of Trends
in Functional Differentiation (1800-2000)*
ESC Rennes School of Business, France
University of Wisconsin-La Crosse, USA
Using the updated Google Book corpus dataset generated in July 2012, we analyze the largest
available corpus of digitalized books to review social macro trends such as the secularization,
politicization, economization, and mediatization of society. These familiar trend statements are
tested through a comparative analysis of word frequency time-series plots for the English,
French, and German language area produced by means of the enhanced Google Ngram Viewer,
the online graphing tool that charts annual word counts as found in the Google Book corpus. The
results a) confirm that the importance of the political system, religion, economy, and mass media
features significant change in time and considerable regional differences and b) suggest that
visions of economized or capitalist societies are intellectual artifacts rather than appropriate
descriptions of society.
Keywords: functional differentiation; Google ngram; culturomics; capitalism; economization;
secularization; politicization; mediatization.
In this article, we draw on the second version of the Google Book corpus dataset to show how
big data analysis may check and challenge old familiar self-definitions of modern society. The
starting point of our venture is the distinction between autonomous function systems such as the
economy, science, art, religion, etc. In fact, this form of functional differentiation is considered a
core concept of modern societies (Leydesdorff, 2002; Beck et al., 2003; Berger, 2003;
Vanderstraeten, 2005; Brier, 2006; Baecker, 2007; Kjaer, 2010; Bergthaller & Schinko, 2011;
Roth, 2015a). Without functional differentiation, there would be no difference between truth and
money, a hospital would be considered the same as a bank, and there would be no sense in the
critiques of doping, corruption, or the selling of indulgences. In like manner, the larger part of
Electronic copy available at: http://ssrn.com/abstract=2798759
contemporary definitions and criticisms of modern society would have to do without their most
basic categories, since all observations of secularization, economization, and mediatization
implicitly refer to an underlying concept of functional differentiation.
Though generally accepted, the idea that particular function systems are more relevant to
society than others is not understood without ambiguity. On the one hand, in the light of the
fundamental equivalence (Vanderstraeten, 2005; Jönhill, 2012) and autonomy (Tsivacou, 2005;
Valentinov, 2012) of the function systems there is no way of arguing that the political system or
the economy is essentially more important than religion or sport, per se. On the other hand, there
seems to be plenty of empirical evidence of such imbalances in terms of the just mentioned trend
This contradiction can be resolved by stating that it is not despite, but because of their basic
equivalence that function systems can be ranked at all because if the function systems were
essentially unequal, they would already be ranked and, therefore, could no longer be ranked. In
this sense, the function systems can be treated as nominal data that feature a skewed distribution
whenever it comes to the analysis of concrete segments of society. Hence, modern societies so
far have been defined in terms of different biases to particular function systems with the most
prominent cases being the definition of society as capitalist (Roth, 2015b). Though there is still
little consensus on the question of whether capitalism results either from the primacy of a
particular form of politics or from the primacy of the economy (Risse, 2003; Wallerstein, 2003;
Foucault, 2008; Urry, 2010; Lash, 2007; Madra & Adaman, 2014), most people would basically
agree on the idea that present societies are subject to an economization of collective goals
(Alexander, 1985). This “increasing influence of economic factors and values on the political
agenda and other areas of society” (Blumler & Kavanagh, 1999:210) includes the economization
- Health (Musick, 1999; Cartier, 2003; Ewert, 2009; Brown et al, 2011),
- Art (Velthuis, 2003; Behnke, 2007; Eikhof & Haunschild, 2007; Kjaer, 2010; De Valick,
- Science (Penders et al., 2009; Berman, 2013),
- Education (Fludernik, 2005; Wilkesmann & Schmid 2012; Spring, 2015),
- Religion (Robertson, 1992; Wannenwetsch, B. 2008), and, as a matter of course,
- Society as a whole (Polanyi, 1957; Habermas & McCarthy, 1985; Schmidt, 1993;
Altvater & Mahnkopf, 1996; Enderle, 1997; Chomsky, 1999; Sayer, 1999; Finch, 2007;
Schimank & Volkmann, 2008).
As a result of this “economic turn” (Smart, 2003) or fethishization of the economy (Foucault,
2008), economization emerges so omnipresent and dominant that even the proliferation of
economics is taken for an indicator of economization (Çalışkan & Callon, 2010)1 or a need of de-
economization (Latour, 2004), respectively. Nonetheless, there is also discussion on further
forms of trend statements and predictions, which includes sometimes concurrent, sometimes
competing definitions of society as mediatized (Dennis, 1978; Eaman, 1987; Castells, 1996;
Chomsky, 1997; Blumler & Kavanagh, 1999; Croteau & Hoynes, 2003; Schulz, 2004; Hjarvard,
2008, 2013; Mazzoleni, 2008; Moon, 2012; Esser & Strömbäck, 2014), politicized (Chomsky,
2000; Blumler & Kavanagh, 1999), militarized (Regan, 1994; Young, 2007; Levy, 2010),
intellectualized (Alexander, 1985), or even aestheticized (Blumler & Kavanagh, 1999; Rocha de
Oliveira, 2009). Anything seems to go as long as there is not too much religion involved. Both the
economization (Robertson, 1992) and the politicization (Thompson, 2006) of religion seem to be
in line with the project of modernity. Any sign of religious recovery, however, apparently
understood as a threat to modernity (Martin, 2005; Martin, 2011; Bracke, 2008), even though
such a “de-secularization” would not challenge functional differentiation and modernity in a
more profound way than a possible “economization of every sphere of existence” (Kane,
In view of the literature, the long-term trends in the discussion of –izations are indeed the
politicization and, most prominently, economization, which take place against the background of
a requiem for religion and are recently complemented by an emerging media boom.
The problem with the corresponding marginalization of the “other areas of society” (Blumler
& Kavanagh, 1999:210) is not only in the fact that “(o)ther social institutions are seen (once
again) as mere puppets in the hands of powerful economic trends and actors” (Stehr, 2002:4) or
as subjects to the respective political counter-performances, but also in the empirical basis of this
discursive bias. So far, the existence of trends is supposed rather than studied. As well, even the
few contributions that call for more “observational research” (Blumler & Kavanagh, 1999:225)
are so much focused on the analysis of dominant or strong function systems that they do hardly
reflect why the respective focus system should be preferred to others and therefore lack the
overview Poul Kjaer (2010:532) is interested in the analysis of function systems:
“(T)he development of a general theory capable of linking them systematically together. When
observed in isolation, the mutual supportive character of these dimensions is not obvious. Only a
more general conceptual framework will make it possible to empirically observe to what degree
the observed phenomenon constitute or potentially will be capable of constituting a ‘higher
In order to approach such a general framework and compensate for the lack of evidence for the
respective diagnoses, the present article will consider both the general idea of biased
constellations of function systems and the observation of particular trends in the significance of
individual function systems. Both are considered hypotheses that still need to be defended.
Subsequent to a theory statement, the present article asks for the empirical validity of the
most popular statements of trends in functional differentiation. Namely the secularization,
politicization, economization and mediatization of society will therefore be re-conceptualized in
terms of hypotheses and tested against the results of a Google Ngram Viewer analysis of the
most frequent function system references in the updated second version of the Google Books
corpus for the years 1800 through 2000. The results not only show that the importance of
individual function systems changes over time and across language areas, but also give reason to
believe that popular observations of an increasing importance of the economic systems are
intellectual artifacts rather than appropriate (self-) descriptions of society.
“Ever since there has been sociological theory it has been concerned with social differentiation”
(Luhmann, 1990:423). Differentiation refers to an intrasystem process of subsystem formation
(Luhmann, 1977, 1997b). The first known forms of subsystems within society were families and
tribes. Families coexisted with other families of the same tribe in the same manner as tribes
coexisted with other tribes. As a result, early societies are said to be differentiated into identical
and co-equal segments of society. However, in the course of the Neolithic revolution, the
situation changed as soon as location decisions condensed to locational advantages and
disadvantages in such a way that individual settlements ascended towards centers, while others
turned into periphery. This center-periphery differentiation was soon complemented and
superposed by stratification, which is commonly associated with the formation of hierarchical
social orders, such as the Indian cast system or the Occidental Estates of the realm. Such
stratified societies defined persons into ranked hereditary communities and allowed for only
limited social mobility, if any at all. Conversely, in the European case, it was mobility that
finally changed the static order. Be it the movable types of Gutenberg’s printing press, the
Central-European rural exodus, or the fact that too many commoners had been granted entry into
the gentry, in all cases, the constitutive distinction of nobles and commoners was for its own part
superposed by a distinction so fundamental to modernity that animals can no longer be divided in
the following way:
“(a) belonging to the Emperor, (b) embalmed, (c) tame, (d) sucking pigs, (e) sirens, (f) fabulous,
(g) stray dogs, (h) included in the present classification, (i) frenzied, (j) innumerable, (k) drawn
with a very fine camelhair brush, (1) et cetera, (m) having just broken the water pitcher, (n) that
from a long way off look like flies” (Borges & Weinberger, 1999:231).
In fact, modern man considers it absurd to assume that a siren and a pig represent the same type
of animal as soon as they both belong to the emperor or are drawn with the same kind of brush.
“Modern society is no longer characterized by a stratification of lineage, clans and families, but
by a differentiation of function systems” (Vanderstraeten, 2005:476). Today, society therefore
cannot only be differentiated into a) similar and equal segments, b) similar and unequal centers
and peripheries, and c) dissimilar and unequal strata, but into d) dissimilar and equal function
systems, as well (see Figure 1).
Figure 1. Social Differentiation (source: Roth, 2015b:113)
Despite their obvious existence, and despite their importance to modern society, there is still
little consensus on necessary and sufficient conditions for the definition of function systems.
Nevertheless, when looking at existing working definitions and non-exhaustive lists of function
systems (Reese-Schäfer, 1999:176f, 2007:120; Künzler, 1987:327, 1989:100f; Andersen,
2003:159; Stichweh, 2005; Baecker, 1994; Henkel, 2010:183; Luhmann, 1997a:11), a list of 10
function systems can be extracted (Roth and Schütz, 2015): Political system, economy, science,
art, religion, legal system, health, sport, education, and mass media. What defines a function
system is “its function for society, which is constitutive for even the most basic legal operations
and thus for the autopoiesis of the legal system. This is why the criterion of function systems is
that they distinguish themselves by their relation to society, and not by their relation to
themselves or to any particular subsystems of society” (Roth and Schütz, 2015:18). Each of the
10 function systems differentiates society by the binary re-coding of communication according to
a specific symbolically generalized communication medium. Each function system applies only
one single code, which it also applies exclusively. For example, science, and only science, is all
about the medium truth, which is binary coded as true or untrue. Scientific programs and theories
decide on when the code of science is properly applied. The function of science is to provide
society with ongoing knowledge communication (see Tab. 1).
Table 1: The Function Systems of Society (source: Roth and Schütz, 2015:24)
While it can be considered impossible to belong to both the estate of the nobles and the estate of
the commoners or to two castes at the same time, persons can be included in all of the function
systems. This multi-inclusiveness of function system applies to organizations as well, which are,
also regarded as multi-referent (Tacke, 2001; Simsa, 2001), polyphonic (Thygesen & Andersen,
2007; Andersen & Born, 2007), or multimedia organizations (Roth et al., 2010; Roth 2012;
Combined with the basic assumption that no function system is essentially more important
than the other, the multimedia character of persons, organizations, and further segments of
society are the basis of the observation of individual differences in the frequencies of function
system references. It is precisely because of the mutual exclusiveness and functional equivalence
of the individual function systems that individual persons, organizations, nation states, or cultural
areas are likely to feature different frequency distributions of particular function system
(p)references. “On this background, it is possible to understand asymmetries, crowding-out
effects, and negative externalities between functionally differentiated spheres as a central source
of tension and conflict in modern society” (Kjaer, 2010:494). The present article will explore
these tension zones and, in doing so, regard society as a system of accessible communications. In
this sense, the article takes the language border as a better distinction than national, i.e.
geopolitical borders, and will refer to the English, French, and German language areas as
The basic assumption of the functional equivalent and mutually exclusive nature of function
systems represents an excellent groundwork for the null hypothesis, which the present article
proposes to test. Representing coequal nominal data, function systems can be assumed equally
relevant to social systems. The null hypothesis is, therefore, as follows:
(H0) Function systems relevancies exhibit a uniform distribution in social system.
On the other hand, empirical research seems to indicate an unequal distribution of function
system relevancies. The alternative hypothesis reads as follows:
(H1) Function systems relevancies exhibit an unequal distribution both in the course of time
(H1.1) and across cultures (H1.2).
Pursuing the alternative hypothesis and linking to the discourse on major trends in functional
differentiation, the article, furthermore, considers the assumed unequal distributions subject to
constant change similar to “changes in economic cycles and international competition [that]
create preferences for certain kinds of language and explanations” (Cornelissen and Kafouros,
2008:14). The second alternative hypothesis, thus, suggests testing the most popular common
senses on trends in functional differentiation:
(H2) Societies feature relatively stable trends in functional differentiation, including (H2.1) the
secularization, (H2.2) the politicization, (H2.3) the economization, and (H2.4) the mediatization
The second hypothesis will be tested against both weak and strong definitions of the respective
trends. Trends will, consequently, be analyzed with regard to their incidence in one or several
cultural areas as well as to their individual course and their relative importance.
The key assumptions proposed in this article are the temporal change of and the intercultural
differences in the importance of particular function systems to particular societies. The major
problem becomes how to measure this importance in a context that can be expected to allow for
somewhat representative information on societies as a whole.
The importance of concepts is often defined in terms of the frequency of their occurrence in
given corpora, which is considered “the simplest and most impartial gauge of word importance”
(Kloumann et al., 2012:1) or the popularity of objects, ideas, and persons (Ophir, 2010;
Bohannon, 2011).2 Given the scope of the present research question and the corresponding scale
of the research field, research utilizing the Internet as the largest possible corpus would make
sense. Despite the fact that web search engines are said to return word frequency estimates that
are highly consistent with established methods (Blair et al., 2002), the problem with Internet
word frequency analyses is that the analysis of this most recent media hardly allows for long-
term trend analyses. Hence, the present discussion of popular trends in functional differentiation
analyses changes the frequency distributions of function system references in the second largest
available text data collection.
Since its start in 2004, the Google Books project has digitalized some 15 million of the
estimated 115 million books ever published. In 2007, a Harvard research team (Michel et al.,
2011) recognized the research potential of the Google Books corpus, performed considerable
quality checks, and finally reported the compilation of a representative corpus of more than five
million books or 500 billion words covering seven language areas and a time span of 600 years.
This corpus was later updated to version 2, issued in July 2012. The development of this
enormous dataset soon raised hopes of a golden age of digital humanities (Johnson, 2010), which
would open up new types of historical knowledge (Ophir, 2010); it has already given birth to the
discipline of culturomics as “the application of high-throughput data collection and analysis to
the study of human culture” (Michel et al., 2011:181). Most culturomic research addresses issues
of literature, language, and culture studies (e.g. Gibbs & Cohen, 2011; Nicholson, 2012). Yet,
there also have been attempts to bring culturomics to the political sciences, for example, by the
use of the Google Book corpus for a retroactive forecasting of social movements such as the
Arab Spring (Leetaru, 2011).
Figure 2. Word Frequency Shares of Deutschland, England, Frankreich, and USA in the 2009
and the 2012 German Language Google Books Corpus (source: own Google Ngram enquiries)
The access to the Google Books corpus is facilitated by the Google Ngram Viewer — an open-
access interface that allows for trending (Manovich, 2012) in terms of the production of
customized time-series plots for entered search terms.
In such way, the Google Ngram Viewer can be used to re-present well-known information in
a compact and intuitive way (see Figure 2).
In contrasting the two charts combined in figure 2, we find that the patterns of the graphs
remain relatively stable between the 2009 and the 2012 versions of the dataset; yet, there are
visible differences, too, for example with regard to the performances of the USA and France in
the 1990s. Moreover, figure 2 clearly illustrates how the importance of the concepts of Germany,
England, France, and the USA varied between the year 1800 and 2000. The graph shows that it
was not until the eve of World War I that the concept of Germany became dominant in German
books. Furthermore, with regard to Germany, we find top peaks of popularity during the early
Nazi era, the Wirtschaftswunder, and the German reunification as well as interim lows after the
lost World Wars and in relation to the Protests of 1968. The overall trend, nonetheless, is
positive, which is not the case for France and England. Both countries lost relevance after World
War II and were overtaken by the USA in the mid-1980s. In this sense, the Google Ngram
Viewer also allows for the qualitative analysis of quantitative data, e.g. in terms of the
assumption that the USA became more important to Germany and perhaps was the most
important occupying power.
Furthermore, the Viewer can be used to detect needs for research as much as for the developing
of research questions (see Figure 3-4).
Figure 3. Word Frequency Shares of money, power, and love in the Overall English Language
Google Books Corpus 2012 (source: own Google Ngram enquiry)
Figure 4. Word Frequency Shares of Geld, Macht, and Liebe (money, power, and love) in the
German Language Google Books Corpus 2012 (source: own Google Ngram enquiry)
Contrasting the German and the English distribution of the concepts of money (blue), power
(red), and love (green) we find that love and money started to feature common characteristics in
the English context as of the late 1940s, while love seems to be closer to power than to money in
the German context, if at all. Far from representing proof for anything, the presented evidence
can, nonetheless, be considered a reason for further research. One big advantage of the updated
version of the Google Ngram Viewer is that the tool does to some extent allow for word context
analyses now. This is useful in the case of power, since it helps with identifying the frequencies
of non-functional meanings such as “electric power” and – if need be – with their exclusion from
the sample (which, however, is not required in the present case because of the only marginal
performances of non-political meanings of power). Again, a comparison of the present figures 3
and 4 with those in the original article (Roth, 2014) shows that the graphs perform almost
identical in the English language case, whereas there are considerable, yet not fundamental,
differences in the German language case.
In some cases, a Google Ngram View might even be used for validity tests of more or less
prominent hypotheses (see Figure 5-7).
Figure 5. Word Frequency Shares of globali(z/s)ation and regionali(z/s)ation in the Overall
English Language Google Books Corpus 2012 (source: own Google Ngram enquiry)
Figure 6. Word Frequency Shares of globalisation and régionalisation in the French Language
Google Books Corpus 2012 (source: own Google Ngram enquiry)
Figure 7. Word Frequency Shares of Globalisierung and Regionalisierung in the German
Language Google Books Corpus 2012 (source: own Google Ngram enquiry)
For example, regionalization is sometimes said to be an effect of globalization (Kacowicz, 1999;
Amin, 1999; Hurrell, 2007). The Google Ngram View of the respective concepts, however, does
not support this idea. Rather, we find evidence for the opposite effect, being that in all the three
language areas, the concept of regionalization enjoyed a modest popularity before globalization
boomed, which is most evident in the French case, which featured a noticeable trend in
regionalization between 1960 and 1975. It is not until the mid-1980s (US-American English) or
the early 1990s (British English, French, German), respectively, that the globalization discourse
reached the level of the respective regionalization discourses. Moreover, the findings suggest that
the globalization discourse could have had its origins in the French discourse and the largest
impact on the German discourse. In all these cases, the graphs feature similar patterns in both
2009 and 2012; what seems to have changed are only the absolute word frequencies (see e.g. the
Testing its own hypotheses, the present article will focus on the analysis of trends in the 2012
versions of the English, French, and German Google Books 1-gram corpora 1800-2000.3
In this setting, the null hypothesis will be rejected in favor of the alternative hypothesis (H1)
when the time-series plots produced by the Google Ngram Viewer feature an unequal
distribution of the occurrence frequencies of all function system designations. Moreover, the
plots will be analyzed with regard to variations of occurrences in the course of time (H1.1) and
across cultures (H1.2).
The hypotheses H2.1-2.4 will be weakly confirmed if the output of the Google Ngram
Viewer query supports the respective trend assumptions in at least two out of three language
areas. In concrete terms, the plots should display a decrease in the importance of religion (H2.1:
secularization) as well as an increase in the importance of the political system (H2.2:
politicization), the economy (H2.3: economization), and the media (H2.4: mediatization),
respectively. A strong confirmation would, moreover, call for (the loss of) a dominant position in
the texture of function system references for all language areas.
Meeting concerns that the function system designations might not represent the most
frequent function system reference in the corpus, the Top 2000 1-grams of the English-language
Google book corpus version 2012 will be scanned for further function system references. The list
of the ten most frequent references to function systems relevant to the hypotheses H2.1-2.4 will
also be entered into the Google Ngram Viewer and the result compared to the outcome of the
query for the names of the function systems.
The ten function systems exhibit unequal occurrence frequencies that vary within and across all
the language areas:
Figure 8. Word Frequency Shares of the Names of the Function Systems in the Overall English
Language Google Books Corpus 2012 (source: own Google Ngram enquiry)
In the English case (see Figure 8), at the beginning of the sample period, religion (not the legal
system as erroneously reported in the original version) is the most dominant function system
followed by art and the political system. A major change to the earlier version of the article is
that we replaced policy by political, which is a stronger and more adequate indicator of the
importance of the political system. The performance of the political system towards the end of
the sample period therefore appears even stronger than in the earlier version. The general picture,
however, remains the same: the political system is dominant, followed by health, which
overtakes education soon before the millennium. Currently, religion ranks ninth, outperformed
even by the formerly marginal economy and mass media system. The performance of the
economy is relatively less pronounced in the 2012 data as compared to the 2009 data.
A closer examination of the function systems relevant to the trend hypotheses 2.1-2.4 calls
attention to the word frequency shares of the political system, the economy, science, religion,
and the mass media system. The decline of religion (ȓ = 0.27/compared to 0.25 as visually
assessed in the earlier version of this article)4 and the rise of the political system (ȓ = 3.5/3.7) are
the most striking trends. Religion is outperformed by the political system during World War I
and by science in the early 1930s, with the latter being a result of slow, but steady growth. At
about the same point in time, the formerly marginal economy takes off (ȓ = 3.57/5.00 between
1930 and 1990). Yet, in 2000 economy and science share about the same numbers of mentions in
the Google Books corpus (approximately 100ppm)5 and are, therefore, both lower-middle
ranking function systems: The political system (320ppm), health (177ppm) and education
(168ppm), and art (119ppm) occupy the first five ranks. On the verge of the millennium, religion
(64ppm) was finally outpaced by the mass media (75ppm). Sport is last (16ppm).
In the French case (see Figure 9, next page), art initially ranks first in front of religion,
justice, and the political system. Between World War I and World War II, it was overtaken by
the political system, whose increase (ȓ = 5.3/4.9) to a frequency of 701ppm in 2000 is remarkable
(the small counter-trend around 1980 is less pronounced and stable than in the 2009 data). The
political system is ranked first uncontestably since the end of the 1920s. In 2000, the political
system occurred three times more often than art. As of the late 1920s, after an increase of
economic communication, the economy ranks third, closely followed by science, education, the
legal system, health, and religion. Mass media and sport come last.
A comparison of the English and the French data uncovers the higher relevance the French
function systems have in their corpus. The word politique, however, must be considered an
exception to this rule because it is both a noun and an adjective. Complemented by the
adjective(s), the English and the German political performance would be about the same as the
French. Still, it is notable that even politique_NOUN, the string used to count only cases in
which the ngram politique appears as noun, remains the clear number one in the French corpus in
the year 2000 (320ppm). Additionally, despite a considerable decline, second-ranked art
(211ppm) is still more important to the French corpus than second and third ranked health and
education are to the English.
Figure 9. Word Frequency Shares of the Names of the Function Systems in the French Language
Google Books Corpus 2012 (source: own Google Ngram enquiry)
The German data display the largest variances and the broadest scope of changes within the three
language areas. In the early 19th century, Germany seems ‘governed’ by the legal system as well
as science, art, and religion. The latter triad appears to share a common destiny in terms of their
collective decline up until the 1960s, when science and art finally separated from the downtrend
of religion, which appears to have stopped since the 1980s. After a most notable increase (ȓ =
7.9/7.0), since the early 1960s, Politik (290ppm) has ranked first, followed by the legal system
(249ppm) as well as art and science (141-161ppm). Economy (132ppm) is ranked fifth due to an
increase in importance dating back to the late 19th century rather than to the late 1920s, like in
the case of both the English and the French.
Figure 10. Word Frequency Shares of the Names of the Function Systems in the German
Language Google Books Corpus 2012 (source: own Google Ngram enquiry)
However, there is also a media trend, which started in the mid-1960s, whose calculation in terms
of a growth factor does not make much sense because, just like in the case of the English and the
French, the media seems to appear from nowhere. In the German case, the same applies to the
late 19th century rise of the economy. At least a comparison with the political curve seems
revealing to some extent, as both curves feature a certain parallelism between World War I and
the mid-1950s (which, however, appears clearer in the 2009 data).
So far, the findings indicate both considerable changes within and significant differences
between the language areas. Moreover, with regard to all language areas, the results seem to
support the hypotheses H2.1 (secularization) and H2.2 (politicization), even if there is some
evidence for the fact that the trends have stopped towards the end of the sample period. Despite
considerable increases in the frequencies of all the three designations of the economic system
and the mass media system, both systems are still far from being dominant in their respective
language areas (see Figure 11). So far, the results would suggest rejecting the hypotheses H2.3
(economization) and 2.4 (mediatization), at least in their strong variant.
Figure 11. Interfunctional Comparative Profiling of the English, French, and German Language
In order to re-test the hypotheses, in particular H2.3 and 2.4, the present article presents the
results of an additional analysis.6 We used a Phython program coded by co-author Jan Berkel to
extract a word frequency list of the entire English-language Google Ngram corpus data from
2012.7 For each of the function systems relevant to the hypotheses H2.1-2.4, the ten most
frequent references in this list were manually identified and then entered into the Google Ngram
Viewer to analyze their performance in the general English corpus using the Google Ngram
Viewer (see. Figure 12-15).
Just like the performance of the concept itself, the broader view of religion also shows a
downtrend of religion that is somehow moderated as of the second half of the 20th century (see
Figure 12). The results are in line with the previous multi-language analysis. Contrasting the
trend of the religious system and the concept of secularization, the finding is that both graphs
move against each other, which is first of all true for God rather than for the other key concepts
of the religious system, though the downtrend is featured by all of the ten most important
Figure 12. Broadband Trending the Religious System against Secularization in the English
Language Google Books Corpus 2012 (source: own Google Ngram enquiry)
The newly identified political references display a stable trend (see Figure 13). The most
important political concept, power, is losing importance between 1800-2000, whereas the term
political is becoming increasingly important in the 20th century. The overall results mean a
significant relativization of the outcome of the previous analysis: The upswing of the concept of
politicization is not attended by signs of an increasing importance of the political system, which
nonetheless appears to be the most dominant function system given the very high word
frequencies of the political keywords (ranging from near 100ppm to 570ppm).
Figure 13. Broadband Trending the Political System against Politicization in the English
Language Google Books Corpus 2012 (source: own Google Ngram enquiry)
Except for the term economic, which starts to rise at about the same time as the German word for
economy, the economic frequencies seem to be stagnating or regressing since a sometimes rise in
the first half of the 20th century (see Figure 14). Obviously, not starting until the early 1980s, the
larger-scale discovery of an economization or commodification of the society might, therefore,
have been late, wrong, or both.
Figure 14. Broadband Trending the Economy against Economization and Commodification in
the English Language Google Books Corpus 2012 (source: own Google Ngram enquiry)
As for the mass media system, many keywords feature an upward trend particularly in the second
half of the 20th century (see Figure 15). Only “pp.”, the abbreviation for pages, displays a
downtrend after 1980, thus maybe indicating a changing referencing culture. Apart from this, the
findings reinforce the earlier results and add to a greater picture of the evolution of the media
system. As a side note, we also find that the lower section of Figure 15 clearly corresponds to the
change of the dominant meaning of the concept mediatization, which originally referred to a
political reorganization of the Holy Roman Empire and now is used to speculate about socio-
cultural change induced by an increasing influence of the mass media system.
Figure 15. Broadband Trending the Mass Media System against Mediatization in the English
Language Google Books Corpus 2012 (source: own Google Ngram enquiry)
Looking at the word frequency shares of each ngram-bundle, again, we find that the economy
can hardly be considered more relevant than secularized religion or even the emerging mass
media system. The still-dominant political terms achieve twice the scores of the economic terms.
The first striking finding of the presented analyses is that the Google Ngram Views of functional
differentiation actually react to history, which is true with regard to both the interregional and the
regional level. On the one hand, all language areas display a decline of religion as well as clear
evidence of the impact of Black Friday on the takeoff of the political system and the economy.
On the other hand, the results also reflect particularities related to the histories of the individual
languages areas: The German time-series plot (see Figure 10) displays a veritable program
change for an entire language area from a realm of priests, poets, and philosophers to a national
economy. This politico-economic campaign started soon after the death of Karl Marx, which is
considerably earlier than in the other two cases. In the English and French language area, it was
not until Black Friday that the economy became at least moderately relevant. In the English case,
it seems to be the two World Wars, the Cold War, and de-colonization that led to an uptrend of
the political system, which again stopped and moderately declined around GATT Uruguay and
Perestroika. In the French case, the plot even seems to tell the story of the beginning and end of
the Fourth Republic as a struggle of the political and the legal systems8.
The results, therefore, not only support the hypothesis that modern societies feature an
uneven distribution of function system references (H1), but also show that these distributions
themselves are subject to changes in time and across cultural borders. The hypotheses H1.1 and
H1.2 are, hence, supported by the results.
Both the word frequency plots of the names of the function systems (see Figure 8-10) and the
broadband trending of the ten most frequent religious words in the Google Books corpus (see
Figure 11) support the secularization hypothesis (H2.1). The dethroning of both the
denomination and the most frequent religious terms is evident throughout and across the entire
sample(s). Referring to both a state and a trend, the secularization hypothesis can, as a result, be
corroborated both in the weak and the strong variant. While the first method also supports the
idea of a politicization of the three societies, the politicization hypotheses (H2.2) must be
relativized in view of the results of the second. In this sense, the hypotheses can only be
defended because of the dominant position the political system (still) takes up in terms of word
frequency shares in all of the three societies. In this sense, politicization is a state description
rather than a trend statement. As a result, H2.2 is corroborated in its weak variant.
Despite the remarkable growth figures the economic word frequency shares feature in all of
the three corpora, the results of both methods suggest rejecting the economization hypothesis
(H2.3) in both the strong and the weak variant. In fact, in two cases, the period of the potential
trend was rather short (E: 1930-1990; F: 1925-1975). The trends are stopped in all of the three
language areas, in none of which the economy ever reached a dominant position throughout the
entire 200 years. The only economic term that displays a potentially ongoing uptrend is the term
business (see Figure 14). Even this statement, however, has to be relativized because it is true
only in the English case.
The mediatization of society (H2.4), in return, however, seems to be an actual trend. Starting
in the 1940s (English), 1960s (German), and 1980s (French), the terms media, Medien, and
médias feature a modest, but constant uptrend. On a larger scale, the results are also consistent
with the analysis of the function system denominations. While the media system never reaches a
dominant position in either of the language areas, the trend remains uninterrupted throughout the
entire sample period. Even books are still popular in books. The mediatization hypothesis is, as a
result, weakly corroborated by the present Google Ngram views of functional differentiation.
LIMITATIONS AND FUTURE RESEARCH QUESTIONS
One of the most serious limitations of the first version Google Ngram Viewer was that it did not
allow for case-sensitive queries in ngrams. Ngrams also could not be bundled into one single
graph, just as the queries could not be for lemmas.9 Today, the enhanced version of the Google
Ngram Viewer can be used to bundle individual ngrams into one graph, in which case, however,
the also updated option of case-sensitive research does not apply. Furthermore, the enthusiasm
for analyzing combined ngram graphs is limited by the fact that a query may apparently not
comprise more than round about 30 ngrams. Still, even in this limited form, the option to
compare is attractive (see Figure 16):
Figure 16. Word Frequency Shares of Selected Ngram Bundles in the English Language Google
Books Corpus 2012 (source: own Google Ngram enquiry)
In order to create Figure 16, we entered the following string into the Google Ngram form:
In looking at the performances of the combined keywords relevant to our hypotheses, we find
that today’s dominant position of the political system is a side-effect of secularization rather than
an indicator of an increasing importance of political issues. Furthermore, we see that the
economy is far from being a dominant factor. Even within this reduced sample of four out of ten
function systems, it is third to the mass media system. This finding contradicts intuitions
according to which our lives are dominated by the economic principle(s).
Figure 15 is a fine indicator that further combined ngram queries may well lead to extremely
insightful results. Still, however, future research might also be interested in using or developing
interfaces that allow for a visualization of trends of keyword bundles comprising more than 30
ngrams. In fact, such tools would be needed to create a compelling chart comparing the
performances of robust bundles of 5-10 keywords of all of the 10 function systems of society.
Further critique may come into play with the fact that studying the key media of the
Gutenberg Galaxy might not be the key to the emerging Internet society. However, there is
evidence that online and offline contents do not differ all that much (Stern, 2004). At the same
time, the advantages of a books corpus are evident: Book content can be considered subject to
stricter selection and, therefore, a better indicator for importance. Moreover, the Internet corpus
does allow for research within the sample period of several hundred years. Further research
limits are, as a matter of course, related to restricted language competences and the corpus data
size. In fact, it would be best to compare all language areas and analyze 2- or more-grams, e.g.
with regard to word co-occurrences, in order to get much deeper context-information. Future
research on “Big Data” (Boyd & Crawford, 2012) in general and the present topic in particular
could, therefore, call upon international cooperation and access to more powerful computer
Future expeditions in the corpus might then not only open up further and more snapshots of
trends in functional differentiation, but also allow for an answer to the question of whether or not
functional differentiation is indeed a master trend in present societies. In this sense, an extended
Google Ngram View of modern societies could inform on the actual state and trend of modernity
The English, French, and German societies actually display politicization, at least in terms of a
prevailing, though not trendy, high weight of political communication. A recently re-enforced
mediatization can also be observed in all the three cases, albeit as a trend rather than a state
description. Secularization seems to characterize both a state (as religion, actually, is not
dominant anymore) and a trend, which, however, might have stopped during the last years of the
sample period (George, 2005).
The biggest surprise resulting from the presented research certainly is that an analysis of the
largest available text corpus does not corroborate the perhaps most prominent state description
and trend prediction related to functional differentiation: The data does not point to an
economized or capitalist society. Starting as a trend in line with the rising popularity of the early
socialist movement, the “economization of society” never resulted in a high weight of the
economy and stopped rather early (German: 1950ies, French: 1970ies, English: 1990ies).
If it is true that economists “perform” economies, then the presented results also support the
idea that economization critics perform economization and that both parties jointly perform an
economy bias in social sciences that is incongruent with the average to marginal relevance of the
research object. This issue is particularly critical in the context of foresight and futures studies,
where it is important to avoid third order risks of giving the right answers to the wrong questions
(Roth and Kaivo-oja, in press). Our findings therefore suggest thinking twice before we continue
considering economic preeminent issues of present and future societies. In this respect, it is also
congruent that the different dictions and notations of economization or commodification are as
marginal entries in the English language corpus as misspellings or exotic forenames are. This
fact is emphasized not to imply an only marginal relevance of research in economic risks and
benefits, but rather as suggestion to consider re-focusing research foci and drawing increased
attention to function systems beyond the politico-economic double stars of social science. Maybe
even the solution to the present “economic” “crises” is not in more, but rather in less attention to
the economy (Roth, 2015b; Roth, in press). In any case, further critiques of economization or
economic colonialism necessarily contribute to economization and are, therefore, (drivers of) the
problem they try to solve. If high weights of particular function systems are indeed a problem,
then vigilance is needed with regard to the political system rather than to the economy. Even
more light, however, could be shed on those areas over which the strong interest in (the interplay
of) the political and the economic system has casted large clouds throughout the last decades.
Apart from a certainly necessary re-cultivation of neglected landscapes of functional
differentiation, there is hardly any reason why researchers should be biased to particular function
systems and there is even less reason for resentments towards a mediocre function system or for
complicity with the most dominant, respectively.
In this sense, the present article gives us a first indication of the transformative power of
computer communication. If big data analysis actually corroborates the finding that the idea of
an economization of society is an artifact rather than a fact, then this will indeed change the face
of modern society.
Aderhold, J. (2004). Form und Funktion sozialer Netzwerke in Wirtschaft und Gesellschaft. VS
Alexander, J. C. (1985). Neofunctionalism. Thousand Oaks: Sage Publications.
Altvater. E., & Mahnkopf, B. (1996). Grenzen der Globalisierung. Ökonomie, Ökologie und
Politik in der Weltgesellschaft. Westfälisches Dampfboot: Münster.
Amin, S. (1999). Regionalization in response to polarizing globalization. Hettne, B., Inotai, A.,
& Sunkel, O. (eds). Globalism and the new regionalism. Palgrave Macmillan:
Andersen, N. Å. (2003). Polyphonic Organizations. In Hernes, T., & Bakken, T. (eds)
Autopoietic Organization Theory. CBS: Copenhagen.
Andersen, N. Å., & Born, A. W. (2007). Heterophony and the Postponed Organization
Organizing autopoietic systems. TAMARA: Journal of Critical Postmodern Organization
Science 6(1/2), 176-187.
Baecker, D. (1994). Soziale Hilfe als Funktionssystem der Gesellschaft. Zeitschrift für Soziologie
Baecker, D. (2007). Communication With Computers, or How Next Society Calls for an
Understanding of Temporal Form. Soziale Systeme 13(1-2), 407-418.
Baeza-Yates R., & Ribeiro-Neto, B. (1999). Modern information retrieval. Addison Wesley:
Beck, U., Bonss, W., & Lau, C. (2003). The Theory of Reflexive Modernization. Theory,
Culture & Society 20(2), 1-33.
Behnke, C. (2007). Corporate Art Collecting: A Survey of German-Speaking Companies. The
Journal of Arts Management, Law, and Society 37(3), 225-244.
Berger, J. (2003). Neuerliche Anfragen an die Theorie der funktionalen Differenzierung. In
Giegel, H.-J., & Schimank, U. (eds) Beobachter der Moderne. Beiträge zu Luhmanns
'Die Gesellschaft der Gesellschaft'. Suhkamp: Frankfurt am Main.
Bergthaller, H., & Schinko, C. (2011). Introduction: From National Cultures to the Semantics of
Modern Society. In Bergthaller, H., & Schinko, C. (eds). Addressing Modernity. Social
Systems Theory and U.S. Cultures. Edition Rodopi: Amsterdam and New York.
Berman, E. P. (2014). Not Just Neoliberalism: Economization in US Science and Technology
Policy. Science, Technology, and Human Values 39(3), 397-431.
Blair, I., Urland, G., & Ma, J. (2002). Using Internet search engines to estimate word frequency.
Behavior Research Methods 34(2), 286-290.
Blok, A. (2011). Clash of the eco-sciences: carbon marketization, environmental NGOs and
performativity as politics. Economy and Society 40(3), 451-476.
Blumler, J. G., & Kavanagh, D. (1999). The Third Age of Political Communication: Influences
and Features. Political Communication 16(3), 209-230.
Bohannon, J. (2011). The Science Hall of Fame. Science 331(6014), 143.
Bohn, C. (2004). Clothing as Medium of Communication. University of Lucerne. Lucerne.
Retrieved April 1, 2013, from http://www.unilu.ch/files/clothing-as_medium3.pdf
Borges, J. L., & Weinberger, E. (1999). Selected Non-Fictions: Jorge Luis Borges. Viking: New
Boyd, D., & Crawford, K. (2012). Critical Questions for Big Data. Information, Communication
& Society 15(5), 662-679.
Bracke, S. (2008). Conjugating the Modern/ Religious, Conceptualizing Female Religious
Agency. Theory, Culture & Society 25(6), 51-67.
Brier, S. (2006). Construction of Knowledge in the Mass Media. Systemic Problems in the Post-
Modern Power-Struggle between the Symbolic Generalized Media in the Agora: The
Lomborg Case of Environmental Science and Politics. Systems Research and Behavioral
Science 23(5), 667-684
Brown, N., Machin, L. & Mcleod, D. (2011). Immunitary Bioeconomy: The Economisation of
Life in the International Cord Blood Market. Social Science & Medicine 72.7, 1115-122.
Burkart, G. (2005). Die Familie in der Systemtheorie. In Runkel, G., & Burkart, G. (eds).
Funktionssysteme der Gesellschaft. Beiträge zur Systemtheorie von Niklas Luhmann. VS:
Çalışkan, K., & Callon, M. (2009). Economization, part 1: shifting attention from the economy
towards processes of economization. Economy and Society 38(3), 369-398.
Çalışkan, K., & Callon, M. (2010). Economization, part 2: a research programme for the study of
markets. Economy and Society 39(1), 1-32.
Callon, M. (1998). Introduction: The embeddedness of economic markets in economics. In
Callon, M. et al. (eds). The Laws of the Markets. Blackwell: Oxford.
Callon, M. (2007). An Essay on the Growing Contribution of Economic Markets to the
Proliferation of the Social. Theory, Culture & Society 24(7-8), 139-163.
Cartier, C. (2003). From home to hospital and back again: economic restructuring, end of life,
and the gendered problems of place-switching health services. Social Science & Medicine
Castells, M. (1996). Rise of The Network Society. Blackwell Publishers: Cambridge.
Chomsky, N. (1997). Media Control. The Spectacular Achievements of Propaganda. Seven
Stories Press: New York.
Chomsky, N. (1999). Profit over People: Neoliberalism and Global Order. Seven Stories Press:
Chomsky, N. (2000). Rogue States: The Rule of Force in World Affairs. South End Press:
Cornelissen, J. P., & Kafouros, M. (2008). The Emergent Organization: Primary and Complex
Metaphors in Theorizing about Organizations. Organization Studies 29(7), 957-978.
Croteau, D., & Hoynes, W. (2003). Media Society: Industries, Images and Audiences. Sage:
Davies, M. (2011-). Google Books (American English) Corpus (155 billion words, 1810-2009).
Available online at http://googlebooks.byu.edu/.
Dennis, E. E. (1978). The media society: evidence about mass communication in America. W. C.
De Valick, M. (2014). Film Festivals, Bourdieu, and the Economization of Culture. Revue
Canadienne D'études Cinématographiques / Canadian Journal of Film Studies, 23(1),
Eaman, R. A. (1987). The media society: basic issues and controversies. Butterworth: London.
Eikhof, D. R., & Haunschild, A. (2007). For Art's Sake! Artistic and Economic Logics in
Creative Production. Journal of Organizational Behavior, 28(5), 523–538.
Enderle, G. (1997). A Worldwide Survey of Business Ethics in the 1990s. Journal of Business
Ethics 16(14), 1475-1483.
Esser F. (ed) & Strömbäck, J. (ed.) (2014). Mediatization of Politics: Understanding the
Transformation of Western Democracies. Palgrave Macmillan: Basingstoke.
Ewert, B. (2009). Economization and Marketization in the German Healthcare System: How Do
Users Respond? German Policy Studies/Politikfeldanalyse 5.1 (2009): 21-44.
Finch, J. H. (2007). Economic sociology as a strange other to both sociology and economics.
History of the Human Sciences 20(2), 123-140.
Fludernik, M. (2005). Threatening the University—The Liberal Arts and the Economization of
Culture. New Literary History 36(1), 57-70.
Foucault, M. (2008). The Birth of Biopolitics: Lectures at the College de France 1978–1979.
Palgrave Macmillan: Houndsmill.
Fuchs, P. (1999). Intervention und Erfahrung. Suhrkamp: Frankfurt am Main.
Fuchs, P. (2000). Form und Funktion von Beratung, Soziale Systeme 6(2), 349-368.
George, S. E. (2005). Believe It or Not: Virtual Religion in the 21st Century. International
Journal of Technology and Human Interaction 1(1), 62-71.
Gibbs, F. W., & Cohen, D. J. (2011). A conversation with data: Prospecting Victorian words and
ideas. Victorian Studies, 54(1), 69-77.
Habermas, J., & McCarthy, T. (1985). The Theory of Communicative Action, Vol. 2: Lifeworld
and System: A Critique of Functionalist Reason. Beacon Press: Boston.
Henkel, A. (2010). Systemtheoretische Methodologie: Beobachtung mit Systemreferenz
Gesellschaft. In John, R., Henkel, A., & Rückert-John, J. (eds). Die Methodologien des
Systems. VS: Wiesbaden.
Hjarvard, S. (2008). The Mediatization of Society. Nordicom review 29(2), 105-134.
Hjarvard, S. (2013). The Mediatization of Culture and Society. Routledge: New York.
Hurrell, A. (2007). One World? Many Worlds? The Place of Regions in the Study of
International Society. International Affairs, 83(1), 127–146.
Jönhill, J. I. (2012). Inclusion and Exclusion - A Guiding Distinction to the Understanding of
Issues of Cultural Background. Systems Research and Behavioral Science 29(4), 387–
Johnson, C. Y. (2010). In billions of words, digital allies find tale. The Boston Globe.
Kacowicz, A. M. (1999). Regionalization, Globalization, and Nationalism: Convergent,
Divergent, or Overlapping? Alternatives: Global, Local, Political 24(4), 527-555.
Kane, C. L. (2010). Programming the Beautiful, Theory, Culture & Society 27(1), 73-93.
Kjaer, P. F. (2010). The Metamorphosis of the Functional Synthesis: A Continental European
Perspective on Governance, Law, and the Political in the Transnational Space. Wisconsin
Law Review 2, 489-1555.
Kloumann, I. M., Danforth, C. M, & Harris, K. D., et al. (2012). Positivity of the English
language. PloS one 7(1), e29484.
Künzler, J. (1987). Grundlagenprobleme der Theorie symbolisch generalisierter
Kommunikationsmedien bei Niklas Luhmann. Zeitschrift für Soziologie 16(5), 317-333.
Künzler, J. (1989). Medien und Gesellschaft. Die Medienkonzepte von Talcott Parsons, Jürgen
Habermas und Niklas Luhmann. Enke: Stuttgart.
Lash, S. (2007). Capitalism and Metaphysics. Theory, Culture & Society 24(5), 1-26.
Latour, B. (2004). Politics of Nature: How to Bring the Sciences Into Democracy. Cambridge:
Harvard University Press.
Leetaru, K. (2011). Culturomics 2.0: Forecasting large-scale human behavior using global news
media tone in time and space. First Monday, 16(9), available at:
http://www.firstmonday.dk/ojs/index.php/fm/article/view/3663/3040. Date accessed: 20
Leupold, A. (1983). Liebe und Partnerschaft: Formen der Codierung von Ehen. Zeitschrift für
Soziologie 12(4), 297-327.
Levy, Y. (2010). The Essence of the "Market Army". Public Administration Review, 70(3), 378–
Lewandowski, S. (2004). Sexualität in den Zeiten funktionaler Differenzierung. Eine
systemtheoretische Analyse. Transcript: Bielefeld.
Leydesdorff, L. (2002). The Communication Turn in the Theory of Social Systems. Systems
Research and Behavioral Science 19(2), 129-136
Luhmann, N. (1977). Differentiation of Society. The Canadian Journal of Sociology / Cahiers
canadiens de sociologie 2(1), 29-53.
Luhmann, N. (1990). The paradox of system differentiation and the evolution of society. In
Alexander, J. C., & Colomy, P. (eds). Differentiation Theory and Social Change:
Comparative and Historical Perspectives. Columbia UP: New York.
Luhmann, N. (1997a). Die Gesellschaft der Gesellschaft. Band 1 und 2. Suhrkamp: Frankfurt am
Luhmann, N. (1997b). Globalization or World society: How to conceive of modern society?
International Review of Sociology 7(1), 67-79.
Luhmann, N., & Barrett, R. (2012). Theory of Society. Palo Alto: Stanford University Press.
Maass, O. (2009). Die Soziale Arbeit als Funktionssystem der Gesellschaft. Carl Auer:
Madra, Y.A. & Adaman, F. (2014). Neoliberal Reason and Its Forms: De-Politicisation Through
Economisation. Antipode 46(3), 691-716. "
Manovich, L. (2012). Trending: The Promises and the Challenges of Big Social Data. In Gold,
M. K. (ed). Debates in the Digital Humanities. The University of Minnesota Press:
Martin, D. (2005). On Secularization: Towards a Revised General Theory. Ashgate: Surrey.
Martin, D. (2011). The Future of Christianity: Reflections on Violence and Democracy, Religion
and Secularization. Ashgate: Surrey.
Mayntz, R. (1988). Funktionelle Teilsysteme in der Theorie sozialer Systeme. In Mayntz, R.,
Rosewitz, B., & Schimank, U., et al. (eds). Differenzierung und Verselbständigung. Zur
Entwicklung gesellschaftlicher Teilsysteme. Campus: Frankfurt am Main.
Mazzoleni, G. (2008). Mediatization of society. In Donsbach, W. (ed). The International
Encyclopedia of Communication. Wiley: Hoboken.
Michel, J.-B., Shen, Y. K., & Aiden, A. P., et al. 2011. Quantitative Analysis of Culture Using
Millions of Digitized Books. Science 331(6014), 176-182.
Moon, Young Bin (2012). The Mediatized Co-Mediatizer: Anthropology in Niklas Luhmann’s
World. Zygon: Journal of Religion and Science 47(2), 438- 466.
Musick, D. W. (1999). Teaching medical ethics: A review of the literature from North American
medical schools with emphasis on education. Medicine, Health Care and Philosophy
Nicholson, B. (2012). Counting culture; or, how to read Victorian newspapers from a distance.
Journal of Victorian Culture, 17(2), 238-246.
Ophir, S. (2010). A New Type of Historical Knowledge. The Information Society 26(2), 144-
Penders, B., Verbakel, J. M. A., & Nelis, A. (2009). The Social Study of Corporate Science: A
Research Manifesto. Bulletin of Science, Technology & Society 29(6), 439-446.
Polanyi, K. (1957). The great transformation. Beacon Press: New York.
Regan, P.M. (1994). Organizing Societies for War: The Processes and Consequences of Societal
Militarization. Prager: Westport.
Reese-Schäfer, W. (1999). Luhmann zur Einführung. Junius: Hamburg.
Reese-Schäfer ,W. (2005). Die Moral der Gesellschaft: Paradigm Lost. Junius: Hamburg.
Reese-Schäfer, W. (2007). Politisches Denken heute: Zivilgesellschaft, Globalisierung und
Menschenrechte. Oldenbourg: München.
Risse, T. (2003). The Euro between national and European identity. Journal of European Public
Policy 10(4), 487-505.
Robertson, R. (1992). The Economization of Religion? Reflections on the Promise and
Limitations of the Economic Approach. Social Compass 39(1), 147-157.
Rocha de Oliveira, P. (2009). The Aestheticization of Reality. South Atlantic Quarterly 108(2),
Roth, S., Scheiber, L., & Wetzel, R. (2010). Organisation multimedial: Zum polyphonen
Programm der nächsten Organisation. Carl Auer: Heidelberg.
Roth, S. (2014a). Fashionable functions. A Google ngram view of trends in functional
differentiation (1800-2000), International Journal of Technology and Human Interaction
Roth, S. (2014b), The multifunctional organization: Two cases for a critical update for research
programs in management and organization, Tamara Journal for Critical Organization
Inquiry 12(3), 37-54.
Roth, S. (2015a), Foreword: Trends in functional differentiation, Cybernetics and Human
Knowing 22(4), 5-10.
Roth, S. (2015b). Free economy! On 3628800 alternatives of and to capitalism, Journal of
Interdisciplinary Economics 27(2), 107-128.
Roth, S. (in press). Growth and function. A viral research program for next organisations,
International Journal of Technology Management, forthcoming.
Roth, S. and Kaivo-oja, J. (in press). Is the future a political economy? Functional analysis of
three leading foresight and futures studies journals, Futures, online first October 22,
Roth, S. and Schütz, A. (2015), Ten systems. Toward a canon of function systems, Cybernetics
and Human Knowing 22(4), 11-31.
Sayer, A. (1999). Valuing culture and economy. In Ray, L., & Sayer, A. (eds). Culture and
economy after the cultural turn. Sage: London.
Scherr, A. (2001). Soziale Arbeit als organisierte Hilfe. In Tacke, V. (ed). Organisation und
gesellschaftliche Differenzierung. Westdeutscher Verlag: Opladen.
Schimank, U., & Volkmann, U. (2008). Ökonomisierung der Gesellschaft. In Maurer, A.
Handbuch der Wirtschaftssoziologie. VS: Wiesbaden
Schmidt, C. (1993). On Economization and Ecologization as Civilizing Processes.
Environmental Values 2(1), 33-46.
Schulz, W. (2004). Reconstructing mediatization as an analytical concept. European journal of
communication 19(1), 87-101.
Schwelger, R. (2008). Moralisches Handeln von Unternehmen. Eine Weiterentwicklung des
Modells und des Ansatzes der Ökonomischen Ethik auf Basis der neuen Systemtheorie
und Institutionenökonomik. Gabler: Wiesbaden.
Simsa, R. (2001). Gesellschaftliche Funktionen und Einflussformen von Nonprofit-
Organisationen. Peter Lang: Bern.
Smart, B. (2003). An Economic Turn. Journal of Classical Sociology 3(1), 47-66.
Spring, J ( 2015). The Economization of Education: Human Capital, Global Corporations, and
Skills-Based Schooling. Routledge: New York.
Stehr, N. (2002). Knowledge and Economic Conduct: The Social Foundations of the Modern
Economy. University of Toronto Press: Toronto.
Stern, S. R. (2004). Expressions of identity online: Prominent features and gender differences in
adolescents’ World Wide Web home pages. Journal of Broadcasting & Electronic Media
Stichweh R. (2005). Inklusion und Exklusion: Studien zur Gesellschaftstheorie. Transcript:
Tacke, V. (2001). Funktionale Differenzierung als Schema der Beobachtung. In Tacke, V. (ed).
Organisation und gesellschaftliche Differenzierung. Westdeutscher Verlag: Wiesbaden.
Thompson, G. F. (2006). Religious fundamentalisms, territories and ‘globalization’. Economy
and Society 36(1), 19-50.
Thygesen, N. T., & Andersen, N. Å. (2007). The polyphonic effects of technological changes in
public sector organizations: A system theoretical approach. Ephemera 7(2), 326-345.
Tsivacou, I. (2005). The Ideal of Autonomy from the Viewpoint of Functional
Differentiation/Integration of Society. Systems Research and Behavioral Science 22(6),
Tyrell, H. (1979). Familie und gesellschaftliche Differenzierung. In Pross, H. (ed). Familie -
wohin? Leistungen, Leistungsdefizite und Leistungswandlungen der Familie in
hochindustrialisierten Gesellschaften. Rowohlt: Reinbeck.
Urry, J. (2010). Consuming the Planet to Excess. Theory, Culture & Society 27(2-3), 191-212.
Valentinov, V. (2012). The Complexity-Sustainability Trade-Off in Niklas Luhmann’s Social
Systems Theory. Systems Research and Behavioral Science (Online first).
Vanderstraeten, R. (2005). System and Environment: Notes on the Autopoiesis of Modern
Society. Systems Research and Behavioral Science 22(6), 471-481.
Velthuis, O. (2003). Symbolic meanings of prices: Constructing the value of contemporary art in
Amsterdam and New York galleries. Theory and Society 32(2), 181-215.
von Rosenberg, F. (2009). Habitus und Distinktion in Peergroups: Ein Beitrag zur
rekonstruktiven Schul- und Jugendkulturforschung. Logos: Berlin.
Wagner, T. (2006). Die Soziale Arbeit der Sozialen Arbeit? – Ein kurzer Blick auf die (Selbst-)
Beobachtung eines Funktionssystems. Manuscript. Retrieved April 1, 2013, from
Wallerstein, I. (2003). Decline of American Power: The U.S. in a Chaotic World. New Press:
Wannenwetsch, B. (2008). Inwardness and Commodification: How Romanticist Hermeneutics
Prepared the Way for the Culture of Managerialism — a Theological Analysis. Studies in
Christian Ethics 21.1 (2008): 26-44.
Wilkesmann, U., & Schmid, C. J.. (2012). The impacts of new governance on teaching at
German universities. Findings from a national survey. Higher Education, 63(1), 33–52.
Young, E. S. M. (2007). A Final Period to the Union: The Militarism and Militarization of the
United States of America and its Effects on the United States Coast Guard and its
People. Race, Gender & Class, 14(3/4), 131–138.
1 The fade-out of the fact that the proliferation of a scientific discipline represents a process of
scientification rather than economization can also be observed in Anders Blok’s (2011) reply to Michel
Callon’s (Callon, 2007; Callon, 1998; Çalışkan & Callon, 2009, 2010) performativity of economics
program. Blok is right in pointing out that markets are not only performed by economists, but also by
politicians, however, he also takes economists for representatives of the economy and hopes that an
ethically guidance of both performers and counter-performers will “generate not only civilized markets,
but also civilized politics” (Blok, 2011:271), in the end.
2 The present interpretation of word importance does not refer to where word importance is inversely
related to word frequency (Baeza-Yates & Ribeiro-Neto, 1999). Such an indexical approach to word
importance would only make sense if the present article was interested in comparing the discriminatory
abilities of the concepts involved. All function systems, however, are on the same level of analysis and,
therefore, feature the same degree of discriminatory power.
3 The data from 1500-1800 is likely to feature biases due to insufficient sample sizes: "The oldest works
were published in the 1500s. The early decades are represented by only a few books per year, comprising
several hundred thousand words. By 1800, the corpus grows to 98 million words per year; by 1900, 1.8
billion; and by 2000, 11 billion” (Michel et al., 2011:176).
4 Fold change of the relative word frequency calculated in terms of the ratio of the word frequency of
2000 and 1800: ȓ = ℮(2000)/℮(1800). Due to newly added features, the word frequencies did not have to
be visually assessed anymore; rather, we could rely on precise measures as indicated by the enhanced
Google Ngram Viewer.
5 The abbreviation ppm represents “parts per million”, i.e. the relative word frequency per million words
in the Google Book corpus. The unit per million is used to avoid longer chains of digits after the decimal
point. The most common English words account for estimated 45’000ppm in version 1 and measured
46’400ppm in version 2 of the data (4.5%/4.6%: the), 30’000ppm/29’400ppm (3.9%/2.9%: of) and
23’000ppm/22’800 (2.3%/2.3%: and). Time, the most frequent noun in the English-language Google
Book corpus, has an appearance of 1’160ppm. Political (320ppm) hence appears about four times less
6 In the original version of this article (Roth, 2014), this analysis was based on the English Google Million
corpus (only, as other corpora proved unmanageable with the available hardware). First, the 10 separate
fractions of this user-friendlier corpus were merged to a 8GB SPSS file. The corpus was then transformed
into a ranked word frequency list of books published between 1800 and 2000. The 2,000 most frequent
entries of this list were, thereupon, qualitatively analyzed for their function system references. As a first
result, some 230 words with a clear function system reference were identified. The only function system
missing among the 2,000 most important words of the corpus was sport, with the designation itself as the
first reference ranking far beyond the 6,000th position.
7 The code will be made available in a work-in-progress article entitle “Futures of a shared memory. A
global brain wave measurement (1800-2000)”.
8 politique is both a noun and an adjective; the adding of adjectives like légal/e or juridique, however, has
only minor effects on the big picture.
9 A search engine for the Google Books corpus designed by Marc Davies (2011-) allows for these and
further options, however, for the American English corpus only.