Birth of the cool: a two-centuries decline in emotional expression in Anglophone fiction

How Male and Female Literary Authors Write About Affect Across Cultures and Over Historical Periods

A wealth of literature suggests the existence of sex differences in how emotions are experienced, recognized, expressed, and regulated. However, to what extent these differences result from the put in place of stereotypes and social rules is still a matter of debate. Literature is an essential cultural institution, a transposition of the social life of people but also of their intimate affective experiences, which can serve to address questions of psychological relevance. Here, we created a large corpus of literary fiction enriched by authors' metadata to measure the extent to which culture influences how men and women write about emotion. Our results show that even though before the twenty-first century and across 116 countries women more than men have written about affect, starting from 2000, this difference has diminished substantially. Also, in the past, women's narratives were more positively laden and less arousing. While the difference in arousal is ubiquitous and still present nowadays, sex differences in valence vary as a function of culture and have dissolved in recent years. Altogether, these findings suggest that historic evolution is associated with men and women writing similarly about emotions and reveal a sizable impact of culture on the affective characteristics of the lexicon.

Supplementary Information

The online version contains supplementary material available at 10.1007/s42761-023-00219-9.

Large language models show human-like content biases in transmission chain experiments

As the use of large language models (LLMs) grows, it is important to examine whether they exhibit biases in their output. Research in cultural evolution, using transmission chain experiments, demonstrates that humans have biases to attend to, remember, and transmit some types of content over others. Here, in five preregistered experiments using material from previous studies with human participants, we use the same, transmission chain-like methodology, and find that the LLM ChatGPT-3 shows biases analogous to humans for content that is gender-stereotype-consistent, social, negative, threat-related, and biologically counterintuitive, over other content. The presence of these biases in LLM output suggests that such content is widespread in its training data and could have consequential downstream effects, by magnifying preexisting human tendencies for cognitively appealing and not necessarily informative, or valuable, content.

Using portraits to quantify the changes of generalized social trust in European history: A replication study

A portrait is an exercise of impression management: the sitter can choose the impression she or he wants to create in the eyes of others': competence, trustworthiness, dominance, etc. Indirectly, this choice informs us about the qualities that were specifically valued at the time the portrait was created. In a previous paper, we have shown that cues of perceived trustworthiness in portraits increased in time during the modern period in Europe, meaning that people probably granted more importance to be seen as a trustworthy person. Moreover, this increase is correlated to economic development. In this study, we aim to replicate this result, using more controlled databases: 1) a newly created database of European head-of-state sovereigns (N = 966, from 1400 to 2020), that is a database of individuals holding the same social position across time and countries, and 2) a database of very high-quality portraits digitized with the same technique, and coming from the same Museum, the Chateau de Versailles database (N = 2,291, from 1483 to 1938). Using mixed effects linear models, we observed in the first dataset that the modeled perceived facial trustworthiness of these sovereigns' faces increased over time (b = 0.182 ± 0.04 s.e.m., t(201) = 4.40, p < 0.001). On the opposite, no effect of time was detected on the portraits of the Château de Versailles (b = - 0.02 ± 0.03 s.e.m., t(759) = - 0.85, p > .250). We conclude by discussing the potential of this new technique to uncover long-term behavioral changes in history, as well as its limitations.

Supernatural explanations across 114 societies are more common for natural than social phenomena

Humans across the globe use supernatural beliefs to explain the world around them. This article explores whether cultural groups invoke the supernatural more to explain natural phenomena (for example, storms, disease outbreaks) or social phenomena (for example, murder, warfare). Quantitative analysis of ethnographic text across 114 geographically and culturally diverse societies found that supernatural explanations are more prevalent for natural than for social phenomena, consistent with theories that ground the origin of religious belief in a human tendency to perceive intent and agency in the natural world. Despite the dominance of supernatural explanations of natural phenomena, supernatural explanations of social phenomena were especially prevalent in urbanized societies with more socially complex and anonymous groups. Our results show how people use supernatural beliefs as explanatory tools in non-industrial societies, and how these applications vary across small-scale communities versus large and urbanized groups.

Valence-dependent mutation in lexical evolution

A central goal of linguistics is to understand how words evolve. Past research has found that macro-level factors such as frequency of word usage and population size explain the pace of lexical evolution. Here we focus on cognitive and affective factors, testing whether valence (positivity-negativity) explains lexical evolution rates. Using estimates of cognate replacement rates for 200 concepts on an Indo-European language tree spanning six to ten millennia, we find that negative valence correlates with faster cognate replacement. This association holds when controlling for frequency of use, and follow-up analyses show that it is most robust for adjectives ('dirty' versus 'clean'; 'bad' versus 'good'); it does not consistently reach statistical significance for verbs, and never reaches significance for nouns. We also present experiments showing that individuals are more likely to replace words for negative versus positive concepts. Our findings suggest that emotional valence affects micro-level guided variation, which drives macro-level valence-dependent mutation in adjectives.

Sentiment analysis of the Twitter response to Netflix's Our Planet documentary

The role of nature documentaries in shaping public attitudes and behavior toward conservation and wildlife issues is unclear. We analyzed the emotional content of over 2 million tweets related to Our Planet, a major nature documentary released on Netflix, with dictionary and rule-based automatic sentiment analysis. We also compared the sentiment associated with species mentioned in Our Planet and a set of control species with similar features but not mentioned in the documentary. Tweets were largely negative in sentiment at the time of release of the series. This effect was primarily linked to the highly skewed distributions of retweets and, in particular, to a single negatively valenced and massively retweeted tweet (>150,000 retweets). Species mentioned in Our Planet were associated with more negative sentiment than the control species, and this effect coincided with a short period following the airing of the series. Our results are consistent with a general negativity bias in cultural transmission and document the difficulty of evoking positive sentiment, on social media and elsewhere, in response to environmental problems.

Valence norms for 3,600 English words collected during the COVID-19 pandemic: Effects of age and the pandemic

The topic of affective development over the lifespan is at the forefront of psychological science. One of the intriguing findings in this area is superior emotion regulation and increased positivity in older rather than younger adults. This paper aims to contribute to the empirical base of studies on the role of affect in cognition. We report a new dataset of valence (positivity) ratings to 3,600 English words collected from North American and British English-speaking younger (below 65 years of age) and older adults (65 years of age and older) during the COVID-19 pandemic. This dataset represents a broad range of valence and a rich selection of semantic categories. Our analyses of the new data pitted against comparable pre-pandemic (2013) data from younger counterparts reveal differences in the overall distribution of valence related both to age and the psychological fallout of the pandemic. Thus, we found at the group level that older participants produced higher valence ratings overall than their younger counterparts before and especially during the pandemic. Moreover, valence ratings saw a super-linear increase after the age of 65. Together, these findings provide new evidence for emotion regulation throughout adulthood, including a novel demonstration of greater emotional resilience in older adults to the stressors of the pandemic.

How to Develop Reliable Instruments to Measure the Cultural Evolution of Preferences and Feelings in History?

While we cannot directly measure the psychological preferences of individuals, and the moral, emotional, and cognitive tendencies of people from the past, we can use cultural artifacts as a window to the zeitgeist of societies in particular historical periods. At present, an increasing number of digitized texts spanning several centuries is available for a computerized analysis. In addition, developments form historical economics have enabled increasingly precise estimations of sociodemographic realities from the past. Crossing these datasets offer a powerful tool to test how the environment changes psychology and vice versa. However, designing the appropriate proxies of relevant psychological constructs is not trivial. The gold standard to measure psychological constructs in modern texts - Linguistic Inquiry and Word Count (LIWC) - has been validated by psychometric experimentation with modern participants. However, as a tool to investigate the psychology of the past, the LIWC is limited in two main aspects: (1) it does not cover the entire range of relevant psychological dimensions and (2) the meaning, spelling, and pragmatic use of certain words depend on the historical period from which the fiction work is sampled. These LIWC limitations make the design of custom tools inevitable. However, without psychometric validation, there is uncertainty regarding what exactly is being measured. To overcome these pitfalls, we suggest several internal and external validation procedures, to be conducted prior to diachronic analyses. First, the semantic adequacy of search terms in bags-of-words approaches should be verified by training semantic vector spaces with the historical text corpus using tools like word2vec. Second, we propose factor analyses to evaluate the internal consistency between distinct bag-of-words proxying the same underlying psychological construct. Third, these proxies can be externally validated using prior knowledge on the differences between genres or other literary dimensions. Finally, while LIWC is limited in the analysis of historical documents, it can be used as a sanity check for external validation of custom measures. This procedure allows a robust estimation of psychological constructs and how they change throughout history. Together with historical economics, it also increases our power in testing the relationship between environmental change and the expression of psychological traits from the past.

From Storytelling to Facebook : Content Biases When Retelling or Sharing a Story

Cultural evolution researchers use transmission chain experiments to investigate which content is more likely to survive when transmitted from one individual to another. These experiments resemble oral storytelling, wherein individuals need to understand, memorize, and reproduce the content. However, prominent contemporary forms of cultural transmission-think an online sharing-only involve the willingness to transmit the content. Here I present two fully preregistered online experiments that explicitly investigated the differences between these two modalities of transmission. The first experiment (N = 1,080 participants) examined whether negative content, information eliciting disgust, and threat-related information were better transmitted than their neutral counterpart in a traditional transmission chain setup. The second experiment (N = 1,200 participants) used the same material, but participants were asked whether or not they would share the content in two conditions: in a large anonymous social network or with their friends, in their favorite social network. Negative content was both better transmitted in transmission chain experiments and shared more than its neutral counterpart. Threat-related information was successful in transmission chain experiments but not when sharing, and finally, information eliciting disgust was not advantaged in either. Overall, the results present a composite picture, suggesting that the interactions between the specific content and the medium of transmission are important and, possibly, that content biases are stronger when memorization and reproduction are involved in the transmission-as in oral transmission-than when they are not-as in online sharing. Negative content seems to be reliably favored in both modalities of transmission.

Semantics of European poetry is shaped by conservative forces: The relationship between poetic meter and meaning in accentual-syllabic verse

Recent advances in cultural analytics and large-scale computational studies of art, literature and film often show that long-term change in the features of artistic works happens gradually. These findings suggest that conservative forces that shape creative domains might be underestimated. To this end, we provide the first large-scale formal evidence of the association between poetic meter and semantics in 18-19th century European literatures, using Czech, German and Russian collections with additional data from English poetry and early modern Dutch songs. Our study traces this association through a series of unsupervised classifications using the abstracted semantic features of poems that are inferred for individual texts with the aid of topic modeling. Topics alone enable recognition of the meters in each observed language, as may be seen from the same-meter samples clustering together (median Adjusted Rand Index between 0.48 and 1 across traditions). In addition, this study shows that the strength of the association between form and meaning tends to decrease over time. This may reflect a shift in aesthetic conventions between the 18th and 19th centuries as individual innovation was increasingly favored in literature. Despite this decline, it remains possible to recognize semantics of the meters from past or future, which suggests the continuity in meter-meaning relationships while also revealing the historical variability of conditions across languages. This paper argues that distinct metrical forms, which are often copied in a language over centuries, also maintain long-term semantic inertia in poetry. Our findings highlight the role of the formal features of cultural items in influencing the pace and shape of cultural evolution.

From Text to Thought: How Analyzing Language Can Advance Psychological Science

Humans have been using language for millennia but have only just begun to scratch the surface of what natural language can reveal about the mind. Here we propose that language offers a unique window into psychology. After briefly summarizing the legacy of language analyses in psychological science, we show how methodological advances have made these analyses more feasible and insightful than ever before. In particular, we describe how two forms of language analysis—natural-language processing and comparative linguistics—are contributing to how we understand topics as diverse as emotion, creativity, and religion and overcoming obstacles related to statistical power and culturally diverse samples. We summarize resources for learning both of these methods and highlight the best way to combine language analysis with more traditional psychological paradigms. Applying language analysis to large-scale and cross-cultural datasets promises to provide major breakthroughs in psychological science.

Socially Situated Transmission: The Bias to Transmit Negative Information is Moderated by the Social Context

Cultural evolutionary theory has identified a range of cognitive biases that guide human social learning. Naturalistic and experimental studies indicate transmission biases favoring negative and positive information. To address these conflicting findings, the present study takes a socially situated view of information transmission, which predicts that bias expression will depend on the social context. We report a large-scale experiment (N = 425) that manipulated the social context and examined its effect on the transmission of the positive and negative information contained in a narrative text. In each social context, information was progressively lost as it was transmitted from person to person, but negative information survived better than positive information, supporting a negative transmission bias. Importantly, the negative transmission bias was moderated by the social context: Higher social connectivity weakened the bias to transmit negative information, supporting a socially situated account of information transmission. Our findings indicate that our evolved cognitive preferences can be moderated by our social goals.

Beyond social learning

Cultural evolution requires the social transmission of information. For this reason, scholars have emphasized social learning when explaining how and why culture evolves. Yet cultural evolution results from many mechanisms operating in concert. Here, we argue that the emphasis on social learning has distracted scholars from appreciating both the full range of mechanisms contributing to cultural evolution and how interactions among those mechanisms and other factors affect the output of cultural evolution. We examine understudied mechanisms and other factors and call for a more inclusive programme of investigation that probes multiple levels of the organization, spanning the neural, cognitive-behavioural and populational levels. To guide our discussion, we focus on factors involved in three core topics of cultural evolution: the emergence of culture, the emergence of cumulative cultural evolution and the design of cultural traits. Studying mechanisms across levels can add explanatory power while revealing gaps and misconceptions in our knowledge. This article is part of the theme issue 'Foundations of cultural evolution'.

The rise of prosociality in fiction preceded democratic revolutions in Early Modern Europe

The English and French Revolutions represent a turning point in history, marking the beginning of the modern rise of democracy. Recent advances in cultural evolution have put forward the idea that the early modern revolutions may be the product of a long-term psychological shift, from hierarchical and dominance-based interactions to democratic and trust-based relationships. In this study, we tested this hypothesis by analyzing theater plays during the early modern period in England and France. We found an increase in cooperation-related words over time relative to dominance-related words in both countries. Furthermore, we found that the accelerated rise of cooperation-related words preceded both the English Civil War (1642) and the French Revolution (1789). Finally, we found that rising per capita gross domestic product (GDPpc) generally led to an increase in cooperation-related words. These results highlight the likely role of long-term psychological and economic changes in explaining the rise of early modern democracies.

Psychological origins of the Industrial Revolution: More work is needed!

I am grateful to have received so many stimulating commentaries from interested colleagues regarding the psychological origins of the Industrial Revolution and the role of evolutionary theory in understanding historical phenomena. Commentators criticized, extended, and explored the implications of the perspective I presented, and I wholeheartedly agree with many commentaries that more work is needed. In this response, I thus focus on what is needed to further test the psychological origins of the Industrial Revolution. Specifically, I argue, in agreement with many commentators, that we need: (1) better data about standards of living, psychological preferences, and innovation rates (sect. R1); (2) better models to understand the role of resources (and not just mortality) in driving cultural evolution and the multiple aspects of the behavioral constellation of affluence (sect. R2); and (3) better predictions and better statistical instruments to disentangle the possible mechanisms behind the rise of innovativeness (genetic selection, rational choice, and phenotypic plasticity) (sect. R3).

Cultural evolution of emotional expression in 50 years of song lyrics

Popular music offers a rich source of data that provides insights into long-term cultural evolutionary dynamics. One major trend in popular music, as well as other cultural products such as literary fiction, is an increase over time in negatively valenced emotional content, and a decrease in positively valenced emotional content. Here we use two large datasets containing lyrics from n = 4913 and n = 159,015 pop songs respectively and spanning 1965-2015, to test whether cultural transmission biases derived from the cultural evolution literature can explain this trend towards emotional negativity. We find some evidence of content bias (negative lyrics do better in the charts), prestige bias (best-selling artists are copied) and success bias (best-selling songs are copied) in the proliferation of negative lyrics. However, the effects of prestige and success bias largely disappear when unbiased transmission is included in the models, which assumes that the occurrence of negative lyrics is predicted by their past frequency. We conclude that the proliferation of negative song lyrics may be explained partly by content bias, and partly by undirected, unbiased cultural transmission.

Guideline for improving the reliability of Google Ngram studies: Evidence from religious terms

The Google Books Ngram Viewer (Google Ngram) is a search engine that charts word frequencies from a large corpus of books and thereby allows for the examination of cultural change as it is reflected in books. While the tool's massive corpus of data (about 8 million books or 6% of all books ever published) has been used in various scientific studies, concerns about the accuracy of results have simultaneously emerged. This paper reviews the literature and serves as a guideline for improving Google Ngram studies by suggesting five methodological procedures suited to increase the reliability of results. In particular, we recommend the use of (I) different language corpora, (II) cross-checks on different corpora from the same language, (III) word inflections, (IV) synonyms, and (V) a standardization procedure that accounts for both the influx of data and unequal weights of word frequencies. Further, we outline how to combine these procedures and address the risk of potential biases arising from censorship and propaganda. As an example of the proposed procedures, we examine the cross-cultural expression of religion via religious terms for the years 1900 to 2000. Special emphasis is placed on the situation during World War II. In line with the strand of literature that emphasizes the decline of collectivistic values, our results suggest an overall decrease of religion's importance. However, religion re-gains importance during times of crisis such as World War II. By comparing the results obtained through the different methods, we illustrate that applying and particularly combining our suggested procedures increase the reliability of results and prevents authors from deriving wrong assumptions.