Saturday, May 29, 2021

Millions of People's Location Data Revealed a 'Universal' Pattern In Study: The universal visitation law of human mobility

The universal visitation law of human mobility. Markus Schläpfer, Lei Dong, Kevin O’Keeffe, Paolo Santi, Michael Szell, Hadrien Salat, Samuel Anklesaria, Mohammad Vazifeh, Carlo Ratti & Geoffrey B. West. Nature, volume 593, pages522–527. May 26 2021.

Abstract: Human mobility impacts many aspects of a city, from its spatial structure1,2,3 to its response to an epidemic4,5,6,7. It is also ultimately key to social interactions8, innovation9,10 and productivity11. However, our quantitative understanding of the aggregate movements of individuals remains incomplete. Existing models—such as the gravity law12,13 or the radiation model14—concentrate on the purely spatial dependence of mobility flows and do not capture the varying frequencies of recurrent visits to the same locations. Here we reveal a simple and robust scaling law that captures the temporal and spatial spectrum of population movement on the basis of large-scale mobility data from diverse cities around the globe. According to this law, the number of visitors to any location decreases as the inverse square of the product of their visiting frequency and travel distance. We further show that the spatio-temporal flows to different locations give rise to prominent spatial clusters with an area distribution that follows Zipf’s law15. Finally, we build an individual mobility model based on exploration and preferential return to provide a mechanistic explanation for the discovered scaling law and the emerging spatial structure. Our findings corroborate long-standing conjectures in human geography (such as central place theory16 and Weber’s theory of emergent optimality10) and allow for predictions of recurrent flows, providing a basis for applications in urban planning, traffic engineering and the mitigation of epidemic diseases.

Popular version: Millions of People's Location Data Revealed a 'Universal' Pattern In Study

Wisdom: When thinking about others, we often take a perspective of an impartial observer, a third person viewing the events from afar; it seems a good idea to work on our issues from a distant observer perspective

Grossmann, Igor. 2021. “Wisdom: Situational, Dispositional, or Both?.” PsyArXiv. May 28. doi:10.31234/

Abstract: Some people think wisdom is a stable and invariable individual disposition. Others view wisdom as deeply embedded in culture, experiences, and situations, and treat these features as mutually making up wisdom. Who is right and what are the implications for measurement, training and the fundamental (essentialist vs. constructivist) nature of wisdom itself? In this chapter, we will review evidence concerning the dispositional versus situational approaches to study wisdom. Even though main features of wisdom show some stability, there is also a profound and systematic variability in response to situational demands. We will also learn about a novel theoretical framework, conceptualizing dispositions as a distribution of situation-specific responses, thereby resolving the dispositional versus situational debate on the nature of wisdom. Drawing on these insights, we will conclude by reflect on recommendations for best measurement practices and ways to boost and train wisdom in everyday life.

Boosting and training wisdom

One of the most exciting implications of cross-situational variability in wisdom is that we can possibly shape situations to our benefit. Here, the insight about wisdom being lower when dealing with personal issues appears troublesome: In many domains of our lives, we cannot always defer decisions to someone else. What to do? As we discussed earlier when introducing the idea of naïve realism in perception of reality, people tend to subjectively represent and construct the events they encounter 8. The notion of subjective construal can help shed possible light on the difference in wisdom when reflecting on person- and non-person-centric situations. Ethan Kross and I reasoned that the reason wisdom appeared heighted in reflection on non-personal challenges concerns a particular vantage point one adopts when construing other people’s problems compared to personal problems. When thinking about others, we often take a perspective of an impartial observer, a third person viewing the events from a far. In contrast, when we reflect on personal issues we typically do so from an immersed, first-person perspective. If this difference in the subjective vantage point is elemental for the self-other asymmetry in manifest wisdom, it may be possible to boost wisdom in reflection on personal issues by construing personal situations as an “impartial observer.”

We first sought to test this idea in the context of job prospects at the peak of "great recession" in the U.S., asking college seniors, none of whom had a secured job at this point, to consider their future career prospects 43. Participants were randomly assigned to two conditions. In one condition, we instructed participants to reflect on their job prospects from a perspective of a "distant observer," envisioning the situation unfolding from a far. In the control condition, seniors envisioned the situation unfolding before their own eyes. What we found is that compared to the control group, the “distant observer” instructions prompted greater wisdom – greater recognition of limits of their knowledge and consideration of things may unfold and change. In follow-up set of experiments, we showed equivalent results when instructing participants to reflect on a polarized political issue at a peak of 2008 U.S. presidential election 43, trust and infidelity conflicts 10, and personal autobiographical experiences – i.e., recent unresolved conflicts people experienced in their own lives 44,45. In each case, linguistic and temporal prompts promoting a distant observer vantage point (e.g., by using a third-person language “he”/”she” or perspective of “one year from now”) fostered wisdom (recognition of the limits of one's knowledge and recognition of change) in reflections on hypothetical and autobiographical issues compared to prompts promoting an immersed vantage point (e.g., by using a first-person language “me”/”mine” or a perspective of “here and now”). Moreover, using this manipulation, we were able to attenuate the self-other asymmetry discussed above. That is, observer vantage point led to greater wisdom for both personal and a friend’s problems, reducing self-other asymmetry 10; Studies 2-3. It appears that experimental instructions altered the perception of the situation—from exclusively self-focused to a situation considering viewpoints of other persons involved, in turn recreating wisdom-enhancing contexts in one's mind. Overall, it appears that a wide range of construal-altering instructions (see Figure 4) increases participants' ability for applied central features of wisdom in hypothetical and real-world situations, both in the context of interpersonal and intergroup conflicts.

fig 4

Can the distanced observer construal be trained to promote changes in wisdom over time? Building on the insights from the contextual view of wisdom, my colleagues and I decided to address this question 46. Given that people experience a range of issues in their lives, we reasoned that an effective shift in subjective construal toward a vantage point of an impartial observer requires repeated practice of wisdom-enhancing strategies over time. In turn, practice-driven shifts in subjective construal should promote greater wisdom after the practice. We tested this idea in a set of randomized control trial (RCT) intervention. In each study, participants reflected on their interpersonal conflicts twice – before and after the intervention, and we analyzed their reflections for presence of wisdom-related themes. In-between these measurement points, participants were randomly assigned to the third-person intervention condition or the first-person control condition (in the second study we also added no instruction control condition). In each condition, participants were instructed to keep a diary, each day writing a short reflection on the most significant (positive or negative) issue of the day. Based on the earlier experimental work, in the intervention condition participants had to write the diary using third-person language, reflecting on the event from an observer perspective. In the control condition(s), participants wrote their diary in a first-person, as one would typically do. Figure 5 shows the results we saw in the first study, which demonstrate that this month-long intervention impacted a range of features of wisdom, resulting in post-intervention growth in wisdom in the third-person condition compared to the control condition. These results were statistically accounted by a shift toward more inclusive subjective construal of the interpersonal conflicts participants reflected on in the experimental conditions. In the spirit of humility, it is worth highlighting that these training results are preliminary and require further replications and extensions to other cultures. None the less, they are encouraging, for the first time providing empirical support from randomized control trials for training-based shifts in wisdom over time.

fig 5

Larger is not better: No mate preference by European Common Frog (Rana temporaria) males; plus large rate of failures to mate

Larger is not better: No mate preference by European Common Frog (Rana temporaria) males. Carolin Dittrich, Oliver Roumldel. bioRxiv, May 28 2021.

Abstract: According to classical sexual selection theory, females are the choosy sex in most species. Choosiness is defined as the individual effort to invest energy and time to assess potential mates. In explosive breeding anurans, high intrasexual competition between males leads to a sexual coercion ruled mating system, where males could have evolved preferences for specific female traits. In the current study, we tested male mating preference in the explosive breeding European Common Frog without intrasexual competition. We hypothesized that males show preferences towards larger female body size in the absence of male competition. We conducted mate choice experiments, placing a male and two differently sized females in a box and recorded their mating behavior. Males did not show any preference considering female body size, neither in the attempt to grab a female nor during the formation of pairs. We witnessed a high failure rate of male mating attempts, which might make the evolution of mate choice too costly. However, small males are faster in attempting females, which could be an alternative strategy to get access to females, because their larger competitors have an advantage during scramble competition. Nonetheless, in successfully formed pairs, the females were on average larger than the males, an observation which deviated from our null-model where pairs should be of similar size if mating would be random. This indicates that selection takes place, independent from male mating preference or scramble competition.

26 Introduction

27 Research on sexual selection, exploring the mechanisms that lead to female/male mate choice and the evolution of 
28 different mating systems that facilitate non-random mating, has increased considerably in recent years (Janetos 1980; 
29 Ryan and Keddy-Hector 1992; Paul 2002; Edward and Chapman 2011). Studies addressing the theory of sexual 
30 selection revealed that females are the choosy sex in most species. This is mainly based on one assumption, the 
31 evolution of anisogamy, where males produce many small (cheap) gametes and females less but larger (expensive) 
32 gametes. Thus, females invest more energy in the production of eggs than males invest in the production of sperm 
33 (Trivers 1972). In consequence, reproduction is more costly to females and they should choose the 'fittest' male to 
34 mate with. This includes those with the best possible genes to improve her offspring’s fitness and/or those who can 
35 provide vital resources (e.g. territory, nesting place, food, parental care) to increase offspring survivability and 
36 attractiveness, thereby increasing the female´s personal fitness (Fisher 1958; Hedrick 1988; Møller and Alatalo 
37 1999). Here, choosiness is defined as an individual’s active effort to invest energy and time to assess potential mates, 
38 whereas preference is defined as an intrinsic, passive attractiveness towards specific traits of the opposite sex 
39 (Jennions and Petrie 1997; Cotton et al. 2006). However, female preferences can be overridden by dominant 
40 intrasexual competition (Qvarnström and Forsgren 1998; Härdling and Kokko 2005; Formica et al. 2016). 
41 Preferences can enhance the evolution of different mating strategies and tactics to increase reproductive output with 
42 behavioral plasticity; depending on sex, age, physiological state or operational sex ratio (Parker 1982; Gross 1996). 
43 Nevertheless, newer studies suggest that males can be choosy too, if mate availability is high and simultaneous 
44 sampling possible (Barry and Kokko 2010), if there is variation in female quality/fecundity (Krupa 1995; Johnstone 
45 et al. 1996), and if the benefits of choosing between females is higher then the costs associated with assessing 
46 females (Edward and Chapmann 2011, and references therein). Some prerequisites are the presence of males’ ability 
47 to detect differences and a preference for particular female traits. Body size can be such a trait, i.e. indicating 
48 longevity based on good genes which could be heritable (Kokko and Lindström 1996; Møller and Alatalo 1999). 
49 However, body size usually is based on a variety of genes and environmental processes, but might simply indicate 
50 higher fecundity (Peters 1986; Shine 1988; Nali et al. 2014). Mating with a larger female thus may increase a male’s 
51 individual fitness. A male’s choice however, should not only be based on such trivial correlation, it will be impacted 
52 by trade-offs concerning its mating chances, and thus individual males indeed may follow very different strategies to 
53 access females. Some examples of male tactics are satellite males, usually being smaller than their competitors (Arak 
54 1983; Halliday and Tejedo 1995), mate-guarding (Parker 1974), prudent mate choice (Fawcett and Johnstone 2003; 
55 Härdling and Kokko 2005), clutch piracy (Vieites et al. 2004) or even functional necrophilia (Izzo et al. 2012). 
56 Mating systems in amphibians are diverse, and apart from environmental parameters, mostly depend on female 
57 availability over time (Wells 2007). In frog and toad species (anurans) with long breeding periods (prolonged 
58 breeders) female mate choice seems to be the rule (Wells 1977). At any given time, a few females actively choose 
59 among many calling males, often based on call characteristics (Toledo et al. 2015; Ryan et al. 2019), the quality of 
60 defended territories, or the availability of other resources to judge the males (Howard 1978; Kirkpatrick and Ryan 
61 1991; Kokko and Jennions 2008; da Rocha et al. 2018). In lek-breeding anurans, the males aggregate in displaying 
62 arenas that do not contain any resources required by females. Females visiting these arenas 'sample' several males 
63 and choose a male to mate with (Bourne 1992). In lek-mating systems the operational sex ratio is highly skewed 
64 towards males and individual males are not able to monopolize females, leading to higher intrasexual competition 
65 (Emlen and Oring 1977). In contrast, in species with a short breeding period (explosive breeders) males are actively 
66 searching for mates and engage in direct male-male competition over the arriving females. Explosive breeding is 
67 characterized by an almost equal operational sex ratio, synchronized receptiveness of females and low sexual 
68 selection (Emlen and Oring 1977). In theory all males are able to mate and reproduce, but larger/more dominant 
69 males have an advantage to access and dominate receptive females during scramble competition leading to a 
70 variation in male mating success (Berven 1981; Olson et al. 1986; Höglund 1989; Vagi and Hettyey 2016). 
71 Therefore, some males are considered to sexually dominate the females in explosive breeding systems, leaving little 
72 room for male and female mate choice if the cost for mate sampling are too high (Dechaume-Moncharmont et al. 
73 2016). Nevertheless, male mate preferences could have evolved in explosive breeders, because female fecundity 
74 highly dependents on female body size in most anuran species (Krupa 1995; Nali et al. 2014). Simultaneous 
75 sampling of preferred females might be particular possible during the peak mating time because female availability 
76 should then be highest (Arntzen 1999; Barry and Kokko 2006). All males should prefer larger females to increase 
77 their own fitness according to adaptation theory, although preferences could be obscured by high intrasexual 
78 competition. On the other hand, costs associated with mate choice depend on male density and the frequency of 
79 different mating tactics within a breeding aggregation (Arak 1983; Höglund and Robertson 1988), as well as for 
80 instance male’s individual predation risk (Magnhagen 1991; Bernal et al. 2007), all factors which may vary already 
81 during a short breeding season (Olson et al. 1986; Vojar et al. 2015). 
82 In this study, we investigate the mating preference of the European Common Frog (Rana temporaria) because it is an 
83 excellent example of an explosive breeder with male-male competition. Although former studies suggest a lack of 
84 male mate preferences in this species (Elmberg 1991), we observed non-random mating by body size and found 
85 indications of male mate preference and different mating tactics in former experiments (Dittrich et al. 2018). Larger 
86 females were paired more frequently than smaller ones and smaller sized males showed a different mating tactic to 
87 get access to females (Dittrich et al. 2018). Here, we hypothesize that all males will prefer larger females 
88 independent of their own body size, when intrasexual competition is absent and males are presented to differently 
89 sized females. Additionally, we predict small males to be faster in attempting a female to increase their chances to 
90 keep an exclusive access to the female during scramble competition

The Cartesian Folk Theater: People believe that consciousness happens in a single, confined area (vs. multiple dispersed areas) in the human brain, and that it (partly) happens after the brain finished analyzing all available information(partly) happens after the brain finished analyzing all available information

Forstmann, Matthias, and Pascal Burgmer. 2021. “The Cartesian Folk Theater: People Conceptualize Consciousness as a Spatio-temporally Localized Process in the Human Brain.” PsyArXiv. May 28. doi:10.31234/

Abstract: The present research (total N = 2,057) tested whether people’s folk conception of consciousness aligns with the notion of a “Cartesian Theater” (Dennett, 1991). More precisely, we tested the hypotheses that people believe that consciousness happens in a single, confined area (vs. multiple dispersed areas) in the human brain, and that it (partly) happens after the brain finished analyzing all available information. Further, we investigated how these beliefs are related to participants’ neuroscientific knowledge as well as their reliance on intuition, and which rationale they use to explain their responses. Using a computer-administered drawing task, we found that participants located consciousness, but not unrelated neurological processes (Studies 1a & 1b) or unconscious thinking (Study 2) in a single, confined area in the prefrontal cortex, and that they considered most of the brain not involved in consciousness. Participants mostly relied on their intuitions when responding, and they were not affected by prior knowledge about the brain. Additionally, they considered the conscious experience of sensory stimuli to happen in a spatially more confined area than the corresponding computational analysis of these stimuli (Study 3). Furthermore, participants’ explicit beliefs about spatial and temporal localization of consciousness (i.e., consciousness happening after the computational analysis of sensory information is completed) are independent, yet positively correlated beliefs (Study 4). Using a more elaborate measure for temporal localization of conscious experience, our final study confirmed that people believe consciousness to partly happen even after information processing is done (Study 5).

Increasing Population Densities Predict Decreasing Fertility Rates over Time: A 174-nation Investigation

Rotella, Amanda M., Michael E. W. Varnum, PhD, Oliver Sng, and Igor Grossmann. 2020. “Increasing Population Densities Predict Decreasing Fertility Rates over Time: A 174-nation Investigation.” PsyArXiv. August 5. doi:10.31234/

Abstract: Fertility rates have been declining worldwide over the past fifty years, part of a phenomenon known as “the demographic transition.” Prior work suggests that this decline is related to population density. In the present study, we draw on life history theory to examine the relationship between population density and fertility across 174 countries over 69 years (1950 to 2019). We find a robust association between density and fertility over time, both within- and between-countries. That is, increases in population density are associated with declines in fertility rates, controlling for a variety of socioeconomic, socioecological, geographic, population-based, and female empowerment variables. We also tested predictions about environmental boundary conditions. In harsher living conditions (e.g., higher homicide or pathogen rates), the effect of increased population density on fertility rates was attenuated. The density-fertility association was also moderated by religiousness and strength of social norms, where the relationship between density and fertility was attenuated in countries with high religiosity and strong social norms. We discuss why and when changes in population density may influence fertility rates and the broader implications of this work.

Cultural evolution of emotional expression in 50 years of song lyrics

Cultural evolution of emotional expression in 50 years of song lyrics. Charlotte O. Brand, Alberto Acerbi, Alex Mesoudi. Evolutionary Human Sciences , Volume 1 , 2019 , e11. Nov 7 2019.

Abstract: 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.


We analysed the emotional content of song lyrics in over 160,000 songs spanning the years 1965–2015. We found that the frequency of negative words increased over time, whilst the frequency of positive words decreased over time, and asked whether these patterns could be attributed to cultural transmission biases such as success bias, prestige bias, content bias or unbiased transmission. In the billboard dataset, containing top-100 songs from 1965 to 2015, we found an effect of unbiased transmission on positive lyrics, and an effect of content bias on negative lyrics. For the larger mxm databases we only found weak effects of unbiased transmission for both negative and positive lyrics.

The effects we found in all models are extremely small. This is partly because we analysed the data on the scale of each word, negating any need for averaging over lyrics and songs. Thus, the relative increase or decrease in the log odds is understandably small. Furthermore, our implementation of transmission biases is necessarily indirect and simplified given that we lack direct observations of song lyrics being copied. It is therefore unsurprising that the effects vastly reduced or disappeared when controlling for unbiased transmission, given how many other factors must be at play in the generation of song lyrics, both directional biases such as those we explored here and random processes (Bentley et al. 2007). For example, prestige can be realised in myriad ways (Jiménez and Mesoudi 2019), particularly in the music industry. The effect of various recording companies, the extent of media attention outside of the charts and the amount of money spent on music promotion may all play a significant role in an artist's apparent prestige, and is not necessarily restricted to the content of their music. Our implementation of ‘prestige’ as predominance in the charts therefore only captures one specific aspect of musical prestige.

The effect of unbiased transmission is, however, the largest and most consistent in all of our models. This result suggests there may be an effect of random drift, or random copying, in the emotional content of song lyrics over time. This is consistent with previous work showing that random copying can explain changes in the popularity of dog breeds, baby names and popular music (Bentley et al. 2007; Hahn and Bentley 2003), as well as archaeological pottery and technological patents (Bentley et al. 2004). Thus, rather than song-writers being influenced by the most prestigious or successful artists, they may simply be influenced by the emotional content of any of the available song lyrics in the previous timestep, which may happen to increase in negativity or decrease in positivity owing to small fluctuations. As in previous work, our results do not provide evidence of literal random copying by individuals as we do not have direct access to individual's copying decisions. Instead, random drift is posed as a baseline against which to compare evidence of other copying biases. It is possible that the population-wide patterns are not a result of unanimous random copying, but owing to a multitude of idiosyncratic causes that collectively cancel each other out to create the appearance of random copying (Hoppitt and Laland 2013). In this sense, any small fluctuation in negative words owing to a particular historical event, or owing to the emergence of a more negatively biased genre, may have caused an initial increase in negative lyrics, which became exacerbated by random drift.

The presence of a content bias in the likelihood of negative lyrics occurring in the billboard songs is noteworthy. This result suggests that songs with more negative lyrics are more successful in general, perhaps reflecting either a general negativity bias (Bebbington et al. 2017; Fessler et al. 2014) or an art-specific, or music-specific, negativity bias. Similar trends favouring negative emotions vs positive ones in other artistic domains support our finding. As mentioned above, Dodds and Danforth (2010) documented a decrease in frequency of positively valenced words, and an increase in negatively valenced ones in pop song lyrics (a similar result was found in DeWall et al. 2011). Morin and Acerbi (2017) found a similar pattern in centuries of literary fiction, with a general decrease in the frequency of words denoting emotions, explained by a decrease in words denoting positive emotions, whereas the frequency of negative words remained constant. It is worth noting that we were unable to look for content bias (with our implementation) in the mxm data as there was no ranking system. One possible way of determining the popularity or use of a song could be to look at how many times, or how often, its lyrics are searched for, and whether this correlates with negative content.

In general, the idea that negative emotions would be privileged in art is consistent with the hypothesis that artistic expressions may have an adaptive function, in particular as simulation of social interactions (Mar and Oatley 2008). According to this view, developed with literary fiction in mind but potentially generalisable to other expressive forms, art would provide hypothetical scenarios where we can test and train, with no risk, our cognitive and emotional reactions. From this perspective, simulating negative events is more useful than simulating positive ones (Clasen 2017; Gottschall 2012). Art expressing negative emotions, in addition, may hold more value for audiences seeking comfort from the knowledge that others also experience negative emotions. Indeed, studies have shown that people underestimate the prevalence of others’ negative emotions, and this underestimation exacerbates loneliness and decreases life satisfaction (Jordan et al. 2011). Furthermore, suppressing rather than reappraising negative emotions decreases self-esteem and increases sadness (Nezlek and Kuppens 2008)(Nezlek & Kuppens, 2008). This hypothesis is worth investigating in future research.

Our varying effects models suggested that most of the variation lay between artists. However, genre also showed considerable variation. We were unable to control for genre in the billboard data as genre information was not available with this dataset. This could provide a partial explanation for the differing results between the billboard and mxm datasets; indeed, Dodds and Danforth (2010) attributed the decrease in emotional valence within pop song lyrics to the emergence of more negative genres such as heavy metal and punk. Future work investigating the variation of emotional expression between different genres of music would be valuable. A further limitation of this study is that we restricted our analysis to comparing the content of each song with that of the songs from the previous three years of songs. Mechanistically this suggests that songs that are currently in the charts influence song-writers who are writing within three years of chart success, assuming that the time it takes to get from the song-writing process to chart success is three years or less. It is possible that these effects are stronger or weaker at different time points, such as within one or five years of chart success. Furthermore, although we controlled for artist, many songs in the billboard charts are in fact written by specially designated song-writers, such as Max Martin.

Overall this research contributes to the growing body of work attempting to quantitatively study trends in the domain of music (Youngblood 2019; Savage 2019; Mauch et al. 2015; Ravignani et al. 2017). Our starting result of an increase in negative emotions and decrease in positive ones in song lyrics is paired with similar findings regarding acoustic qualities. Using the same Billboard top-100 songs that we analysed, Schellenberg and von Scheve (2012) found an increase in minor mode and a decrease in the average tempo, which indicates that the songs become more sad-sounding through time. This seems to be part of a longer trend in Western classical music, where the use of the minor mode increased over a 150-year period from 1750 to 1900 (Horn and Huron 2015). The relationship between minor tone and negative valence of lyrics has been also studied, and confirmed, quantitatively (Kolchinsky et al. 2017). Analogously, studying more than 500,000 songs released in the UK between 1985 and 2015, Interiano et al. (2018) found a similar decrease in ‘happiness’ and ‘brightness’, coupled with a slight increase in ‘sadness’ (these high-level features result from algorithms analysing low-level acoustic features, such as the tempo, the tonality, etc.). They also found the puzzling result that, despite a general trend towards sadder songs, the successful hits are, on average, happier than the rest of the songs. In the same way, whereas we found that the higher the position in the billboard chart the more negative a song is, billboard songs are as a whole more positive than the songs in the mxm dataset, which contains more (and less successful) songs.

In this study we used cultural evolutionary theory to try to explain patterns in one of the most pervasive of human cultural practices, music production. More specifically, we tried to detect whether any particular transmission bias best explained the changing patterns of emotional expression over time. We conclude that, although we found weak evidence of success and prestige biases, these were overwhelmed by an effect for unbiased transmission. The presence of a content bias for negative lyrics remained, and this may be a contributing factor to the increasing in negative lyrics over time. A potential explanation for these results is that a multitude of transmission biases and other causes are at play. It is likely that small shifts, for example owing to historical events or the emergence of particular genres, may have nudged the production and transmission of negative and positive lyrics in opposite directions, and random copying exacerbated this trajectory. These possibilities should be explored more in future work. Overall, the exercise of precisely analysing large datasets to explain cultural change, if refined on relatively benign cultural trends such as pop music, could eventually be more expertly applied to areas of greater societal importance and impact, such as shifts in political beliefs or moral preferences.