“Cultural Evolution in the Digital Age” theme session at the CESC 2017

I am organising a one-hour theme session at the inaugural conference of the Cultural Evolution Society that will take place in Jena in September. The outline programme of the conference has been just published on their website. The session – “Cultural Evolution in the Digital Age” – is scheduled for the first day, Wednesday the 13th, in the morning. Plenty of other interesting talks and events all throughout the conference (but of course I am quite biased…). Below the excellent abstracts of the three talks that will constitute the session. See you in Jena!

Predicting the replicative success of Twitter hashtags from their intrinsic properties

Maciej Pokornowski

Since its launch in 2006, Twitter has developed into one of the largest contemporary internet services in terms of user base and the volume of data generated daily. As the platform evolved, its users created a means for marking tweets that belong to the same topic: hashtags (#). Since the generation of hashtags is virtually unconstrained, alternative forms of tags emerge to mark reference to a single topic that then continue to replicate at rates that are highly and non-randomly differential. An intriguing question is whether the extent of replicative success of hashtags can be predicted from their intrinsic properties. Here we present a “longitudinal” study of the replication dynamics of Twitter hashtags related to the topic of the US Affordable Care Act (ACA). We consolidated findings from existing research and systematized them into 12 hypotheses related to the influence of hashtag-internal properties – such as length, sentiment, context-dependency, uniqueness, or topic-relevance – on various measures of replication success, or “fitness criteria”; we then proceeded to test these hypotheses on a corpus of over 4 million tweets spanning 5 years. To extract the data, we developed and employed an original method of crawling historical tweets. This eventually led to a model of hashtag fitness, which was successful at predicting the fitness of tweets established with an independent measure (hashtagify.me). We conclude by proposing a medium-independent, formal, frequency-based definition of the “fitness” of hashtags.

Inferring processes of cultural transmission: the critical role of rare variants 

Understanding how social information is used in human populations is one of the challenges in cultural evolution. Fine-grained individual-level data, detailing who learns from whom, would be most suited to answer this question empirically but this kind of data is difficult to obtain especially in pre-modern contexts. Therefore inference procedures have often been based on population-level data in form of frequency distributions of a number of different variants of a cultural trait at a certain point in time or of time-series that describe the dynamics of the frequency change of cultural variants over time, often comprising sparse samples from the whole population. In this talk we demonstrate that there exist theoretical limits to the accuracy of the inference of underlying processes of cultural transmission from aggregated data highlighting the problem of equifinality especially in situations of sparse data. Crucially we show the importance of rare variants for inferential questions. The presence, or absence, of rare variants as well as the spread behaviour of innovations carry a stronger signature about underlying processes than the dynamic of high-frequency variants. On the example of the choice of baby names, we illustrate that the consistency between empirical data, summarised by the so-called progeny, and hypotheses about cultural evolution such as neutral evolution or novelty biases depends entirely on the completeness of the data set considered. Analyses based on only the most popular variants, as is often the case in studies of cultural evolution, can provide misleading evidence for underlying processes of cultural transmission.

Lexical transformations in blogspace: a case study in short-term cultural evolution

Camille RothSébastien Lerique

Cultural Attraction Theory (CAT) introduced the notion of cultural attractors to provide a conceptual link between individual- and population-level cultural evolution processes. In this framework, attractors drive the collective convergence of representations towards specific areas despite imprecise information transmission between individuals. However, validating the existence of such attractors remains an empirical challenge which has generated a diverse number of approaches. The current deluge of digital traces offers a compelling opening in this problem and, conversely, CAT is in a good position to shed light on information propagation in online communities by introducing a more cognitively informed viewpoint on transmission behavior.

Here, we focus on the linguistic domain and study the transformation of quotations when they are copied from website to website (blogs or online media), using a large dataset of timestamped posts. By coding words with well-studied lexical features such as word frequency or age of acquisition, and inferring the most probable links between occurrences of minimally different quotations, we show that substitutions introduced by authors in those quotations are both consistent with the hypothesis of cultural attractors (words are attracted to feature-specific values) and with known lexical effects (for instance, words harder to recall in lists have a higher tendency to be substituted, whereas words easier to recall are produced instead).
Using an original “data science” approach, we more broadly demonstrate an empirical connection between psycholinguistics, digital media, and the field of cultural evolution.

Leave a Reply

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

WordPress.com Logo

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

Twitter picture

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

Facebook photo

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

Connecting to %s