Research

At the core, I’m interested in how human beings and other animals develop, grow, intermingle and communicate.

This page first explains the overarching questions that drive me and the sources from which I draw inspiration. For more tangible output, jump straight to the projects I’ve developed, the publications these led to, or my education.


Questions

I would like to understand what makes you you, what makes me me, and what happens in and between us two when we spend time together, when we talk, walk, interact with one another. What happens in a group of more than two people. How what we just exchanged weaves itself in our future interactions. While the end target of this are small groups or dyads, any explanation will most likely include elements of larger groups and dynamics, since all communication takes place in communities where their patterns emerge.

This interest brings me to questions like:

(Yes, there’s circularity in wanting to understand the process of understanding, but progress on these questions can still be made!)


Epistemic sources

Some of these contradict each other, but I find them all inspiring:


Projects

2017 to now — Node-word-2-vec

Social media platforms present themselves as large networks of users communicating and interacting through written messages. Their interactions are highly structured, and linked to a variety of factors such as their socio-demographic variables, their linguistic usage patterns, the distribution of their interests, or the structure of interaction networks (either conversation- or follower-based).

To our knowledge, current approaches to describing the dynamics of such systems do not capture the dependencies between these dimensions, as each corresponds to a different type of information on nodes and edges of a network. However, linguistic usage patterns are deeply linked to socio-economic variables, and topological network structure is a fundamental and coevolving component of the spread of information in the system. So there is much to be gained by integrating all the available information to identify patterns in the system.

We focus on the relationship between language evolution and network structure, using a francophone Twitter dataset of 200M+ tweets and 2M+ users collected over a two-year period. We explore the capabilities of deep learning approaches — which have not yet been fully applied to network-structured data — to unify the different sources of information available in Twitter and explore the relationship between topology-based communities and patterns of language use. The goal is to construct an embedding of users that lets us infer correlations between linguistic variables, network structure, and socio-economic attributes.

This project is my current work focus, and you can watch progress on the github repository!

2015-2017 — Gistr

To investigate interpretation in particular contexts, and the effects it can have at the global scale such as cultural attractors, we built an online chinese whispers game where people iteratively memorise and rewrite short pieces of text, to see how they are gradually transformed as they are transmitted.

It’s presented as a Game With a Purpose, available a gistr.io. If you’re interested, head over to the project wiki! This was my first non-trivial project using Elm, and it’s been a delightful experience.

Gistr Welcome page Gistr Experiment Instructions Gistr Tree Exploration

The high-quality data coming from this experiment let us model the transformations that occur when people are asked to remember and rewrite short stories. In particular, we showed that those transformations can be described as a combination of simple operations that lead into each other and form word filiations. Breaking down complex linguistic changes into such simple operations also paves the way for better models of short-term cultural change.

The full details are in Chapter 3 of my thesis.

Transformations made to one story in the experiment
Breakdown of an individual transformation

2013-2016 — Brains copy-paste

Online news is full of quotes from politicians or other famous people. Quite often though, those quotes are transformed just a tiny little bit when they are copied from one source to another article — blog or news outlet. This process is most likely unconscious, but it’s not random: the changes made are significant, and we measured some of them with data-mining techniques. Looking at substitutions from one word to another shows that more complex words tend to be replaced with simpler, better known words.

This led to our Semantic drift of quotations in blogspace paper. The whole analysis is free software and thoroughly documented, so that it can be remixed or can serve as inspiration or as an example for other projects. The code lives here, the documentation here.

Substitution detection models
Feature susceptibilities POS susceptibilities

2012-2014 — Daydreaming

Mind-wandering is something we all do about 50% of our waking time: at any moment of the day, while reading a book, or working, or during any activity, you can disconnect from the immediate environment and start thinking about the past, the future, or people and places far away. Most of the time you won’t even realise it until a few moments have passed, or even not at all! Getting to know more about this phenomenon has been traditionally quite challenging: unconscious mind-wandering isn’t something you can trigger on command in the laboratory.

Vincent Adam and I had partnered to start building Android apps for cognitive science, and after a chance encounter with Jonathan Smallwood we started a project to gather better data on mind-wandering thanks to smartphones! The project later transitioned to Mikaël Bastian and Jérôme Sackur for the scientific side, partnering with Gislain Delaire for the design. Eventually created the Daydreaming app, which would ask you at random moments of the day if you were mind-wandering or not (and delve into the details if you were). Launching it was a nice adventure, and led to another paper, Language facilitates introspection, which investigates the effect of mind-wandering in language vs. in images on the probability that you’ll realise that you are currently mind-wandering.

Daydreaming question example Daydreaming rhythms results example Daydreaming thought type results example

Here too, the whole process was open and all the parts are released as free software: the code is here. The project also featured as a pilot experiment for the Science en Poche project Vincent and I helped bootstrap with David Chavalarias, securing an “Émergence(s)” grant from the City of Paris later on.

2009-2011 — Rugby

Studying the social and cognitive sides of life together is hard, and a substantial reason for that is that our metaphors and categories might have to change a lot before we can reach satisfying descriptions of cognition-with-social: notably, the separation between the two might not be so obvious. Now another factor is that historically few people have been interested in this topic. So there is much to do, and exhibiting a few strong experimental links between the social and the cognitive is a good place to start to attract interest and spark more debate.

A few years back, with Julien Clément and Florence Weber, we started a project aiming to do just that: experimentally show a strong link between the culture we are brought up in and the low-level automatic perception we have of a situation. Such links are probably present everywhere, and the literature in social science is rife with starting points (take Bourdieu’s habitus for instance), but eliciting one robustly requires a situation where the link is particularly strong and well-identified.

So based on Julien Clément’s PhD dissertation (Le rugby de Samoa: les techniques du corps entre “Fa’Asamoa” et mondialisation du sport), we aimed to study the way rugby players can perceive a situation differently depending on their upbringing. Samoan players, for instance, contrast strongly with French players: their individual technical competence can be quite superior to that of French players, while being unable to play some team situations that the French would consider textbook, for instance when particular group perception is necessary to win. Julien’s work suggests that this strongly relates to players having different low-level perceptions of a given situation, depending on the rugby they have been trained to play, and the social milieu they were brought up in.

The goal for the project was to study this hypothesis with the tools of neuroscience, but we had to abort shortly after the preliminary phase because of timing constraints and the lack of clear funding possibilities (among others, funding for a PhD on the subject). The sole output of the endeavour was my Rugby et Cognition: Enquête auprès du Racing Métro 92 internship report studying the Parisian rugby club Racing Metro 92, as a first step to identify potential situations where we could develop this idea (since flying international rugby players across the globe for an EEG or MRI is in no lab’s budget). Certainly an unrealised opportunity, which might come back to life if the future allows!


Publications

Peer-reviewed journals

2018

Sébastien Lerique and Camille Roth (2018). The semantic drift of quotations in blogspace: a case study in short-term cultural evolution. Cognitive Science, 42(1), 188-219. [Publisher version]

2017

Mikaël Bastian, Sébastien Lerique, Vincent Adam, Michael S. Franklin, Jonathan W. Schooler, and Jérôme Sackur (2017). Language facilitates introspection: verbal mind-wandering has privileged access to consciousness. Consciousness and Cognition, 49, 86-97. [Publisher version]

2016

Sébastien Lerique (2016). Pour une étude de la dynamique du sens : Réflexions épistémologiques sur la mémétique et l’épidémiologie des représentations. Travaux de linguistique, 73(2), 45-68. [Publisher version]

Communications

Upcoming

Sébastien Lerique and Camille Roth (upcoming). Descriptive modelling of utterance transformations in chains: short-term linguistic evolution in a large-scale online experiment. HBES 2018, Amsterdam, July 2018.

Past

Sébastien Lerique, Éric Fleury and Márton Karsai (2018). Linguistic and social network coevolution: joint analysis of heterogenous sources of information in Twitter. NetSci 2018, Paris, June 2018.

Sébastien Lerique and Camille Roth (2017). Lexical transformations in blogspace: a case study in short-term cultural evolution. Inaugural Cultural Evolution Society Conference, Max Planck Institute for the Science of Human History, Jena, September 2017.

Sébastien Lerique and Camille Roth (2016). Cultural attractors by iterated sentence reformulation: elements of the cognitive story in complex contagion. Data Driven Approach to Networks and Language, ENS Lyon, May 2016.

Sébastien Lerique (2015). Pour une étude du contexte d’interprétation. Les Mèmes Langagiers workshop, Université Paris-Sorbonne, December 2015.

Sébastien Lerique and Camille Roth (2015). Empirical Attractors in Sentence Reformulation Processes. Conference on Complex Systems 2015, Arizona State University, September 2015.

Sébastien Lerique (2014). Cultural Epidemiology. A quest for ontologies where the social and cognitive sciences can meet. Journée scientifique des doctorants de 1è année de l’ED3C, Paris, March 2014.

Vincent Adam, Sébastien Lerique, and David Chavalarias (2014). Pocket Science. Easing development of political, scientific, and citizen apps. 3rd Citizen Cyberscience Summit, University College London, February 2014.

Vincent Adam, Sébastien Lerique, Florence Ruby, Haakon Engen, Gislain Delaire, Tal Yarkoni, and Jonathan Smallwood (2013). Daydreaming: Exploring mind-wandering with smartphones. Brainhack 2013, Paris, October 2013.

Project reports

Sébastien Lerique (2016). The Gistr Platform. Deliverable for the Algopol project ANR-12-CORD-0018.

Academic theses

Sébastien Lerique (2017). Epidemiology of representations: an empirical approach. Thesis submitted for the degree of Ph.D. in Cognitive Science, supervised by Jean-Pierre Nadal and Camille Roth. École des Hautes Études en Sciences Sociales, Paris. Read the jury’s report — here are the jury members:

Sébastien Lerique (2012). Comment les cerveaux copient-collent-ils ? Dérive sémantique des citations dans la blogosphère. Thesis for the M.Sc. in Cognitive Science Cogmaster, supervised by Camille Roth. École Normale Supérieure de Paris.

Sébastien Lerique (2011). Rugby et Cognition. Enquête auprès du Racing Métro 92. Research internship report, supervised by Florence Weber and Julien Clément. École Normale Supérieure de Paris.

Media

13th October 2017. When the Internet sheds light on the transformation of messages. Cogito n°3 (Sciences Po research newsletter).


Education

See my CV for more details. Here is the shorter version: