The enactive approach of cognition is based upon two main ideas. Firstly, the mind is considered as embodied so it cannot be reduced to the brain activity, and similarly, the body cannot be only regarded as a sensorimotor reservoir. Secondly, the lived experience of the cognitive agent, which is embedded into an environment, influences its actions.
In the case of social cognition, the Perceptual Crossing Experiment (PCE) has proved successful for considering the embodiment of the mind (Rohde 2010). Besides, some modifications of the PCE have been able to take into account the subjective experience of the participants as well as support the idea that for social cognition to exist, social interaction is required (De Jaegher et al. 2010, (Froese et al. 2014).
When studying a complex system, such as the human mind, and specifically the interaction-dominant dynamics observed in social cognition, a systems biology approach renders useful because the relevant aspects lie on the interaction between the components of the system and the emergence of collective behaviour, instead of focusing on just the individual components (Fossion & Zapata in press)
Thus, the interaction itself has become a process thoroughly studied by the PCE that has led to the development of new theories on social cognition, like the second-person neuroscience where the interaction dynamics between two subjects are the field of study (Scilbach et al. 2013).
However, recent work (Bedia et al. 2014) has claimed that the analysis of the PCE, both in the simulation and the behavioural fashions, has implicitly assumed that the emergence of social engagement can be reduced to a single time scale.
Therefore, we have adopted a time-series analysis perspective for studying social interaction in order to consider more time-scales and search for long-range correlations and fractal dynamics in the PCE.
Following the methodology used in Bedia et al. 2014, and trying to replicate their results, we have looked for 1/f noise in the time-series of the players from the PCE carried out by Froese et al. in 2014. Moreover, we have studied the correlation between the time-series obtained and the Perceptual Awareness Scale (PAS) used in the same experiment.
We found that, overall, the time series of those players who both had a PAS score of 4 (meaning a clear experience of each other´s presence) showed a scale invariant behaviour and long-range correlations (1/f noise) measured by Detrended Fluctuation Analysis (DFA). So we were able to obtain an objective quantification (by means of time-series analysis) of the subjective impression of the players (PAS values).
All the data were obtained from Froese et al. 2014
Firstly, we plotted the time-series from the positions of both players for every team (n=17) and every trial (n=15). Due to the characteristics of the PCE set up, a correction has to be made in order to obtain a time-series that could be analysed.
Secondly, the instantaneous velocity of each player was calculated by subtracting the successive positions for every point of the time-series (derivative of position). Then, the relative velocity was obtained by calculating the difference between the players’ velocities. The velocities were preferred to the original positions because the latter had too much trend that was unfavourable for the analysis to be made.
Thirdly, we integrate the relative velocities in order to obtain the relative positions, since the positions would not give false positives of human-human interaction.
Finally, DFA was calculated for the time-series of the integrated relative velocity for those trials in which both players shared the PAS values.
The first region of the log F(n) plot corresponds to one oscillation of the time-series of the relative velocity (one player “palpating” the other). This behaviour is quite predictable and therefore a slope with a slope of 2 is fitted almost perfectly; this in turn is compatible with the alfa exponent of Brownian noise. However, at larger scales (ie. the second region of the log F (n) plot) various oscillations are taken into account. And at this point, the behaviour shows a 1/f noise pattern, with an alfa exponent very close to 1 (See Figure 1). These results, though, were only observed for those trials in which both players graded their experience (by means of the PAS) with a value of 4 (lilac), but not when they reported a joint value of 3 (green) or 2 (red); (see Figure 2).
The obtained results support the use of time-series analysis for studying the dynamics of dyadic embodied social interaction because this is a way of objectively measuring the subjective aspects of the interaction process. Also, the use of time-series avoids a short scale analysis due to the statistical methods that can be applied to them.
The correspondence found between the PAS values and the behaviour of the time-series is opposite when taking into account the correctness of the clicks made by the players during the trials (see Froese et al 2014 for detailed information).
A more thoroughly statistical analysis is required in order to better characterise the time-series properties. It is known that the presence of 1/f noise thought to be ubiquitous in human performance is neither necessary nor sufficient evidence of interaction-dominant dynamics (Ihlen & Vereijken 2010). Since we are dealing with the dynamics of social interaction, interaction-dominant and multifractal analysis approach might be of use in our case.
So far, the work presented has only focused on the behavioural component of social interaction. However, the neural correlates are also important for having a holistic approach to social interaction so the time-series analysis of the EEG from the players while engaged in interaction might reveal more information about social cognition, and thus help to get a better understanding of how the human mind works (Dumas et al. 2014).
Moreover, there are many disorders in which the social interaction is altered, such as autism or congenital facial palsy (Krueger & Michael 2012), so a better understanding of social interaction might be useful for new diagnostic and therapeutic methods which would help people who suffer these kind of impairments.
BIBLIOGRAFÍA DE APOYO:
Barandiaran, X. (2008) Mental Life. A naturalized approach to the autonomy of cognitive agents (University of the Basque Country, Spain).
Bedia, M. G., Aguilera, M., Gómez, T., Larrode, D. G., & Seron, F. (2014). Quantifying long-range correlations and 1/f patterns in a minimal experiment of social interaction. Frontiers in Psychology, 5(November), 1–12.
De Jaegher, H., Di Paolo, E., & Gallagher, S. (2010). Can social interaction constitute social cognition? Trends in Cognitive Sciences, 14(10), 441–447.
Dixon, J. a., Holden, J. G., Mirman, D., & Stephen, D. G. (2012). Multifractal Dynamics in the Emergence of Cognitive Structure. Topics in Cognitive Science, 4, 51–62.
Dumas, G., Kelso, J. a S., & Nadel, J. (2014). Tackling the social cognition paradox through multi-scale approaches. Frontiers in Psychology,5(August), 1–4.
Fossion, R., et al.(2010). Scale invariance as a symmetry in physical and biological systems: Listening to photons, bubbles and heartbeats. AIP Conference Proceedings, 1323(2010), 74–90.
Fossion R. & Zapata-Fonseca L. (in press). The scientific method, to be published in: Aging research – Methodological issues, Springer
Froese, T., & Di Paolo, E. A. (2011). The enactive approach: Theoretical sketches from cell to society. Pragmatics & Cognition, 19(1), 1-36
Froese, T., Iizuka, H., & Ikegami, T. (2014). Embodied social interaction constitutes social cognition in pairs of humans: A minimalist virtual reality experiment. Scientific Reports, 4, 3672.
Froese, T., Iizuka, H., & Ikegami, T. (2014). Using minimal human-computer interfaces for studying the interactive development of social awareness. Frontiers in Psychology, 5(September), 1–13.
Froese, T. (2015). Lecture on Participatory sense-making as a part of the masters course “Autonomous agents, social systems and the new cognitive science”. IIMAS, UNAM
Ihlen, E. A. F., & Vereijken, B. (2010). Interaction-Dominant Dynamics in Human Cognition : Beyond 1 / f Fluctuation. . Journal of Experimental Psychology: General, 139(3), 436–463.
Kello, C. T., Brown, G. D. a, Ferrer-i-Cancho, R., Holden, J. G., Linkenkaer-Hansen, K., Rhodes, T., & Van Orden, G. C. (2010). Scaling laws in cognitive sciences. Trends in Cognitive Sciences, 14(5), 223–232.
Krueger, J., & Michael, J. (2012). Gestural coupling and social cognition: Möbius Syndrome as a case study. Frontiers in Human Neuroscience, 6(April), 1–14.
Liebovitch, L. S., & Scheurle, D. (2000). Two lessons from fractals and chaos. Complexity, 5(4), 34–43.
Peng, C. K., Havlin, S., Stanley, H. E., & Goldberger, a L. (1995). Quantification of scaling exponents and crossover phenomena in nonstationary heartbeat time series. Chaos (Woodbury, N.Y.), 5(1), 82–87.
Rohde, M. (2010). Enaction, Embodiment, Evolutionary Robotics. Simulations Models for a Post-Cognitivist Science of Mind. Paris, France: Atlantis Press/World Scientific.
Schilbach, L., Timmermans, B., Reddy, V., Costall, A., Bente, G., Schlicht, T., & Vogeley, K. (2013). Toward a second-person neuroscience. The Behavioral and Brain Sciences, 36(4), 393–414.
Thompson, E. (2007). Mind in Life: Biology, Phenomenology, and the Sciences of Mind. Cambridge, MA: Harvard University Press