Tom Froese

New evidence of the interactive basis of direct social perception

Cognitivists have started to agree with enactivists that the minds of others are given in perceptual experience, for instance when we see happiness in another’s smile. However, their accounts of the basis of this experience diverge considerably. Cognitivists treat the experience of other minds as a product of the perceiver’s brain-based information processing, and hence as a property of the individual alone. Enactivists, on the other hand, include social interaction as a constitutive part of the dynamical basis of the experience of other minds, and hence are open to the possibility that one and the same social experience is shared across subjects. Accordingly, these two approaches make divergent predictions about the role of others in the constitution of social perception, which can be tested with a suitable real-time social interaction paradigm, like the perceptual crossing experiment. Here I report results of a new study with this paradigm that support the interactive basis of social perception. Variations in subjective clarity of the experience of the other is better accounted for at the level of the dyad, rather than at the individual level, and mutual recognition correlates with collective sensorimotor integration. I argue that this supports a direct perception account of social experience in which the other herself, rather than an internal representation of the other, is experienced.

Erik Rietveld & Julian Kiverstein

Skilled We-Intentionality: Joint Action Without Shared Intentions and Goals

What kind of understanding of other people is required if we are to succeed in acting jointly with them? In particular, is it necessary to make inferences about their intentions and goals? In a recent paper, Katja Abramova and Marc Slors have challenged the claim that mindreading is necessary for action coordination (Abramova & Slors 2015). They argue that people can succeed in performing a task together based on their perception of each other’s responsiveness to relevant affordances. Abramova and Slors however agree with a core component of the standard philosophical analysis of joint action that when people act jointly they must do so on the basis of a shared intention or goal. We will argue that the concept of goals and intentions doesn’t explain what makes skilled agents selectively open and responsive to relevant affordances. We will suggest this explanatory work can be done instead by our concept of skilled intentionality. When people engage in activities together we will argue, it is the shared relevant affordances, and the wider context in which they are situated (the “landscape of affordances”) that does the work of explaining how they coordinate with each other.

Bart Geurts

Evolutionary pragmatics

According to the received view, human communication is essentially a form of “mind reading”. On this view, speakers make statements, promises, and requests in order to express beliefs, intentions, desires, and so on, and communication succeeds if their audience understands what mental states are being expressed. It is well known, however, that the received view raises developmental issues. If understanding a promise is essentially a matter of grasping the speaker’s intentions, children must be able to attribute intentions before they can understand promises. But how likely is it that two-year olds have this skill, and how do they learn to attribute intentions and other psychological states if not through communication?

Since there are no working models of how children might learn to attribute mental states before they start dealing with speech acts, it is tempting to stipulate that mental-state attribution doesn’t have to be learned in the first place: it’s in the genes. But this flight into nativism merely passes the buck from development to evolution, and the same questions that arose with respect to the ontogeny of communication and mental-state attribution now arise with respect to their phylogeny.

As an alternative to the received view, I propose that human communication is first and foremost a matter of negotiating commitments, rather than one of conveying intentions, beliefs, and other mental states. Every speech act causes the speaker to become committed to the hearer to act on a propositional content. Hence, commitments are relations between speakers, hearers, and propositions. Their purpose is to enable speakers and hearers to coordinate their actions: communication is coordinated action for action coordination. In recent work, I have applied this framework to several major topics in pragmatics, including speech acts, implicatures, and common ground, as well as some less prominent ones, like fiction and self talk. The chief objective of my lecture will be to argue that a commitment-based approach to pragmatics offers a sounder basis for understanding the evolution of human communication than a mind-reading approach.

Fred Hasselman

Radical Embodied Computation: Reproduction of Similarity by Analogy as an Order Generating Mechanism in Complex Systems

Radical Embodied Computation (REC++) is a term I use to describe various efforts to merge some of the more radical post-cognitivist, non-representational theoretical perspectives on human behaviour and cognition, such as, Radical Embodied Cognition (REC; Chemero, 2009) and Physical Intelligence (PI; Turvey & Carello, 2012) with the theoretical perspective in physics in which reality is essentially made up of physical information, and lawful behaviour is considered the result of natural computation (see e.g. Wolfram, 2002; Pfeifer et al., 2007; Polani et al., 2007; Verlinde, 2011). The main goal is to try to understand human behaviour in terms of physically plausible order generating processes, not in terms of virtual computational architectures that are merely assumed, but never shown to be realised by the nervous system.

One of the areas of interest is the study of so-called bio-physical codes and their relation to the emergence of meaning. For example, the transfer RNA molecule can be said to connect the world of nucleotides (DNA) to the world of amino acids (cf. Barbieri, 2003). Stated in terms of information theory, one can say a message is transmitted from the world of DNA, which is encoded, channelled and decoded through the structure of the tRNA molecule, to the destination world of amino acids. The system as a whole is a physical realisation of encoded meaningful information, it is quite literally the embodiment of contextual meaning: The structure of tRNA “represents” the similarities (redundancies) that exist between the two worlds.

In this talk I will explore whether a mechanism similar to the tRNA molecule exists that can help understand complex adaptive behaviour of agent-environment systems, specifically, the phenomenon of Complexity Matching. Complexity Matching is a resonance phenomenon first described in simulation studies that evidenced that interacting systems that share similar complexities (e.g. fractal scaling in structure or behaviour), this leads to a maximization of information exchange (cf. Aquino et al., 2011). Moreover, adaptive interacting systems tend to attune (or match) their complexities in order to enhance their coordination. Especially the latter phenomenon has been observed in empirical studies as a matching of the complex correlational structure recorded in time series of human physiology and performance, known as (multi-) fractal scaling. Examples include: Synchronisation and coordination of movement (Coey et al., 2016; Almurad et al., 2018), dyadic conversation (Abney et al., 2014), social interaction (Zapata-Fonseca et al., 2016) and speech perception (Hasselman, 2015; Ward et al., 2018).

I will discuss the conjecture to consider scaling exponents as codes connecting two different “worlds” that are separated by lags of time in the context of results obtained from speech perception experiments. I’ll explore whether self-similar information representation (reproduction of similarity by analogy) could constitute a resolution of the representation problem and argue that human neurophysiology of the auditory senses indeed points to a sensitivity to detect acoustic similarities that unfold on different temporal scales.

Mark Blokpoel

Ambiguity helps: How unaligned pragmatic communicators can understand each other

Imagine you want to tell your friend an anecdote about a class mate. Neither of you know the person’s name. How can you communicate who you are referring to? You could refer to features of the person (e.g., ‘the one with red hair’ or `the tall one’) or use more metaphorical terms (i.e., ‘the teacher's favorite’ or ‘the gamer’). Of all the things you could say, in principle, how do you decide what to say? Ambiguity is often viewed as a problem that needs to be overcome. However, in this talk I put forth the possibility that ambiguity may be a helpful feature of a communication system. To investigate how communicative success is affected by asymmetry, ambiguity and the agent's cognitive ability to perform higher-order pragmatic reasoning, we used cognitive, interactive agent-based simulations. We find that higher-order asymmetrical agents can understand each other better than zero-order agents with exactly matched lexicons, but only when their lexicons are ambiguous. This counter-intuitive role of ambiguity under asymmetry shows the importance of exploring the boundary conditions under which referential communication between two interacting agents can be successful.