Ongoing microtwinning projects

Collaborators:

• [ESR] Marta Kąpielska
  • University of Warsaw (UW)
• Johane Nedergaard
  • Aarhus University (AU)
• Anna Borghi
  • Sapienza University of Rome (Uniroma1)

Abstract:

Background. Many researchers have suggested that the processing of
abstract concepts depends on language with a recent study (Villani et
al., 2019) arguing that some abstract concepts, such as those that are
negotiated through social interactions, could be more reliant on
language than others, such as physical, spatiotemporal, and
quantitative concepts.

Aims. In this study, we aim to test the role of language in the processing
of different categories of abstract concepts by having participants solve
odd-one-out problems while engaging in either verbal or nonverbal
secondary interference tasks.

Methods and procedures. We conduct the study online. In the main part
of the experiment, we ask participants to alternate between odd-one-out
abstract concept trials and 1-back matching verbal or nonverbal
interference trials. We also include a control condition with no
interference.

Predictions. If linguistic resources are needed for processing abstract
concepts, we predict that response time for the odd-one-out task will be
slower and accuracy with be lower in the verbal interference condition
compared with the nonverbal interference and the control condition.
Further, if some categories of abstract concepts rely more on linguistic
resources than others, we predict that there will be significant
differences in how much the categories are affected by verbal
interference.

Implications. If our hypotheses are confirmed, this will provide support
for the idea that language – both as it occurs internally and between
people – plays an important role in the formation and negotiation of
abstract concepts as well as the processing of them.

Collaborators:

• [ESR] Jędrzej Miecznikowski
  • University of Warsaw (UW)
• [ESR] Valentina Rossi
  • Sapienza University of Rome (Uniroma1)
• Julian Zubek
  • University of Warsaw (UW)
• Joanna Rączaszek-Leonardi
  • University of Warsaw (UW)
• Anna Borghi
  • Sapienza University of Rome (Uniroma1)
• Chiara Fini
  • Sapienza University of Rome (Uniroma1)

Abstract:

More abstract words are generally slower to process and more difficult
to recall (concreteness effect) and, due to their difficulty, they might
generate more uncertainty as to their meaning. Some authors have
proposed that the metacognitive awareness of the inadequacy of our
knowledge (Shea, 2018) might lead to relying more on other people
(social metacognition Borghi et al., 2018, 2019). A recent study
supports this hypothesis, showing that participants’ movement is more
synchronous with an experimenter when participants have to guess
which abstract vs. concrete concepts an image refers to (Fini et al.,
under review).

We plan to perform a movement tracking study (or, due to COVID
restrictions, to analyse previously collected data) with concrete and
abstract concepts during an interaction/conversation between two
people who use concrete and abstract concepts.
We will pursue two different aims.

1) We intend to investigate whether abstract concepts elicit more
prosocial behavior due to the stronger need to rely on others to
comprehend them. Therefore, we intend to explore whether the
coordination in behavior within dyads is different while using
abstract vs. concrete words. We expect more overall
coordination, not necessarily synchronous; we also intend to
investigate whether some turn-taking might appear during
meaning negotiation.

2) We also intend to simultaneously track different body parts
(head, torso, hands, etc.), possibly trying to distinguish between
different groups of concepts based on that. For example, verify
that some concepts elicit more movement in participants’ hands
(event on the individual level, disregarding synchronisation).
This would also make it possible to filter out task-based
movement.

The study might have wide implications, both from the perspective of
processes necessary to make sense of abstract concepts (not only
intrapersonal but also interpersonal?), and from the perspective
of pragmatics (do we deliberately use abstract words rather than
concrete ones to influence the relation somehow?).

References

Fini, C., Era, V., da Rold, F., Candidi, M., & Borghi, A. M. (2020). Abstract
concepts in interaction: The need of others when guessing abstract concepts
smooths dyadic motor interactions.
Rączaszek-Leonardi J, Krzesicka J, Klamann N, Ziembowicz K, Denkiewicz M,
Kukiełka M and Zubek J (2019) Cultural Artifacts Transform Embodied
Practice: How a Sommelier Card Shapes the Behavior of Dyads Engaged in
Wine Tasting. Front. Psychol. 10:2671. doi: 10.3389/fpsyg.2019.02671

Collaborators:

• [ESR] Elisa Złamańska
  • University of Warsaw (UW)
• Albertyna Osińska
  • University of Warsaw (UW)
• Anna Borghi
  • Sapienza University of Rome (Uniroma1)
• Clauda Mazzuca
  • Sapienza University of Rome (Uniroma1)

Abstract:

Research shows that abstract and concrete concepts rely on the
sensorimotor, emotional and social systems as well as interoceptive
experience (Borghi et al., 2018; Villani et al., 2021). However, there are
significant differences in how these two kinds of expressions are
perceived. In comparison to concrete concepts, abstract expressions
activate the previously mentioned systems to a greater extent,
especially the social and interoceptive dimensions (Villani et al., 2021).

These concepts are more difficult to learn, since they are more detached
from reality and do not have a single, concrete referent in the physical
world. As a consequence, abstract concepts require more time to be
processed and in memory tasks they are recalled less accurately (Villani
et al., 2021).

A different level of abstractness is represented by the family of
metaphors. This kind of figurative speech is especially grounded in the
social dimension. Metaphors help to detect mental states and emotions
as well as create a sense of interpersonal closeness or intimacy (Bowes
& Katz, 2015). Moreover, they evoke stronger affective responses to
their literal equivalents, which can be seen in the enhanced heart rate
(Rojo et al., 2014) or activation of brain areas related to emotional
processing (Citron & Goldberg, 2014; Citron et al., 2019).

Idiomatic expressions represent the least creative kind of figurative
speech, however they still share similarities with metaphors (Citron et
al., 2019). For instance, idioms strongly rely on the social dimension,
since the relationship between their lexical and phrasal meaning is often
arbitrary and it has to be learned (Citron et al., 2019). Furthermore, a
recent neuroimaging research showed that idioms, in comparison to
their literal translation, evoke an enhanced emotional engagement in
the brain like metaphors (Citron et al., 2019). Idiomatic expressions

This project has received funding from the European Union’s Horizon 2020
research and innovation programme under grant agreement No 952324.

From Social Interaction to Abstract Concepts and Words:
Towards Human-centered Technology Development
related to body parts also activate the relevant sensorimotor
representation in the cortex (Boulenger et al., 2009).

In our study we explore idioms’ processing, especially their affective
and interoceptive aspects and how they differ from abstract concepts.

To investigate this problem we focus on idiomatic expressions related to
the human body, in particular the one including words connected with
the head (e.g. “over one’s head”), arms (e.g. “all thumbs”) and legs (e.g.
Achilles’ heel”). We expect that idioms will represent a distinct level of
abstraction, i.e. they will differ from single-word abstract concepts (e.g.
processing the idiomatic expression “all thumbs” will be represented
differently than the single-word abstract concept “clumsy”). We
presume that idioms will be more grounded in the interoceptive
dimensions and they will be linked to a higher emotional engagement.

Collaborators:

• [ESR] Katarzyna Lipowska
  • University of Warsaw (UW)
• Angelo Cangelosi
  • University of Manchester (UoM)
• Anna Borghi
  • Sapienza University of Rome (Uniroma1)
• Caterina Villani
  • Sapienza University of Rome (Uniroma1)

Abstract:

Recent studies have shown that, during the processing of more abstract
concepts, the mouth motor system is engaged, while the hand motor
system is more involved during more concrete concepts processing
(Borghi et al., 2019; Dreyer&Pulvermuller, 2018; Ghio, Vaghi &
Tettamanti, 2013). The engagement of the mouth motor system has
been ascribed to the involvement of inner speech due to the difficulty of
abstract concepts (Borghi et al., 2018). Inner speech would help us
monitor for our knowledge and search for possible conceptual
meanings in working memory.

Numerical concepts represent an exception, since, despite their
abstractness, they engage the hand effector, likely because of the
finger counting habit (Fischer &Brugger, 2011; Fischer &Shaki, 2018;
De La Cruz, Di Nuovo, Di Nuovo & Cangelosi, 2014). However, most of
the experiments in numerical cognition focus on numbers below 10,
while the linguistic system might be more involved for high numbers.
The project is aimed at exploring whether numbers below and above
ten might differently involve the hand and mouth effectors because of
the possible stronger engagement of language and inner speech with
higher numbers.

Collaborators:

• [ESR] Wiktor Rorot
  • University of Warsaw (UW)
• Wenjie Huang
  • University of Manchester (UoM)
• Angelo Cangelosi
  • University of Manchester (UoM)
• Joanna Rączaszek-Leonardi
  • University of Warsaw (UW)

Abstract:

Neuroscientific research highlights strong dependence between spatial
cognition and abstract concepts. Initially suggested in the idea of
“cognitive maps”, these representations have been extensively studied
and significant support for their existence in the human brain has been
already accumulated. The central thread of this research, in the work of
e.g., Peter Gärdenfors and Edvard Moser, has focused on the role of
hippocampal formation and the correspondence between allocentric
space representation and encoding the properties and relations of
concepts.

Allocentric frame of reference is one of two frames of reference
employed by human (and in general mammal) cognitive system. It
represents space in a detached way, similarly to a map representation.
Allocentric space representation is primarily located in the
hippocampus, and employs specialized cells which represent individual
places, paths, and the structure of the environment.
The other, egocentric system is centered on the position of the subject
in the environment and encodes spatial properties in a relational way.
It uses a variety of different neural systems, most significantly located
in the posterior parietal cortex.

Most researchers believe that the operation of the two systems is
complementary and the brain employs one or the other depending on
the task and its complexity (e.g., amount and size of objects perceived).
However, some, e.g., Flavia Filimon, argue that in fact all frames of
reference are inherently egocentric, while others, e.g., Alain Berthoz,
argue reversely – that the allocentric system is in fact developmentally
primary.

To this date, most research in neuroscience is focused on the
allocentric space representation, while egocentric space and relations
between the two frames of reference remain understudied.
The goal of the current project is to address the question of the

From Social Interaction to Abstract Concepts and Words:
Towards Human-centered Technology Development
possible role of egocentric frame of reference for abstraction using a
robotic model. Employing a robotic platform is important for the
project as many researchers underscore the embodied and embedded
character of egocentric frame of reference. Specifically, this project
builds on the model of egocentric space perception proposed by Rick
Grush and later developed by Wiktor Rorot, called Skill Theory.
According to this account, perception of egocentric space is grounded
in skills and sensorimotor states of the subject.

The theoretical analysis will be supplemented by proof-of-principle
computer simulations and experiments inspired by the experiments
performed previously by Ruth Schulz in mobile robots. In this project,
the simulations and experiments will be performed in the iCub
platform, a cognitive humanoid platform for developmental robotics,
available at the University of Manchester. The iCub has already been
employed, although independently, in studies on spatial cognition (e.g.
on hemispatial neglect) and abstraction. Furthermore, the work
performed by Wenjie Huang on information flow for attentional
processes employing the global workspace model provides important
connections to conscious experience and allows for examining the
significance of attention in sensorimotor grounding of concepts.

Collaborators:

• [ESR] Maria Bancerek
  • University of Warsaw (UW)
• Wenxuan Mou
  • Aarhus University (AU)
• Angelo Cangelosi
  • University of Manchester (UoM)
• Julian Zubek
  • University of Warsaw (UW)

Abstract:

Through mother-infant interactions, infants learn to use language to
interact with the environment. By being involved in various social
scenarios, an infant learns to navigate in its surroundings and to
communicate with other people [1]. This, in turn, can lead to abstract
concepts acquisition when the interactions become more complex and
require greater language flexibility. The aim of the project is to underpin
the role of social interactions in the development of symbolic
communication by numerical modelling of early stage mother-child
interactions.

Proposed model consists of two neural networks coupled in a form of
dialogue. The child network, equipped with a set of vital parameters,
relies on the mother network to fulfil its needs, i.e. the mother is able
to change the vital parameters of the child to keep them at the desired
level. However, since the mother network has no direct insight into the
child’s state, communication between the two networks is necessary.

By analogue to the real-life mother-infant dyads, initially,
communication is originated by the mother, who uses language to
engage the infant while taking care of it. The infant learns to relate
messages communicated by the mother to a change of its state –
certain symbolic expressions are then linked to an embodied
experience. As the child learns, the interaction evolves which paves the
way for the acquisition of more abstract terms. Under such conditions,
language learning is facilitated through an active involvement of the
child. Once the child learns to relate language expressions to its own
state, it is able to communicate its needs in a way understandable for
the mother, which plays a key role in its well-being. Therefore, to
comfortably navigate in its surroundings, the child has to assimilate the
mother’s language grounding it in its own experience.

The main objective of the study is to observe how symbolic
communication emerges in an asymmetric interaction, and explore
communication’s imprint on the system. The coupled networks model
could help in understanding the dynamics of language learning process
by an individual immersed in the social landscape, as well as outlining
the relation between embodiment and symbolic communication. This
may help to clarify the hypotheses regarding the developmental order
of acquisition of various types of concepts, and illustrate processes of
abstraction allowing learning concepts beyond the immediate
experience [2].

1. Rączaszek-Leonardi, J., Deacon, T. W. Ungrounding symbols in
language development: implications for modeling emergent symbolic
communication in artificial systems, conference paper (2018)
2. Di Nuovo, A., Cangelosi, A. Abstract Concept Learning in Cognitive
Robots. Curr Robot Rep 2, 1–8 (2021).

Collaborators:

• [ESR] Urszula Kuczma
  • University of Warsaw (UW)
• Kristian Tylén
  • Aarhus University (AU)
• Joanna Rączaszek-Leonardi
  • University of Warsaw (UW)

Abstract:

The project we propose investigates the relation between abstraction
and creation of novel ideas. As an exemplary form of ideation we use
divergent thinking. Divergent thinking (Guilford, 1956) is the process of
spontaneous generation of ideas based on exploration of possibilities. It
is believed to be based on the ability to create remote semantic
associations (Wang et al. 2018). We believe these associations are
tightly connected to the process of generalization, feature extraction
and abstraction in general. We are interested mostly in distant
associations, regarded as long semantic jumps between domains. We
believe that there is an underlying dimension for each pair of
associated concepts along which these associations can be aligned and
we want to investigate these dimensions as forms of abstraction. The
associative dimensions potentially linking concepts vary from context to
context and are possibly not reducible to an exhaustive list. Yet, the
project is based on the assumption that successful divergent thinking
processes might be based on the ability to multiply the dimensions of a
solution space to make it possible to connect any two semantic nodes
by at least one dimension. We would like to analyze the
above-mentioned semantically-distant retrieval both in terms of its
roots and timeframe.

The project will consist of a theoretical analysis and a set of
experiments. The analysis will aim at relating the concept of abstraction
to divergent thinking processes on the ground of existing theories of
abstraction. The experiments will rely on a verbal fluency task. Upon
completion of the task, participants are asked to introspect on the
motivations of their associations. That will allow us to study divergent
thinking at a metacognition level by mapping the number and quality of
associative links and the relation to task performance (how many
candidate solutions the participant produces within a set time frame). A
link of such associative processes to creativity is currently widely
disputed (Smeekens, Kane, 2016, Fox and Beaty, 2019).

This comparison and analysis following it will help us anchor the position of abstracting
processes within the boundaries of discussion about creative ideation.
Further, we want to test how interaction in dyads influences
performance on the task. We hypothesize that social interaction might
motivate a larger amount of possible associations between concepts
resulting in a wider range of produced concepts and a faster process.

This prediction relates to the phenomenon of the “adjacent possible”
(Kauffman, 1995), that is, when a response of one subject induces a
specific association in the other that opens up a whole new sub-domain
of possible ideas. Our project will inform discussions in current theories
of abstraction and aims to find common ground between some of the
approaches for better understanding of nuances that occur during
divergent thinking processes. The dimensions identified will be related
to the ideas of modality of abstraction (Lynott, Connell et al., 2019),
type-token interpretation (Barsalou, 2003) and some aspects of
current connectionist approaches (Lowney II et al., 2020).

Collaborators:

• [ESR] Borys Jastrzębski
  • University of Warsaw (UW)
• Kristian Tylén
  • Aarhus University (AU)
• Anna Borghi
  • Sapienza University of Rome (Uniroma1)

Abstract:

A number of researchers have approached the elusive problem of the
origin of abstract concepts as well as their role in the development,
evolution and adaptability of natural languages. Many of those make
extensive use of lexicons as the source of evidence in their analyses
(e.g. Blondin-Massé et al. 2008, Goddard and Wierzbicka 2014, Youn et
al. 2016, Lupyan and Winter 2018, Borghi 2019). However, to a large
extent, their methodologies differ substantially enough not to be readily
applicable to each other’s research programmes, despite similar
corpora and convergent aims. In this Microtwinning project, we aim to
bridge the theoretical gap between different methodologies of research
on the grounding of abstract symbols and address a few hypotheses
that arise in between, especially on the modality and the type of
grounding at work in the so-called grounding kernels of the
natural-language lexicons and their evolution in time, using an
integrative approach.

First, we focus on the approaches initiated by the works of Stevan
Harnad and Anna Wierzbicka. We will generate grounding kernels of
several contemporary European languages, starting with Polish, Danish
and Italian, from monolingual dictionaries according to the procedure
described in Blondin-Massé et al. 2008 and Picard et al. 2009. This is to
establish whether their findings for English remain relevant for
languages of varying levels of similarity. Then, we will compare the
contents of the common grounding kernel and the kernel core with the
semantic primes (lexical universals) identified by Wierzbicka in her
Natural Semantic Metalanguage, and their exponents in the respective
languages, to calculate the level of overlap. In addition to the
knowledge of the mutual relation of the two research models, this study
will provide information about the stability and the order of time involved
in the evolution of the grounding kernel. Are the cores universal and
stable in time or will they differ significantly between relatively close
languages?

Second, we extend the study from step one to include historical data.
For that purpose, we will curate a number of historical dictionaries or
comprehensive language corpora and pass them to the network model prepared for contemporary languages. We will compare the grounding kernels for one language across centuries and between several
languages in the same period. The results thus obtained will inform us
about the kind of changes involved in the evolution of a language, as
well as provide more fine-grained data on the time-frame and the
precise modes of evolution of the grounding kernels.

Third, we will conduct a polysemy network analysis of the same
contemporary European languages using retranslation (after Youn et al.
2016). We will construct the models on the modern and historical data
and calculate the overlap with Harnad’s grounding kernels and
Wierzbicka’s semantic primes.

Finally, we will check the cores obtained in the three steps above
against psycholinguistic databases (e.g. MRC, Lancaster Sensorimotor
Norms and Brysbaert et al. 2014 measures) to establish their
concreteness levels, types of modality of grounding and age of
acquisition, at least for their English counterparts.

References
Blondin-Massé, A., Chicoisne, G., Gargouri, Y., Harnad, S., Picard, O.,
& Marcotte, O. (2008). How is meaning grounded in dictionary
definitions? Les Cahiers du GERARD, G-2008-48.
Borghi, A.M. (2019). Linguistic relativity and abstract words. Paradigmi,
37(3), 429–448.
Brysbaert M, Warriner AB, Kuperman V. 2014 Concreteness ratings for
40 thousand generally known English word lemmas.Behav. Res.
Methods46, 904–911. (doi:10.3758/s13428-013-0403-5)
Goddard, C. and Wierzbicka, A. (2014). Words and Meanings: Lexical
Semantics Across Domains, Languages, and Cultures. Oxford: Oxford
University Press.
Lupyan, G. and Winter, B. (2018). Language is more abstract than you
think, or,why aren’t languages more iconic? Phil.Trans. R. Soc. B373:
20170137. http://dx.doi.org/10.1098/rstb.2017.0137.
Picard, O., Blondin-Massé, A., Harnad, S., Marcotte, O., Chicoisne, G.,
Gargouri, Y. (2009). Hierarchies in Dictionary Definition Space.
arXiv:0911.5703.
Youn, H., Suttond, L., Smith, E., Moore, C., Wilkins, J. F., Maddieson, I.,
Croft, W. and Bhattacharya, T. (2016). On the universal structure of human lexical semantics. PNAS.
www.pnas.org/cgi/doi/10.1073/pnas.1520752113

Collaborators:

• [ESR] Michał Pięta-Lendzion
  • University of Warsaw (UW)
• [ESR] Małgorzata Pięta-Lendzion
  • University of Warsaw (UW)
• Joanna Rączaszek-Leonardi
  • University of Warsaw (UW)
• Travis Wiltshire [external collaborator]
  • Tilburg University

Abstract:

Interpersonal coordination on various dimensions (e.g. vocal, physiological) during interaction proves to be an important factor that underlies satisfaction with an interaction, team performance, and the feeling of sharing the same emotions (Fusaroli et al. 2015). In psychotherapeutic setting interpersonal coordination seems to have a complex influence on the feeling of being understood, psychotherapy outcome, and intensity of symptoms. In particular coordination of movement proves to be strongly connected to psychotherapy outcomes (Wiltshire et al. 2020) while vocal synchrony (reflected in the raised pitch of a voice) can raise the intensity of symptoms in social anxiety disorder (Schoenherr et al. 2021).

The main goal of the project is to explore the embodiment of abstract concepts during interpersonal group sessions. In particular, we want to explore the heterogeneity of vocal and physiological (i.e., pulse, movement) coordination among interpersonal group participants
during sessions, and its’ relationships to group participation satisfaction.

The main hypothesis for this research project is an adaptive role of thephysiological synchrony of participants during the interactions.

Particularly, we suppose that linking abstract psychological constructsthat are discussed in a group session to personal experience, e.g., whenparticipants share personal stories illustrating the experience, leads togreater physiological synchrony of participants that engage in theinteraction (Kallen et al. 2016). On the contrary, the escalation of vocalsynchrony in participants in the context of negative affectivity leads tolower compatibility of physiological indicators and thus, will bereflected in lower subjective therapy satisfaction and/or a sense ofbeing misunderstood and lower subjective emotional involvementduring the sessions (Schoenherr et al. 2021).

Collaborators:

• [ESR] Peter Thestrup Waade
  • Aarhus University (AU)
• Julian Zubek
  • University of Warsaw (UW)
• Kristian Tylén
  • Aarhus University (AU)
• Rebekah Baglini
  • Aarhus University (AU)
• Fernando Rosas [external collaborator]
  • University of Sussex
• Olivia Foster vander Elst
  • Aarhus University (AU)

Abstract:

In this project, we will investigate the emergence of synchronicity, synergy and syntax in Lindy Hop pair dancing. As such, the project is related to challenge number 8 in the call, namely social interactive processes of abstraction.

We will use Lindy Hop pairdancing as an example of immediate, intimate, improvised, joyful and creative social interaction, where co-creation of concrete complex movement patterns, sometimes synchronized, leads to the emergence of an abstract synergetic between-dancer phenomenon, and an abstract movement class based syntax.

In the course of the project, we will try to disentangle, and operationalize differently, a few distinct concepts that are often conflated, including:
Synchronicity, that is, the two dancers doing the same thing at the same time (potentially in complex ways)
Coupling, that is, the degree to which the movements of the two dancers depend on each other

Coordination, that is, the two dancers moving in ways that are useful i.e. allows them to perform the dance

Synergy, that is, an emergent between-dancer ‘oneness’, where the movements of one dancer are only meaningful in relation to the other dancer, and where they cannot be reduced to their components

We will use a combination of movement capture technology, heartrate measurements, and subjective ratings, to investigate

1) whether Lindy Hop pairdancing is a synergetic activity, that is, whether it is best described at the pair level (as opposed to the single dancer level)

2) how dance experience and music familiarity affects synchronicity,

synergy and syntax. Is synchronicity and strong entrainment to the music for example necessary for synergy, but only for inexperienced dancers? Are more experienced dancers able to combine longer sequences of movements, in syntactically more complicated phrases – and does this depend on how well they know the music they dance to?

3) how syntax, synchronicity, synergy and complexity affects subjective experience of enjoyment, feeling of connection, and aesthetic ratings. Is it, for example, the complexity of a given dancer’s movements, or rather the synchronization with the partner, their synergy, or the complexity at the pair level, which best predicts which times in a dance are the most pleasant?

4) we will investigate how different leader-follower dynamics (from closed leading with a strong unidirectional lead, to open leading or even switching, where leader-follower roles blend into each other, and which is more improvised and more difficult) affects the above questions, as well as different canonical leader-follower operationalizations (like from the CRQA, or transfer-entropy).

Finally, we will compare our findings with, or maybe replicate them in, data from a similar study on Cuban Salsa which we collaborate with.

Collaborators:

• [ESR] Ewa Nagórska
  • University of Warsaw (UW)
• Joanna Rączaszek-Leonardi
  • University of Warsaw (UW)
• Claudia Mazzuca
  • Sapienza University of Rome (Uniroma1)
• Anne Fausto-Sterling [external collaborator]
  • Brown University
• Damian Kelty-Stephen [external collaborator]
  • The State University of New York-New Paltz

Abstract:

Interactional routines are forms of cooperation during which children’s agency and their ability to form important bonds is shaped, even in early childhood. The types of behaviours, their sequences, and the timing of the interaction is a mark of values realized during them.
Current methods of examining interactions, as well as new methods of dynamical analyses, enable us to check whether these values are
different, depending on if the parents are interacting with girls or boys, from the very early stage of life.

The main goal of the study is checking whether parent-child interactional routines differ in case of interactions between parent and either boys or girls. The study will be conducted in cooperation with Prof. Anne Fausto-Sterling, whose research during the last few decades has been focused on shaping of the gender/sex identity on biological, social and historical levels. Prof. Fausto-Sterling has proposed applying dynamical methods approach to interactions, to verify some theses of her processual model of shaping gender/sex identity (e.g. Fausto-Sterling, 2021). To create the model, she has been working with Dr. Damian Kelty-Stephen, who also joined the project. The connections they discovered, as well as the methodology of their work, had been described in details in a paper by Eason et al. (2020); we plan on applying similar methodology to analysing the interactions between mothers and younger children (up to 12 months old).

The study will be mostly exploratory, but based on previous research by Fausto-Sterling; Mazzuca et al.; and Nomikou, Rohlfing and Rączaszek-Leonardi, one can hypothesize about:
1. the dynamics of multimodal behaviours in interactions (method after: Rohlfing et al., 2019);
2. the dynamics and types of vocalizations
3. the content of ‘messages’ directed at infants (Fausto-Sterling, 2021), and verbalized by mothers (Mazzuca et al., 2020)
4. word frequency

We plan on deriving concrete hypotheses regarding possible differences in interactions with infants of different gender/sex, and, possibly, integrating the model with the theoretical approach to value realizing in interactions (Hodges & Baron, 1992; Hodges & Rączaszek-Leonardi, 2022).

The project might contribute to 1) identifying unconscious tendencies to some behaviours in interactions, such that can favour shaping of gender roles compliant with stereotypes present in the particular culture, and 2) creating a theoretical framework for analysing value realizing in interactions and the processes of early acculturation.

Collaborators:

• [ESR] Michał Pięta-Lendzion
  • University of Warsaw (UW)
• [ESR] Haodong Xie
  • The University of Manchester (UoM)
• Joanna Rączaszek-Leonardi
  • University of Warsaw (UW)

Abstract:

Chicken Game is a setup in the game theory that intends to model a conflict between two individuals. Its most stereotypical illustration is the one in which two drivers are heading directly at each other. One of them has to swerve to avoid a crash which would be deadly for both of them. However, both drivers want to avoid swerving, as one which swerves is considered to be a coward (thus “chicken”). Various payoff matrices can be implemented for this game so the players can be more or less prone to swerve.

We would like to implement Chicken Game or its modification to create a situation in which agents will have to create a common strategy to help them survive and achieve the highest possible payoff. Chicken Game is an anti-coordination game. This means that the payoff is greatest when the players behave differently. We think that such a setup can result in a strategy in which agents will accept the common convention of passing each other and if agents will be able to use signals, some common language will emerge.

After we will create simulations in which agents will develop a common strategy for avoiding crashing, we will observe the evolution of a convention. We want to observe how changes in the environment, agents’ properties, and setup of the Chicken Game influence the behaviour of the agents in the moment of passing each other. We will try to capture the moment of the sign’s emergence.

Collaborators:

• [ESR] Katarzyna Skowrońska
  • University of Warsaw (UW)
• [ESR] Anna Alińska
  • University of Warsaw (UW)
• [ESR] Jakub Strebeyko
  • University of Warsaw (UW)
• Joanna Rączaszek-Leonardi
  • University of Warsaw (UW)
• Kristian Tylén
  • Aarhus University (AU)
• Angelo Delliponti [external collaborator]
  • Nicolaus Copernicus University

Abstract:

The goal of our project is to provide new insights on the elements of the ostensive model of communication and investigate the possibilities of recreating them within the artificial systems.
According to the ostensive model of communication (Scott-Phillips, 2014; Heintz & Scott-Phillips, 2022, Sperber & Wilson, 1986) human language is made possible by the ability of expression and recognition of communicative and informative intentions. In this view language is the result of an evolutionary process which had as its initial stage some kind of non-verbal communication, probably gestural and embodied, in which ostensive signals were expressed in order to communicate to others that we want to draw their attention to something, and therefore to facilitate cooperation (Tomasello, 2010). The project aims at exploring this view on the foundations of language by conducting a semiotic experiment, which would involve both human participants and artificial agents.
The participants will play a simple 2d computer game in which each player controls the movement of a virtual agent. Their task in the game would be to coordinate with other players in a specified way; for example the participants would be divided in two groups and their goal would be to gather all the group members in one place. The information required for successful coordination will not be directly accessible, therefore players will have to communicate with each other and the only way in which they can is through movements of their agent over the board.