Giacomo Melis and Eva Rafetseder

The relation between unreflective and reflective minds

University of Stirling

Some philosophers think that the ability to critically evaluate one’s evidence is necessary to form rational beliefs. They often portray human minds as different in kind from the minds of other animals. Other philosophers allow for a thoroughly unreflective notion of rational belief that applies to young children and non-human animals as well as to human adults and see no difference in kind of minds. It is tempting to explain the difference between these two approaches as hinging on two distinct notions of rational belief and dismiss the disagreement as superficial or merely verbal. But even if such explanation were correct, there remains a question about the relation between the two notions in play. We address this question from philosophical and empirical standpoints. In the first part of the presentation, we suggest that the study of epistemic defeaters may help to explain how unreflective subjects may acquire the capacity for of reflective belief revision. In the second part we describe an experiment built around the ideas illustrated in the first part. The experiment aimed to assess whether young children, pigs and dogs may be capable to ascend to the level of reflective belief-revision based on what they do unreflectively.

Giulia Palazzolo

Is a maximally unified account of animal and human communication achievable and desirable?

University of Warwick

It is a common intuition that nonhuman animals, like humans, communicate. However, vindicating this intuition is not straightforward: in the literature, animal and human communication are typically studied separately, using different–and sometimes inconsistent– accounts of communication. Moreover, none of the dominant accounts of communication in the literature (biological, informational and Gricean) are designed to, or capable of, vindicating the intuition that there is something characteristically common to both animal and human communicative behaviours. Drawing on my analysis of these accounts, I derive two basic constraints on what should count as a plausible Maximally Unified Account of animal and human communication (MUA), discussing two accounts that meet these conditions: Millikan’s theory of intentional signs (2004; 2017) and Green’s theory of organic meaning (2019). Finally, I assess the utility of a MUA in the studies of animal and human communication.

Kirsten Blakey

How Children Revise Their Beliefs in Response to Different Types of Counterevidence 

University of Toronto

Young children revise their beliefs in response to new evidence, but some philosophers argue that rational belief revision requires reflection—pointing to the ability to answer “why” questions, which typically emerges around age six, as the clearest evidence. However, recent work suggests that responding to undermining counterevidence—that challenges the evidence supporting a belief—may itself be a form of reflection. Schleihauf et al. (2022) found that 4- and 5-year-olds were more likely to revise beliefs when faced with undermining evidence rather than positive evidence, suggesting that children reflect on their reasons for belief and the supporting evidence earlier than the “age of reason.” In two studies we examined reflective belief revision using testimonial and non-verbal evidence. In Study 1 (N = 43), we replicated and extended earlier findings: 3- to 6-year-olds were more likely to revise their beliefs when presented with testimony that undermined the evidence supporting their initial belief. Study 2 (N = 84 children, N = 23 adults) used an eye-tracking task in which participants received non-verbal evidence. Adults and 3- to 6-year-olds again revised their beliefs more often after encountering undermining evidence, while 2-year-olds did not. These findings suggest that reflective belief revision can occur in children as young as three, challenging assumptions about the onset of rationality. 

Fredrik Stjernberg

Collingwoodian Creatures 

Linkoping University

Dennett’s “ladder” distinguishes between Darwinian, Skinnerian, Popperian and Gregorian creatures, each with more flexible and advanced cognitive abilities. Perhaps only humans are to be found on the Gregorian level. Gregorian creatures have thinking tools. These tools are abstract (arithmetic, language) or concrete (scissors, computers). Their inner environments are informed by the designed portions of the outer environment.  

Would this identification of Gregorian properties serve as a way of distinguishing the cognitive abilities of humans from non-human animals? Trying to find a specific cognitive trait to distinguish humans from animals has in general been a losing game, and many Gregorian traits can be found in chimps. 

My talk will present reasons to distinguish a fifth level, the Collingwoodian, after R. G. Collingwood. Collingwood wrote: “Every statement that anybody ever makes is made in answer to a question.” It is perhaps surprising that chimps (in the wild) don’t seem to ask questions, whereas our human lives are thoroughly impregnated with this practice; our complex environments wouldn’t be possible to handle without asking for help.  

But with the right prompting, cross-fostered chimps, using sign language, do ask questions. Something has clearly happened with the advent of extra linguistic resources. My talk will be about how asking questions should be understood as one important further step on Dennett’s ladder. 

Rasmus Overmark

Pragmatic relativity in humans and great apes 

University of St Andrews

Recent work proposes that linguistic relativity in mentalizing may arise from pragmatics. Some linguistic devices, such as demonstratives, require their users to attend to their interlocutor’s attention and knowledge for successful use. This means that the pragmatics of such devices train the attention of their users, with possible developmental consequences. 

We propose that pragmatic relativity may arise from the highly popular quotative ‘be like’ (QBL), which can be used to quote utterances and thoughts using “like”, as in: 

I was like, ‘that’s crazy!’ 

QBL does not itself signal whether a thought or an utterance is quoted. This means that we may need to rely on monitoring an interlocutor’s mental states for reliable interpretation of QBL.   

We present results from a new interpretation task on how adults interpret QBL. Our findings suggest that successful interpretation of QBL requires interpreters to pay special attention to contextual information about speaker goals and adherence with social norms. QBL may therefore generate pragmatic relativity. 

If human mentalizing exhibits pragmatic relativity, it may differ from mentalizing in primates with other communication systems. Recent work suggests that non-human primate communication systems involve rich pragmatics. On this backdrop, we ask whether pragmatic relativity is distinctly human. 

Luca Di Vincenzo

Beyond Language: Multimodal Insights into Theory of Mind and Rationality Across Species

Sapienza University of Rome

The relationship between Theory of Mind (ToM), rationality, and multimodality challenges the long-standing assumption that these cognitive capacities are uniquely human and inherently tied to language. Vocal-centric models of cognition have historically marginalized the role of gesture, facial expression, and sensory integration in the emergence of complex socio-cognitive abilities. A multimodal framework suggests that ToM and rational inference may arise from multimodal integration—the coordinated processing of information across sensory modalities—rather than from language alone. Two experimental paradigms were designed to investigate ToM in nonhuman species, focusing on perceptual access and false-belief attribution under ecologically valid, noisy conditions. These environments require subjects to flexibly shift between communicative modalities based on context. Initial trials with Asian elephants indicate sensitivity to others’ perceptual states, while ongoing studies with elephants and capuchin monkeys aim to further assess cognitive flexibility and social reasoning across species. This perspective reframes mental state attribution and rationality as outcomes of sensorimotor and perceptual coordination. Multimodal interaction emerges as a core mechanism in the construction of mental representations and mindreading, offering a biologically grounded alternative to language-centric accounts of social cognition.   

Amanda Seed

University of St Andrews

Cognitive Models and Cognitive Control in Non-human Primates and Children

Christine Caldwell

Cognitive challenges in human cumulative cultural evolution

University of Stirling

Human evolutionary success has been attributed to the capacity to accumulate knowledge and skills over generations via cultural transmission, but explanations for the distinctiveness of human culture remain heavily debated. I will present a series of studies carried out within my research group, investigating relationships between cognitive development and behaviours associated with both acquiring and passing on social information. In most of our experimental work we have endeavoured to capture the kinds of challenges likely to be present in real world cases of cumulative cultural evolution in humans. We find that there are indeed significant cognitive challenges associated with accumulation of knowledge via social learning, illustrated by striking differences in performance between children from different age groups. Our studies suggest that children’s developing capacities for metacognition may be particularly influential in enhancing their ability to find and use social information, and their ability to transmit their expertise to others. This competence appears to develop relatively late and may not be shared by other species. The unique complexity of human culture may therefore be linked to differences in reflective abilities between humans and other species.

Zhengxi Jin

Rationality in association: exploiting the environment’s latent structure 

Kings College London

In this talk, I argue that associative learning, which is widespread across animal species, is ecologically rational: it is optimised to exploit the causal structure of the environment, thereby facilitating decision-making and action control. I begin by clarifying the notion of rationality I will be working with, proposing that it is a matter of how well a subpersonal process takes advantages of those “shortcuts” tailored to local environmental challenges. I then show how the “latent-cause” model of associative learning, as recently developed in cognitive science, fits this conception. On this view, associative learning involves inferring hidden causes behind observable regularities, allowing organisms to partition sensory experiences into chunks that track causally “deep” patterns in their environment. Drawing on philosophical accounts of the metaphysical structure of the world, particularly work on natural kinds, I argue that latent-cause associative learning exploits causally grounded property clusters (natural kinds) in the environment. Finally, I suggest that such associative mechanisms may be precursors to the mental-file structures posited in accounts of human rational thought, as both serve the function of organising information in ways that exploit the environment’s causal structure. The rationality of file-based thoughts, I argue, is similarly an ecological matter, though one that often goes unrecognised. 

Dimensions of Inference: A Unified Account of Inference for Human and Nonhuman Animals 

Ruhr University Bochum

The most common approach to conceptualizing inference in comparative research relies heavily on success testing and ‘binary’ interpretations of the results as reflecting either cognition or ‘mere’ association. Adding to the existing criticisms of this approach, we present a novel framework that allows for a more fine-grained analysis of inference. We proposal a minimal characterization of inference as an informational transfer – a transfer between two informational states mediated by an intermediary representation (IR). The IR is a model of the relevant part of the world that integrates different bits of information into a coherent representation, and it can be characterized along three key dimensions: context-dependence, representational format, and perspectivity. We showcase the explanatory power of this multidimensional framework by applying it to several case studies and showing how the informational transfers can be categorized across the three dimensions. Our approach shifts the question from whether animals can make inferences to how they perform behaviors that seem to require inferences, and which representational structures they employ. The framework offers significant methodological advantages for comparative and developmental psychology by providing a unified account of inference that acknowledges both similarities and differences in cognitive processes across species and developmental stages. 

Martin Niederl

Animal Rationality Through Higher-Order Reasons

University of Chicago

Humans are still singled out as the uniquely rational animals. This is typically understood in terms of humans’ capacity to respond to reasons. Against this philosophical orthodoxy, some philosophers have recently started to argue that certain animals can in fact respond to reasons. Out of those, Melis and Monsó (2024) argue for the most robust conclusion in the literature to date via animals’ capacity to respond to undermining epistemic defeaters. Although we are ultimately sympathetic to this idea, the present article is dedicated to highlighting two principled methodological problems. We begin by highlighting that researchers must answer both a conceptual and a methodological question regarding animals’ capacity for normative cognition in order to appease staunch skeptics. We then turn to Melis and Monsó’s answers to both questions, before providing two principled methodological challenges for studies that allege to support their point. In combination, the Acquisition Problem and the Priority Problem show that salient lower-level explanations of animals’ behavior will be available as long as the experiment involved the animals’ experience with overriding defeaters. Since this is the case in all the paradigms Melis and Monsó cite, the skeptic will remain unimpressed. Finally, we close by suggesting a possible way forward. 

Eylem Özaltum ​

Rational Animals as Linguistic Animals

Koç  University

Human beings are rational and linguistic animals. But must these two characteristics go together?  The philosophers who think so form a rather diverse set. Even the small subset I consider in this paper, namely, Descartes, Turing, Davidson, and Chomsky, do not have much in common. Nevertheless, all these authors, even the ones who are the paradigmatic defenders of rationality as a uniquely human trait, take language onlyas a sufficient condition for attributions of rationality. I want to focus on this often-not-mentioned fact. For Davidson, linguistic communication is not only sufficient as a manifestation of rationality but also, in our case, what facilitates it. For Descartes and Turing, on the other hand, language is a sufficient condition for rationality attributions but does not seem to have this constitutive role to play. For them, thinking is not necessarily a social trait. But then rationality, presumably, could also manifest in some non-social intelligent behavior. And yet, language is the only sufficient condition they give.  I compare these two different ways of thinking of language as a sufficient condition for rationality to answer the following question: What is peculiar to linguistic communication that singles it out as the mark of rationality?

Hassan Saleemi

The Mere Animal and the Rational Animal 

University of Pittsburgh

What is a rational animal? Supposing that rational animals differ in kind, and not merely in degree, from the other animals, I argue that this question must be answered in a particular way. According to additive theorists of rationality, the concept of the animal is fundamental, and then concept of the rational animal is understood derivatively. Transformative theorists of rationality have argued that the additive theorist renders the rational animal unintelligible. I think this is right. They think instead that neither concept is prior to the other. The two are on a footing as species of a genus. I think this is wrong, for this renders the mere animal unintelligible. This suggests we should turn to the remaining possibility, according to which the rational animal is the fundamental concept, and we understand the mere animal derivatively. To put the point in a slogan: we recognize them as animals as we recognize ourselves in them. 

Eric Marcus

The Difference Rationality Makes

Auburn University

According to the Transformative Theory of Rationality, rational animals possess a fundamentally different kind of mind from non-rational animals. I argue that this thesis does not conflict with any of Darwin’s insights, nor is it challenged by “empirical research on, e.g., inferential reasoning in non-human animals, reasons-evaluation in young children [or] philosophical studies highlighting the unreflective bases of rational belief and action.” On the Transformative Theory, although the non-rational and the pre-rational know things on the basis of perception and act—often with remarkable intelligence—in pursuit of their desired ends, they do not speak their minds. It is precisely this capacity for articulation that marks a qualitative transformation in the nature of knowledge, perception, action, and, indeed, the unity of the mind itself.

Paul Harris

Any questions?

Harvard University

Despite his ability to coordinate keyboard symbols to answer questions and to make requests, the chimpanzee Kanzi asked no questions. Also, naturalistic observation of chimpanzee communities has not revealed any spontaneous disposition to engage in question-like behaviors. By contrast, among typically developing human 2-year-olds, verbal question-asking is frequent and non-verbal questions (based on gesture and vocalization) are produced still earlier. Exposure to spoken language (including spoken questions) does not seem to be a precondition for question-asking because deaf children who have created their own (home)-sign system spontaneously ask questions. By implication, asking questions is an untaught, early emerging human capacity – one that enables young humans to seek information from others. Moreover, having asked a question, children respond selectively to pertinent as compared to irrelevant replies, implying that they pose questions with some expectations about what would count as an informative answer. I speculate that children’s ability to imagine as yet unknown possibilities is critical for their question asking and analyze children’s questions about the names of things, about object locations, and about mental states in support of this speculation

Sarah Beck

Associative tool use by human children and Goffin’s cockatoos: comparisons are valuable, despite methodological differences.

University of Birmingham

Close comparisons between participants of different species appear to offer us a way to evaluate relative levels of intelligence. Through careful study design researchers aim to make tasks as similar as possible and then to apply clear and fair criteria to interpret results. Here we reflect on two pairs of associative tool-use studies with young human children and Goffin’s cockatoos. In both we saw a gradual increase in success between 3-4 and 8-9 years of age for the humans, but the cockatoos’ performance differed in that they succeeded on one task but not the other. While human and cockatoo participants were presented with essentially the same task, methodological differences were present due to sample sizes and access to participants, social contexts, and, inevitably, the physical challenges present for the two species. These differences rightly prevent overly simple comparisons between human children and cockatoos in terms of ‘who is better’ at associative tool innovation. But we can use these rich data sets and observations to understand more about innovation and reflective problem solving, without necessarily seeing human thought as uniquely rational.

Andreea Miscov

Do capuchin monkeys use metacognitive judgements in virtual environments?

University of St Andrews

Authors: Andreea Miscov, Emma S. McEwen, Justin Ales, Amanda M. Seed 

Recent primate studies have highlighted how virtual environment (VE) software facilitates studying phenomena in naturalistic contexts, unrestricted by real-world space. We studied memory in capuchin monkeys in a VE delayed-response task. On a touchscreen, subjects viewed three hedge rectangles from a hilltop, one containing grapes. While walking down the hill, the hedges obscured the grapes, requiring monkeys to remember their location for varying delays. We compared ten capuchins’ short-term memory performance in the VE task, and an equivalent physical task. We found the predicted effect of delay in the VE task, and comparable performance to the physical task. Monkeys spontaneously abandoned some trials, most often in long-delay trials. In study two, we will explore this using an opting-out paradigm to disentangle metacognitive judgements from learned associative rules. Monkeys will receive two options: attempt a memory task for a high-value reward or “opt-out” by moving to a low-value reward. We will orthogonally vary delay, number of hedges, and contextual cues to create unique trials. We predict that, if monkeys make metacognitive judgements, their opting-out rates will increase with task difficulty. The comparable performance of monkeys in virtual and real environments, presents VEs as an exciting new tool for comparative memory research. 

Emma McEwen

Do they know what they are looking for?  Developmental change in children’s targeted information seeking in a new torch paradigm. 

University of St Andrews

Authors: Emma S. McEwen, Eleanor J. Jordan, & Amanda M. Seed 

As adults, we are able to evaluate specific pieces of information that we lack and selectively take corrective action to seek such information in a targeted manner.  To understand how this ability develops, we tested 3- to 6-year-old children (N = 56) in a computerised targeted information seeking task in which a squirrel needed to find a snack in 1 of 4 trees. Participants either had to seek information about the location of the snack in the tops of the trees, about spikes on the bottoms of the trees rendering them inaccessible, or about both, using a torch linked to their mouse. Children of all ages readily searched the trees and used the information to find the correct option in the conditions in which they had to find the tree containing the reward, though they performed less well when searching for information about the obstacles. Further, we found that older children tended to search more locations overall, and searching more locations had a positive effect on accuracy, suggesting that older children employed a more exhaustive and systematic strategy than younger children. This supports other research showing a difference between younger and older children in the style of their information seeking. 

Simon Brown

What is the Role of Narrative Understanding in Rationality and is it Uniquely Human? 

London School of Economics and Political Science

What is required for rational decision-making across time?  One view, the Maximisation Picture, is satisfied with reliably making decisions that tend to maximise expected utility across time.  Many animals approximate this ideal.  By contrast, the Narrative Picture holds that rationality requires explicitly conceiving of one’s well-being across time. One argument for this alternative view appeals to the alleged importance of the narrative structure of one’s life, and denies that animals can conceive of such narratives.  I defend the Maximisation Picture, arguing that we can explain the importance of narratives by understanding them as imperfect but useful cognitive tools for approximating utility maximisation. This view avoids various problems faced by the Narrative Picture (for example, narratives typically simplify and can mislead, so directly incorporating them into rationality itself seems odd).  Moreover, it explains away intuitions which appear to support the Narrative Picture, by showing such intuitions would be produced by our cognition misfiring in specific situations.  Finally, it suggests that narratives do not mark a fundamental divide between humans and other animals with respect to intertemporal rationality. 

Tobias Starzask

Ruhr University Bochum

How and why compare: Long-lasting disagreement in animal cognition research and how to solve it

Animal rationality is as debated as any other cognitive capacity in animals, be it belief, theory of mind, causal understanding, mental time travel, or consciousness. Since answers to these questions also depend on empirical observations, one would expect that new empirical data dissolves disagreement over time. However, many debates remain polarized between the romantic view that some animals have human like cognitive capacities and the killjoy view that human cognitive capacities are unique. In this talk I will discuss three potential reasons for this situation: (1) researchers subscribe to different principles of interpretations; (2) the methods to investigate animal cognition necessarily result in a problem of underdetermination, and (3), what I take to be the central problem: disagreement is largely conceptual. Focusing on the conceptual question, I’ll argue that we should not anchor our understanding of cognitive capacities in the way we find them in humans but take animal behavior as informative concerning how we should conceptualize them. This, however, creates the challenge that we want to answer whether or not animals have a certain capacity, while at the same time we don’t really know what this capacity is. 

Mike Dacey

The Purpose of Behavior and the Meaning of Thought: A new Approach to Intentionality

Bates College

I propose a view of intentionality that takes it to have first arisen in purposive behavior. The idea of purposive behavior is adapted from Tolman’s work in the 1920s: it is behavior with what he calls persistence until character. He names his brand of behaviorism “purposive behaviorism,” so I suggest “purposive functionalism.” I interpret purpose in an enactivist spirit, and argue that purposive behavior enacts a thin sort of normativity which can ground mental state functions.

This inverts priority in two common orders of analysis: the intentional meaning of (basic) mental states depends on purpose that is in behavior (not behaviors gaining meaning from intentions) and mental state functions depend on basic norms that are built into purposive behavior (not norms depending on functions). In short, purpose before intentionality, and norms before functions. Because purposive behavior occurs in living beings, biological teleology in other forms may already be present, supporting and explaining the emergence of purposive behavior. In turn, it may allow the later emergence of other kinds of mental state.

---

---

Madison Evans

Deer Xing: Adult Mule Deer Odocoileus hemionus Does Initiate Road Crossings in Anthropogenic Habitat  

Fort Lewis College

Foraging patches for mule deer Odocoileus hemionus in anthropogenic environments are often separated by roads. Interactions with humans on these trafficked areas pose threats to both species. The risky decision to cross such a barrier may be made by adults seeking better forage, neophilic fawns/juveniles, and/or a combination thereof. We distinguish between two alternative hypotheses regarding their decision-making process: one, that adult females (does) initiate road crossings; two, that fawns initiate these crossings. We find that adult does initiate the overwhelming majority of road crossings. Data was collected at Fort Lewis College, (Durango, CO, USA) for two calendar years beginning in summer 2023. Initiation of a group crossing was defined by the first individual to cross a paved road section of the environment while oriented away from 75% of the group. These results suggest that maternal foraging need, more than fawn defense, contributes to these risky group decisions. 

Andrew Bate

Riding waves of social information  

University of Leeds

Observing the decisions and actions of others provides social information that can inform your actions such as whether to follow. We consider a model where all agents simultaneously gather stochastic private information (weighted towards an unknown preference), coming to a decision once sufficiently confident. However, decisions (and indecisions) by agents are observed by all other agents and provide social information. In particular, following a decision by one agent, other agents incorporate this social information with their private information and may follow this decision if the agent becomes sufficiently confident; forming a wave of decisions. This wave (or lack thereof) provides further social information about the private information of other agents, leading to potentially more waves in response to the (in)decision. We explore small groups of agents to see if all this social information leads to quicker and/or better decisions as well as the number of waves of decisions and whether agents remain undecided afterwards. 

Mathilde Prenevost

Children display sudden, insightful problem solving before they understand insightful experiences 

University of Lancaster

Mathilde H. Prenevost¹, Ida B.R. Nilsen², Evalill Bølstad³, Francisco Pons³, Paul L. Harris⁴, & Rolf Reber^{3 1} Lancaster University, ² Stockholm University, ³ University of Oslo, ⁴ Harvard University 

Moments of insight offer a window into human and non-human problem-solving. Insight involves a sudden new understanding, often involving a shift in mental representations. Cognitive psychology has extensively studied insightful problem-solving in adults¹, while comparative research dates back to Köhler’s work with chimpanzees². However, recent work questions whether these lines of research examine the same phenomenon³. Human studies often rely on verbal reports, which are unavailable for animals. Children who cannot yet articulate their insight experiences offers a bridge between these approaches. Children (N = 160, 4-8 years) completed a problem-solving task designed to elicit aha-experiences and additional tasks assessing their understanding of such experiences. In the problem-solving task, children were shown three and asked to find a related solution word. Observed indicators of insight included: sudden answer after delay; gasp before answer; energetically shouting the answer. We also recorded self-reported aha-experiences. While the frequency of observed insights remained stable across ages, understanding of these experiences increased with age. Older children more accurately recognized and understood their aha-moments, suggesting a developmental lag between experiencing and understanding insight. These findings suggest children have insights before they can report them, offering a potential bridge between human and non-human research. 

Miguel  Borrajo Reinaldo 

Listening to Sperm Whales through AI: Epistemological Challenges and the Complexity of Nonhuman Minds  

Autonomous University of Madrid

This paper examines the epistemological assumptions behind efforts to “translate” nonhuman communication, using the Cetacean Translation Initiative (CETI) as a case study. While CETI presents itself as a conservation project, its structure and discourse are organized around the idea that understanding a species requires decoding its vocalizations into linguistic terms. This framework draws on what has been called the code model of communication: the notion that meaning is a message that can be extracted, formalized, and rendered intelligible to humans. I argue that this model imposes a reductive view of animal communication, privileging codifiable signals over relational, embodied, and affective dimensions. By focusing on translation as a technical goal, CETI reorganizes research priorities, diverts attention from immediate ecological threats, and redefines what counts as relevant knowledge. My analysis draws from philosophy of science, critical technology studies, and recent debates on animal cognition to show how translation operates not only as a tool, but as an epistemic regime. The aim is not to reject AI-based research, but to expose the limits and consequences of modelling other species according to human linguistic frameworks. What cannot be translated, in this model, risks being rendered inaudible—and unworthy of protection. 

Zeynep Genç

Norms and Counterfactuals in Development  

Authors: Zeynep Genç1 & Angela Nyhout1 1School of Psychology, University of Kent, CT2 7NP, UK 

Counterfactual reasoning involves thinking about how reality could have been different. Adults show remarkable consistency in the counterfactual possibilities they imagine, highlighting the presence of “fault lines” of reality. For instance, they tend to imagine counterfactuals that undo immoral actions. However, it remains unclear whether this link between morality and the counterfactual imagination is an inherent cognitive feature, present from early childhood. To elucidate the relation between morality and counterfactuals in development, we tested 191 4- to-11-year-olds across two studies. In Study 1, children heard stories in which a moral normviolating and a moral norm-conforming character together bring about a negative outcome. When asked what could have happened differently, children began to change the moral normviolating part of reality over other parts just after the age of 6. In Study 2, we examined whether this effect is unique to moral norm violations or extends to statistical norm violations. Children began mutating statistical norm violations only after age 9, and even 11-year-olds did so less frequently than 8-9-year-olds in Study 1 with moral violations. These findings suggest morality influences counterfactual thinking earlier than statistical norms, supporting the relation between morality and counterfactual imagination in childhood. 

Ignacio Sánchez-Moreno

Talking Dogs Buttons: Interspecies Communication or Illusion?

National University of Distance Education (UNED)

One of the main criticisms of Augmentative Interspecies Communication (AIC) systems, such as talking dog buttons, is the claim that trained use precludes genuine communicative behavior, a challenge I refer to as the “training objection.” However, this critique rests on a series of conceptual weaknesses that have often been overlooked. In this work, I argue that a neo-Gricean framework of intentional communication offers the necessary conceptual tools to distinguish communicative from non-communicative acts, while also providing operational indicators—such as attention-getting behaviors, persistence, and elaboration—to assess the presence of intentionality. While research on dogs’ social-cognitive skills and anecdotal reports suggest evidence of communicative intentionality, current experimental studies on talking dog buttons do not yet offer conclusive support. Here, I propose that incorporating these indicators into experimental design would fortify the methodological rigor of studies involving these systems. 

Marina Bazhydai

Curiosity-driven learning in infancy: insights from gaze-contingent eye-tracking paradigms

University of Lancaster

The cognitive mechanisms and benefits of curiosity-driven learning in infancy are poorly understood. In two studies, using gaze-contingent eye-tracking paradigms, we investigated infants’ moment-to-moment information sampling and the effect of active sampling on novel word learning. In Study 1 (Altmann et al., 2025), 10-12-month-old infants could freely explore two novel categories, triggering a new exemplar from a category by fixating on either of the two associated areas on a computer screen. This gaze-contingent design enabled us to distinguish between exploration – switching from one category to another – and exploitation – consecutively triggering exemplars from the same category. Infants’ moment-to-moment sampling choices were non-random but guided by the infants’ exploration history. Self-generated sequences grouped into three clusters of brief yet explorative, longer exploitative, and overall more balanced sampling patterns. In Study 2 (Bazhydai et al., under review), 20-23-month-olds assigned to a Curiosity condition were given the opportunity to structure their own information seeking, while infants in two other conditions engaged in learning novel words passively. The results indicate no advantage of active information selection on retention of novel words above and beyond passive learning, with infants across all conditions retaining novel words above chance. These findings offer novel insights into infants’ curiosity-driven exploration and learning.

---