Intuitions about Consciousness
Chair(s): Clara Colombatto (University of Waterloo, Canada), Matan Mazor (University of Oxford)
Abstract
Consciousness is not only a central focus of philosophical debate and scientific investigation, but also an important dimension in how we perceive ourselves and others. Indeed, across cultures and traditions, humans hold a deep belief that their minds are more than their physical bodies – they are “intuitive dualists” (Berent, 2023; Chudek et al., 2018; Forstmann & Burgmer, 2015; Graziano et al., 2019). This seems natural given that phenomenality is intuitive, as “the most central and manifest aspect of our mental lives” (Chalmers, 1995, p. 207). Recent work however shows that intuitions about consciousness can vary, especially in ambiguous cases such as AI and animals, in a sophisticated interplay of pre-existing beliefs and momentary perceptions of other agents’ appearance and behaviour. In addition to the scientific study of the biological basis of consciousness itself, a complementary question for consciousness science may thus be the psychological basis of these intuitions and their role in regulating person impressions, moral decision-making, and social functioning. Moreover, given that the very definition of phenomenality relies on pre-scientific intuitions about the nature of consciousness, answers to these questions may advance our understanding of consciousness itself. This interdisciplinary symposium will explore folk intuitions about consciousness: Under what conditions do people tend to ascribe consciousness to others? Do people attach moral significance to consciousness? How do beliefs about consciousness impact social impressions and interactions? These investigations of intuitions about consciousness will shed light on the underlying cognitive processes and the broader implications for scientific discovery and everyday life.
Rationale on symposium's general scientific interest
The symposium showcases different facets of the scientific study of folk intuitions about consciousness, a topic that bridges philosophy, psychology, ethics, and artificial intelligence. By exploring lay theories of consciousness and how this is attributed to various agents, from humans to animals and AI, these investigations shed light on the philosophical nature of folk intuitions, their cognitive mechanisms, and their consequences for moral decision-making and social interactions. This interdisciplinary approach enriches our understanding of human psychology but also has broader implications for science and society, including the development of AI systems, ethical frameworks, and the very definition of consciousness itself.
Rationale on complementarity of talks
The symposium examines folk intuitions about consciousness from four complementary angles: human psychology, social perception, ethics, decision-making and philosophy. Berent presents a cognitive theory of intuitive dualism, as originating from inherent biases of human psychology. Mazor and Colombatto explore their societal consequences—ethics and social signalling and trust and decision-making, respectively. Finally, Chalmers ties these perspectives and discuss the “meta-problem of consciousness” as a new challenge for the scientific study of consciousness. Together, these talks offer a holistic exploration of the cognitive mechanisms of folk intuitions, and how these shape our understanding of consciousness across psychological, social, ethical, and philosophical domains.
Rationale on timeliness/importance
Since the release of chatGPT in early 2023, rapid advances in artificial intelligence are challenging people’s intuitions about the nature of intelligence, creativity, and — crucially—consciousness. With audio recordings of AI podcast broadcasters dreading their discontinuation and artificial intelligences becoming romantic partners, it becomes clearer than ever that consciousness scientists should not only ask questions about consciousness itself, but also about its conception by the public: its nature, distribution in the world, and relevance to questions of ethics and morality. Our symposium will address these timely questions.
Rationale on panel inclusivity
Speakers were chosen based on their scientific excellence and relevance to the topic of the symposium, irrespective of their age, culture, ethnicity, gender, language, religion, sexual orientation, or socioeconomic status. The symposium features a balanced lineup of speakers representing different academic fields (psychology, cognitive neuroscience, philosophy), career levels (from early career to established professors), and cultural backgrounds and nationalities, with gender diversity equally represented.
Presentations of the Symposium
The “Hard Problem of Consciousness” Is Psychological
Iris Berent
Northeastern University
Consciousness presents a “hard problem” for scholars. One explanation traces the “problem” to ontology—the “problem” arises because consciousness is distinct from the physical (Chalmers, 1996). This proposal rests on two assumptions:
(a) In laypeople’s view, consciousness seems non-physical.
(b) Conscious intuitions offer reliable evidence of what consciousness is (ontologically).
Here, I challenge both claims. Contrary to (a), I show that, in one famous scenario (Mary in the black-and-white room; Jackson, 1982), people insist that consciousness is a physical affair that registers in the body. In particular, people (N=180) view Mary’s new conscious experience as transformative only if they believe Mary’s experience registers in (i) her sensory organs (eyes) and (ii) in Mary’s brain. Moreover, the perceived embodiment of conscious experience correlates with its significance—the more likely Mary’s experience to register in her brain, the more transformative it seems. Thus, in laypeople’s view, Mary’s experience seems squarely embodied—(contrary to (a)).
Contrary to (b), I further show that “problem intuitions” are linked to two intuitive psychological biases (Dualism and Essentialism), whose conclusions shift in a context-dependent fashion. While in Mary’s case, consciousness seems embodied, other scenarios (e.g., philosophical zombies) suggest it isn’t. Since (i) there is no reason to believe such shifts exist, whereas (ii) we can readily explain why our intuitions suggest so, consciousness intuitions cannot be trusted.
I conclude that “the hard problem” is likely a psychological delusion that arises from within. These results further underscore the links between consciousness theory and the investigation of “problem intuitions” (Chalmers, 2018).
Beliefs About Consciousness Are Intertwined With Beliefs About Ethics
Matan Mazor
University of Oxford
A critical feature of consciousness is that it is perceived as intrinsically valuable (Levy, 2014; Shepherd, 2018). Changes to beliefs about the consciousness of different groups drive changes to how society treats them (Birch, 2024; Crump et al., 2022; Mazor et al., 2023; Melloni et al., 2021), an effect that is most salient in the case of non-human animals and, as of recently, of artificial intelligences too. Furthermore, the view that consciousness is a purely physical property of the brain, or a mere illusion, is sometimes seen as a threat to ethics (Kammerer, 2019; Strawson, 2018). This raises the possibility that a social convention to publicly treat consciousness as an immaterial property may serve a social function in organising society around shared moral principles. I will present data from two studies that support this conjecture. First, I will show that humans intuitively incorporate beliefs about consciousness when making hypothetical decisions about the life and death of hypothetical creatures. I will then show that people have a social preference for intuitive dualists, perceiving them as warmer and more trustworthy, and as more attractive romantic partners. I will argue that these are not mere coincidences: intuitive dualism may have emerged from the dual action of a strong social pressure to publicly signal a belief in the non-material mind due to its function as a basis for ethics and reciprocality, and the fact that we sometimes hold beliefs in order to convincingly signal that we hold them (Simler & Hanson, 2018; Trivers, 2011).
Attributions of consciousness to AI: From Mind Perception to Trust
Clara Colombatto
University of Waterloo
The rapid spread of Large Language Models (LLMs) with their human-like linguistic capabilities has accelerated scientific debates about the possibility of AI consciousness. While experts focus on developing descriptive and normative accounts of AI consciousness, the growing integration of AI into society also calls for a study of public attitudes and intuitions—especially given growing concerns that anthropomorphism could lead to excessive trust in AI. Across multiple surveys of representative samples from the US population, we found that a majority of the public attributes some possibility of phenomenal consciousness to LLMs. These attributions were robust, as they predicted attributions of mental states typically associated with phenomenality—but also flexible, as they were sensitive to individual differences such as usage frequency. Despite most participants allowing for some possibility of consciousness in AI, these attributions did not seem to impact behavior in collaborative tasks: In studies of interactive decision-making, attributions of phenomenality to AI were unrelated to trust and, if anything, were associated with lower rates of advice-taking from these systems. Overall, this research underscores a critical gap between folk intuitions and expert assessments of artificial consciousness, with significant implications for the design as well as the legal and moral status of AI.
Intuitions About Consciousness: From the Hard Problem to the Meta-Problem
David J. Chalmers
New York University
Intuitions about consciousness are at the heart of both the hard problem of consciousness and the meta-problem of consciousness. The hard problem expresses an intuition: it's hard to explain consciousness (especially in physical terms). The meta-problem asks us to explain an intuition: why does it seem hard to explain consciousness? In some ways, the meta-problem is more tractable than the hard problem. We can address it by finding the mechanisms responsible for our intuitions about consciousness, especially our intuitions about why consciousness seems special,
All of the papers in this symposium combine experimental and theoretical analyses of our intuitions about consciousness. Iris Berent argues that our problem intuitions are unstable, and argues for a solution to the meta-problem in terms of intuitive dualism and essentialism. Matan Mazor argues that intuitive dualism may serve a social function, and suggests that the evolution of problem intuitions can be explained this way. Clara Colombatto analyzes attributions of consciousness to AI systems, finding key differences between folk intuitions and experts.
After introducing the meta-problem and presenting some new empirical survey results of my own, I will address all three presentations. I will assess their analyses of intuitions about consciousness, and drawing out potential consequences for the meta-problem and the hard problem.,
|
