Kristina Šekrst

New Trends in Cognitive Science syllabus

New Trends in Cognitive Science

A graduate reading seminar on minds in places most people would not even consider: humans with no inner voice, bees that play, crows that count, plants that click, fungi that remember, neurons in a dish, and our own brains on ChatGPT. Every paper is from 2020 or later. Taught with Sandro Skansi at the Center for Cognitive Science, University of Zagreb.

Course information

How to pass the course

Students select a paper from the reading list, prepare it, and present it to the class in the week when that paper is scheduled. If a paper feels too challenging, ask your instructors for guidance and use any secondary literature available. When presenting, the goal is to explain what the researchers did and how they came to their conclusions, including the strongest evidence and the weakest aspects, a skill that will serve your own critical analysis of scientific and non-scientific arguments later. You finish your presentation by posing questions for the class to debate.

Papers are chosen at the beginning of the semester so that each unit has at least one presenter. If two readings are joined by and, they are treated as a single paper (these are usually very short follow-ups). Length can mislead: some of the shortest papers are the hardest to present, because word limits force the authors to squeeze a whole study into a few pages and move the details into long supplementary analyses. Longer papers take more time but explain the background, design and findings step by step, so they are often the easier choice.

Participation and discussion are graded throughout the semester. You are expected to get acquainted with the assigned papers before every class, at least enough to follow the main argument, and to leave the details to the presenters.

  • Paper presentation: 50%
  • Participation and discussion (throughout the semester): 30%
  • Final oral exam: 20%

The final exam

The final exam will be unusual, but the coursework we do is, after all, unusual. A mission has returned from beneath the icy surface of Enceladus, bringing back more than just samples. Each of you will receive a dossier about one of the creatures found there. The dossier contains everything the research team observed before the funding ran out (academic reality): body structure, how it moves, its behaviour around food, its response to injuries, what happens when it meets others of its kind, how it communicates. Since the team could not agree on what the creature is, whether it deserves any rights, whether it is sentient, conscious, or even has something close to a mind, the decision is now up to you.

You will present your findings to the Interplanetary Committee for Strange Life Forms (Sandro and Kristina). Decide first whether your creature shows signs of cognition or sentience, and whether there is any reason to think it is conscious. Then consider whether it deserves protection, should be given rights, and what should follow: further study, sending it back, protection under an act. Base your decision on evidence, using everything covered in the course. You may ask the Committee to run one more experiment on the creature, as long as you explain what result you expect and why it would matter for your decision. The probe has a limited amount of fuel, and we spent all our money on AI, so an adequate response needs to show what the evidence supports, the limits of that evidence, the circumstances that would make you change your mind, and the risks involved if you are wrong either way. The Committee will not be responsible for your decisions.

About the readings

Each week you will find recent research papers, published from 2020 onward, on topics that are still being argued about. You are not expected to read every paper in full, though if you do, you will get much more out of the discussions. Before each class, read at least one paper from the unit in full so you can join the discussion, and look at the abstracts of the others. If you are presenting a paper, read it carefully from start to finish.

Research papers can be technical, so you can skip the detailed statistics and methods and focus on the introduction, main results and discussion; we review the key points together in class. Most papers are free to access through the DOI links, you have institutional access for the others, and every PDF is in the Omega courseware. The sequence and contents of the units may change during the semester to suit your background and the pace of the course; any changes are announced in class and on Omega in advance.

What will you get from this course?

By the end of the semester, you will have looked at minds in places most people would not even consider. You will study people who have no inner voice or cannot form mental images, and women who can smell without olfactory bulbs. You will observe bees that play for the sake of playing, crows that can count, octopuses whose arms send signals to one another, and whales and elephants that respond to their names. You will learn about fungi and mold that appear to remember where food was, and about neurons in a dish that learn to play video games.

You will learn how researchers design experiments for people and for organisms that cannot respond in words, and you will see how machine learning finds patterns in plant sounds, elephant rumbles, monkey calls or whale clicks. No prior knowledge of biology or programming is needed; each method is explained in class.

The units cover the newest research in four domains: humans, animals, plants and fungi, and machines. Most recent trends point towards non-standard modes of cognition, in animals, plants, molds, fungi, organoids, and possibly machines. You will be able to evaluate assertions about minds in such creatures or objects, examine what is actually measured, consider other explanations, and ask what evidence would support the claim, and you will know how to proceed when the evidence is uncertain. Since AI is omnipresent, we finish by studying how its use and overuse affects our brain. These skills carry over to your thesis, to scientific papers, to any critical work in cognitive science, and to the everyday evaluation of claims in the media.

Course units

  1. Brains that break the rules
    Is the usual view of the mind maybe just an average of a wide spectrum of varieties?
    We explore people whose brains do not fit standard patterns. Some women, especially left-handed ones, have no visible olfactory bulbs on MRI scans but smell normally. Some people cannot create voluntary visual imagery; others have no inner voice. We also examine critically how much we can trust personal reports about inner experience.

    • Zeman, A. et al. (2020). Phantasia: The psychological significance of lifelong visual imagery vividness extremes. Cortex, 130, 426–440. · DOI
    • Nedergaard, J. S. K., & Lupyan, G. (2024). Not everybody has an inner voice: Behavioral consequences of anendophasia. Psychological Science, 35(7), 780–797. · DOI
    • Weiss, T. et al. (2020). Human olfaction without apparent olfactory bulbs. Neuron, 105(1), 35–45. · DOI
    • de Haan, E. H. F. et al. (2020). Split-brain: What we know now and why this is important for understanding consciousness. Neuropsychology Review, 30, 224–233. · DOI
  2. Dreams you can talk to
    What are sleep and dreaming for, and how could we ever find out what goes on in the mind of a sleeping creature that cannot communicate with us?
    We examine what dreams are and whether other animals might have them. People in the middle of a lucid dream were given questions and simple maths problems and replied with eye movements and facial muscle signals. We look at a machine learning idea on which dreams resemble the noise added during neural network training, helping the brain avoid overfitting to the day’s experiences. Then we move beyond humans: octopuses have short phases of active sleep during which their skin displays patterns, as it does when they are awake, and even hydra, with no brain, shows a sleep-like state.

    • Konkoly, K. R. et al. (2021). Real-time dialogue between experimenters and dreamers during REM sleep. Current Biology, 31(7), 1417–1427. · DOI
    • Hoel, E. (2021). The overfitted brain: Dreams evolved to assist generalization. Patterns, 2(5), 100244. · DOI
    • Pophale, A. et al. (2023). Wake-like skin patterning and neural activity during octopus sleep. Nature, 619, 129–134. · DOI
    • Kanaya, H. J. et al. (2020). A sleep-like state in Hydra unravels conserved sleep mechanisms during the evolutionary development of the central nervous system. Science Advances, 6(41), eabb9415. · DOI
  3. Do we think in words?
    Is language a necessity for thinking?
    When people do maths, solve logical puzzles, listen to music or work out what others believe, the language areas of the brain are generally inactive, and people who lose language after a stroke can still do mathematics and play chess. Large language models, meanwhile, learn grammar and text from data but have trouble with reasoning, which raises the question of whether they challenge Chomsky’s innate grammar. Are language and thought separate?

    • Fedorenko, E., Piantadosi, S. T., & Gibson, E. A. F. (2024). Language is primarily a tool for communication rather than thought. Nature, 630, 575–586. · DOI
    • Mahowald, K. et al. (2024). Dissociating language and thought in large language models. Trends in Cognitive Sciences, 28(6), 517–540. · DOI
  4. Babies and machines
    Which aspects of human knowledge can be acquired through experience alone, and which have to be built in from the beginning?
    What can be learned from a single child’s experience? Headcam footage from a toddler was used to train neural networks, which learned to recognise objects, including broad visual categories, with no prior knowledge. Yet eleven-month-old infants understand that agents act towards goals, which the models failed to grasp, and children readily invent new uses for tools and discover new cause-and-effect relations.

    • Orhan, A. E., & Lake, B. M. (2024). Learning high-level visual representations from a child’s perspective without strong inductive biases. Nature Machine Intelligence, 6, 271–283. · DOI
    • Stojnić, G., Gandhi, K., Yasuda, S., Lake, B. M., & Dillon, M. R. (2023). Commonsense psychology in human infants and machines. Cognition, 235, 105406. · DOI
    • Yiu, E., Kosoy, E., & Gopnik, A. (2024). Transmission versus truth, imitation versus innovation: What children can do that large language and language-and-vision models cannot (yet). Perspectives on Psychological Science, 19(5), 874–883. · DOI
  5. Decoding animal communication with machine learning
    Do other animals possess the basic components of language, and can machine learning detect structure and meaning in forms of communication that we do not understand?
    Computational linguistics applied to animal calls. Sperm whales communicate with clicks known as codas, and machine learning showed that whales respond to and copy one another; we look at Project CETI’s roadmap for a whale-sound dataset and its basic units and structures. Elephants and marmoset monkeys use name-like calls that machine learning techniques could identify, confirmed by strong reactions to calls addressed to them. Bonobo calls also occur in compositional combinations, some of which change the meaning of the other.

    • Sharma, P. et al. (2024). Contextual and combinatorial structure in sperm whale vocalisations. Nature Communications, 15, 3617. · DOI
    • Andreas, J. et al. (2022). Toward understanding the communication in sperm whales. iScience, 25(6), 104393. · DOI
    • Pardo, M. A. et al. (2024). African elephants address one another with individually specific name-like calls. Nature Ecology & Evolution, 8, 1353–1364. · DOI
    • Berthet, M., Surbeck, M., & Townsend, S. W. (2025). Extensive compositionality in the vocal system of bonobos. Science, 388(6742), 104–108. · DOI
  6. Bees that play
    Can insects play, and what kind of evidence would show that a tiny brain has something like a mind?
    What goes on in a brain with fewer than a million neurons. Bumblebees roll small wooden balls with no food reward, and younger bees do it more often than older ones, as we do. Bees recognise shapes by touch after only seeing them, and by sight after only touching them, a clue towards Molyneux’s problem. They also learn from one another to open a two-step puzzle box that none of them manages alone, even after 24 days of trying.

    • Galpayage Dona, H. S., Solvi, C., Kowalewska, A., Mäkelä, K., MaBouDi, H., & Chittka, L. (2022). Do bumble bees play? Animal Behaviour, 194, 239–251. · DOI
    • Solvi, C., Gutierrez Al-Khudhairy, S., & Chittka, L. (2020). Bumble bees display cross-modal object recognition between visual and tactile senses. Science, 367(6480), 910–912. · DOI
    • Bridges, A. D. et al. (2024). Bumblebees socially learn behaviour too complex to innovate alone. Nature, 627, 572–578. · DOI
  7. Clever without a cortex
    Can intelligence, or even consciousness, arise in brains built very differently from ours?
    Crows and cephalopods, the favourite animals of consciousness researchers, developed extremely complex behaviour without the layered cerebral cortex of mammals. Crows make a fixed number of calls when shown a number, and the sound of the first call hints at how many will follow, so they plan ahead. Cuttlefish pass a version of the marshmallow test by waiting up to two minutes for a better treat, and those that wait longest learn fastest. In octopuses, arms pass signals to one another without going through the brain.

    • Liao, D. A., Brecht, K. F., Veit, L., & Nieder, A. (2024). Crows “count” the number of self-generated vocalizations. Science, 384(6698), 874–877. · DOI
    • Schnell, A. K., Boeckle, M., Rivera, M., Clayton, N. S., & Hanlon, R. T. (2021). Cuttlefish exert self-control in a delay of gratification task. Proceedings of the Royal Society B, 288(1946), 20203161. · DOI
    • Kuuspalu, A., Cody, S., & Hale, M. E. (2022). Multiple nerve cords connect the arms of octopuses, providing alternative paths for inter-arm signaling. Current Biology, 32(24), 5415–5421.e3. · DOI
    • Schnell, A. K., Browning, H., Crump, A., Burn, C. C., & Birch, J. (2026). Sentience in cephalopod molluscs: An updated assessment. Biological Reviews, 101, 1311–1333. · DOI
  8. Who can suffer?
    How can we decide which animals are capable of feeling pain, and what should we do when the evidence is uncertain?
    Scientists assess animal feeling through details of perception, a sense of self (the mirror test), how information is used over time, how the animal evaluates what is good or bad for itself, and similar methods. The framework applied to cephalopods has since been used on crabs, lobsters and insects: true crabs show signs of sentience, and cockroaches and flies satisfy six of eight criteria for pain. We look at the UK Animal Welfare (Sentience) Act 2022 and the New York Declaration on Animal Consciousness, which states a realistic possibility of consciousness in all vertebrates and many invertebrates.

    • Birch, J., Schnell, A. K., & Clayton, N. S. (2020). Dimensions of animal consciousness. Trends in Cognitive Sciences, 24(10), 789–801. · DOI
    • Crump, A., Browning, H., Schnell, A. K., Burn, C., & Birch, J. (2022). Sentience in decapod crustaceans: A general framework and review of the evidence. Animal Sentience, 7(32), 1. · DOI
    • Gibbons, M., Crump, A., Barrett, M., Sarlak, S., Birch, J., & Chittka, L. (2022). Can insects feel pain? A review of the neural and behavioural evidence. Advances in Insect Physiology, 63, 155–229. · DOI
    • Kohda, M. et al. (2023). Cleaner fish recognize self in a mirror via self-face recognition like humans. PNAS, 120(7), e2208420120. · DOI
  9. Plant minds and the pea controversy
    Can an organism with no neurons learn, or perhaps even feel? What sort of evidence would lead us to accept this?
    Some researchers hold that plants show abilities once thought specific to animals, and that if plants have even basic cognitive abilities it is worth asking whether they could be conscious. Plant biologists have examined twelve key assertions about plant consciousness and rejected each, on the grounds that consciousness requires learning through reward and punishment and some internal map of the environment, neither of which plants have. The debate sharpened around a well-known experiment in which pea seedlings in a Y-maze appeared to learn that a fan meant light; a replication found no evidence of learning. We also look at tomato and tobacco plants, which emit ultrasonic clicks when dehydrated or cut, which a machine learning model can read for dryness or injury, and which some animals might hear. Finally, we ask whether a signal that carries information counts as communication when no organism has been shown to respond to it.

    • Segundo-Ortin, M., & Calvo, P. (2022). Consciousness and cognition in plants. WIREs Cognitive Science, 13(2), e1578. · DOI
    • Mallatt, J., Blatt, M. R., Draguhn, A., Robinson, D. G., & Taiz, L. (2021). Debunking a myth: Plant consciousness. Protoplasma, 258(3), 459–476. · DOI
    • Markel, K. (2020). Lack of evidence for associative learning in pea plants. eLife, 9, e57614. · DOI and (it’s 2 paragraphs, don’t worry)
    • ↳ Gagliano, M., Vyazovskiy, V. V., Borbély, A. A., Depczynski, M., & Radford, B. (2020). Comment on “Lack of evidence for associative learning in pea plants”. eLife, 9, e61141. · DOI
    • Khait, I. et al. (2023). Sounds emitted by plants under stress are airborne and informative. Cell, 186(7), 1328–1336. · DOI
  10. Fungal memory and slime mold networks
    Can memory exist without a brain, and what exactly do we mean when we say that a fungus or a slime mold has memory?
    Organisms that store information in their physical structure. A mycelium grew from a piece of wood towards a new, larger block; when the original block was moved to fresh soil, the fungus regrew mainly from the side that had faced the new wood, as if it remembered the direction. Fungi also produce spikes of electrical activity: researchers have divided the spikes into “words” and compared them to human language, and have grown mycelium into the electronics of two robots, using the spikes to drive them and to change their movement under ultraviolet light. As a framework, we apply the idea that mycelial networks meet four criteria for minimal cognition through a kind of memory that needs no internal representations.

    • Fukasawa, Y., Savoury, M., & Boddy, L. (2020). Ecological memory and relocation decisions in fungal mycelial networks: Responses to quantity and location of new resources. The ISME Journal, 14(2), 380–388. · DOI
    • Šekrst, K. (2026). Fungal memory and minimal cognition. Topics in Cognitive Science. · DOI
    • Adamatzky, A. (2022). Language of fungi derived from their electrical spiking activity. Royal Society Open Science, 9(4), 211926. · DOI
    • Mishra, A. K. et al. (2024). Sensorimotor control of robots mediated by electrophysiological measurements of fungal mycelia. Science Robotics, 9(93), eadk8019. · DOI

    This week has a paper by one of the instructors. It may seem unfair to defend or attack it, but she suggests doing so anyway, since she knows where its strong and weak points are. Science involves forming decisions on evidence and not on authority, as the pea controversy shows.

  11. Minds in a dish
    If living cells are taken out of a body and put into a different environment, can they learn or “feel”? Is cognition a spectrum?
    Lab-grown biological systems. In the DishBrain experiments, human and mouse neurons on an electrode array were linked to a simulated game of Pong: when the paddle hit the ball the neurons received predictable signals, and when it missed, random ones. Over time the neurons got better at hitting the ball, and cultures without feedback did not improve. The researchers called this sentience, defined as the ability to sense and respond to input, which caused a controversy. We also examine a roadmap for “organoid intelligence”, which proposes three-dimensional brain organoids as biological computers and takes up the ethical issues, and xenobots, small groups of frog stem cells that push loose cells into piles that grow into new xenobots.

    • Kagan, B. J. et al. (2022). In vitro neurons learn and exhibit sentience when embodied in a simulated game-world. Neuron, 110(23), 3952–3969. · DOI and (it’s only 2 pages)
    • ↳ Balci, F., Ben Hamed, S., Boraud, T., et al. (2023). A response to claims of emergent intelligence and sentience in a dish. Neuron, 111(5), 604–605. · DOI
    • Smirnova, L. et al. (2023). Organoid intelligence (OI): The new frontier in biocomputing and intelligence-in-a-dish. Frontiers in Science, 1, 1017235. · DOI
    • Kriegman, S., Blackiston, D., Levin, M., & Bongard, J. (2021). Kinematic self-replication in reconfigurable organisms. PNAS, 118(49), e2112672118. · DOI
  12. Your brain on ChatGPT
    Does the use of AI tools expand our mental capabilities, or make us overly dependent on them?
    We end with our own cognition. On Clark’s view, humans have always used fingers, notes, sketches and devices to aid thinking, and generative AI continues that line. We weigh this against two studies. In the first, 91 university students researched a scientific question with either ChatGPT or Google: the ChatGPT group found the task less demanding but produced markedly weaker reasoning and justifications. In the second, 319 knowledge workers gave 936 examples of AI use at work: those who trusted AI more thought less critically about its output, while those more confident in their own abilities thought more critically. We discuss whether reduced effort is cognitive debt or simply what a well-functioning extended mind looks like.

    • Clark, A. (2025). Extending minds with generative AI. Nature Communications, 16, 4627. · DOI
    • Stadler, M., Bannert, M., & Sailer, M. (2024). Cognitive ease at a cost: LLMs reduce mental effort but compromise depth in student scientific inquiry. Computers in Human Behavior, 160, 108386. · DOI
    • Lee, H.-P. et al. (2025). The impact of generative AI on critical thinking: Self-reported reductions in cognitive effort and confidence effects from a survey of knowledge workers. Proceedings of CHI ’25, Article 1121, 1–22. · DOI
  13. Planet of the apes (and lemurs)
    How much of the human mind is already present in our closest relatives, and how far back in the primate family tree does it reach?
    At the end of the course we are joined by Dr Elena Račevska, the first Croatian primatologist, who studied lemurs in the littoral forests of Madagascar and personality and social dynamics in zoo gorillas. No reading: spend some of that time thinking about how much of yourself you expect to find in a lemur, a gorilla, or a bonobo.