# Surprising Facts About Octopus Cognition

> 99 evidence-aware findings from The Octopus Mind
> Canonical: https://octopuscognition.org/surprising-facts/

## [Neuroanatomy & the Distributed Nervous System](https://octopuscognition.org/sections/neuroanatomy-the-distributed-nervous-system/index.md)

- Two-thirds of an octopus's neurons are in its arms, not its brain — the arms can taste-by-touch, decide, and react locally in under 100 ms without consulting the central brain.
- The vertical lobe alone holds 25 million neurons — more than half of the entire supraesophageal mass — packed into 14% of its volume.
- Octopus vertical-lobe LTP is NMDA-receptor-independent and presynaptically expressed, unlike the canonical mammalian mechanism, and is maintained by a self-sustaining nitric-oxide 'molecular switch.'
- In the VL connectome, 89% of neurons (SAMs) each receive input from just a single frontal-lobe axon — an extreme sparse fan-out architecture.
- The octopus VL is a case of convergent evolution: it solves associative memory with the same three-layer fan-out logic as insect mushroom bodies and the vertebrate cerebellum, but with a completely different, independently evolved circuit.

## [Embodied Cognition and Autonomous Arm Control in Octopuses](https://octopuscognition.org/sections/embodied-cognition-and-autonomous-arm-control-in-octopuses/index.md)

- A severed, brain-disconnected octopus arm still produces a near-normal reaching movement when stimulated—the reach 'program' lives in the arm, not the brain (Sumbre et al. 2001).
- To fetch food, the soft arm temporarily builds a jointed, elbow-like structure with three dynamic joints, and the middle 'elbow' forms exactly where two muscle-activation waves collide (Sumbre et al. 2005/2006).
- Octopuses and humans converged on the same joint-level, quasi-articulated control strategy for point-to-point reaching despite 500 million years of separate evolution.
- Octopus crawling has no rhythm and no gait: Fourier analysis finds no periodicity, and the animal can crawl in any direction independent of which way its body faces, with no preferred 'lead' arm ('push right, go left').
- The long-standing textbook claim that octopuses lack proprioception was overturned in 2020—the central brain does read arm-position information, just not in the way vertebrates do.
- The catchy 'octopus has nine brains' is now considered misleading; the researcher who tested it reframes it as 'one brain and eight very clever arms.'

## [Learning, Memory & Reversal Learning in Octopus](https://octopuscognition.org/sections/learning-memory-reversal-learning-in-octopus/index.md)

- Octopus vertical-lobe LTP is strikingly hippocampus-like yet NMDA-receptor-INDEPENDENT, instead relying on a nitric-oxide 'molecular memory switch' — convergent function, different molecular hardware.
- In reversal learning, octopuses given only positive reinforcement often fail to learn; adding an explicit 'wrong-choice' signal is what unlocks flexible learning (Bublitz et al. 2021).
- The VL connectome shows a 1:12 'fan-out' where each amacrine interneuron gets just a SINGLE input — the opposite of the convergent 'fan-in' seen in the cerebellum and insect mushroom body, implying independently evolved associative circuitry.
- Blocking OR over-driving (saturating) VL plasticity both wreck next-day memory while having opposite effects on same-day learning speed — a clean dissociation of acquisition from consolidation.
- Octopus arms cannot learn object shape or weight by touch — the tactile system encodes only the proportion of receptors firing, so it discriminates texture but is 'blind' to geometry.
- Tactile memory is genuinely long-term: measurable retention persists for months, decaying only 50% at 24 days and 90% at 96 days.

## [Problem Solving & Tool Use](https://octopuscognition.org/sections/problem-solving-tool-use/index.md)

- The coconut octopus adopts a slower, clumsier 'stilt-walking' gait specifically to carry shells—paying an immediate locomotor and predation cost for a shelter it can only use later, the crux of the 'foresight' argument
- Giant Pacific octopuses learn to open human childproof medication bottles (a push-and-turn task) unaided, improving from 55 minutes to 5 minutes
- Gloomy octopuses throw silt and shells with siphon jets and sometimes appear to aim at other octopuses, who occasionally duck—one of very few non-human cases of possibly targeted throwing
- Octopuses can unscrew jar lids from the inside
- 'Inky' the octopus escaped a New Zealand aquarium by crossing the floor and squeezing down a 50 m drainpipe to the ocean

## [Observational Learning & Cognition Controversies in Octopuses](https://octopuscognition.org/sections/observational-learning-cognition-controversies-in-octopuses/index.md)

- Observers reportedly learned FASTER than the demonstrators who had undergone full operant conditioning — a striking, much-quoted claim from the 1992 paper.
- The original author (Fiorito) and his sharpest critics (Biederman & Davey) later co-authored a 1998 follow-up that failed to find a preexposure benefit — a rare adversarial collaboration in comparative cognition.
- Octopuses are asocial, semelparous, and provide no parental care, so a genuine social-learning capacity would be evolutionarily paradoxical.
- Even neurally immature cuttlefish hatchlings (≤5 days) appear to socially modulate predatory behavior, with more observers than demonstrators reaching criterion.

## [Play Behavior and Individual Personality in Octopuses](https://octopuscognition.org/sections/play-behavior-and-individual-personality-in-octopuses/index.md)

- Octopuses were the first invertebrates ever shown to have consistent individual personalities (Mather & Anderson 1993) and the first shown to play (1999) — both from the same aquarist-scientist duo.
- In the founding play study only 2 of 8 octopuses actually played, by jetting water to bounce a pill bottle against the tank current — play was individual, not species-typical.
- Play reliably appears only AFTER exploration is exhausted (days 3–6 in Kuba et al. 2006), supporting the idea that curiosity must be satisfied before an animal 'plays'.
- Play was unaffected by hunger, age, or sex — arguing it is not disguised foraging.
- The bold–shy axis is context-specific in cephalopods: an animal bold under threat is not necessarily bold when feeding, undermining the notion of a single generalizable personality type.
- Juvenile octopus temperament changes with age and carries a heritable (relatedness) signal, despite octopuses being essentially solitary with no parental care.

## [Social Cognition, Octopolis & Signaling](https://octopuscognition.org/sections/social-cognition-octopolis-signaling/index.md)

- A supposedly solitary invertebrate settles disputes with graded color signals: dark = aggressive, pale = submissive, following an almost game-theoretic escalation logic (dark-vs-dark fights; dark-vs-pale ends in retreat).
- Octopuses throw silt and shells using a siphon jet fired from under the arm web, and throws are more vigorous and more likely to hit others in social contexts - one of the only cases of projectile-throwing at conspecifics outside social mammals.
- Females did nearly all the throwing (90 of 101 throws), and dark-bodied (aggressive-state) animals threw hardest (p<0.0001).
- Octopus cyanea punches partner fish to enforce cooperation, and 3D tracking shows leadership is distributed - fish scout and steer while the octopus controls movement timing, not a simple octopus dictatorship.
- Octopolis was seeded by a single human-made metal object; Octlantis proved the phenomenon occurs naturally around rock outcrops too.

## [Camouflage, Skin Vision & Sensory Cognition](https://octopuscognition.org/sections/camouflage-skin-vision-sensory-cognition/index.md)

- Octopuses are, by all eye-based tests, colorblind (single 480 nm pigment) yet produce near-perfect color camouflage.
- Octopus skin contains the same light-sensing opsin as the eye and expands chromatophores to light with no brain involvement (LACE).
- Chromatophores are muscle-driven organs wired directly to brain motor neurons with no synapse, enabling 50 ms color changes.
- Iridophores generate blues and greens the pigment cells cannot, via acetylcholine-triggered conformational changes in the protein reflectin.
- The 'taste-by-touch' receptors evolved from nicotinic acetylcholine receptors but are gated by greasy, water-insoluble molecules instead of neurotransmitter.
- Cuttlefish polarization acuity (10° e-vector) may form a secret communication channel invisible to predators.

## [Sleep, Two-Stage Sleep, and Possible Dreaming in Octopuses](https://octopuscognition.org/sections/sleep-two-stage-sleep-and-possible-dreaming-in-octopuses/index.md)

- Active-sleep LFP brain activity in the octopus is nearly indistinguishable from waking activity (Pearson R up to 0.95 in high-frequency bands), yet arousal thresholds are highest during this stage.
- Quiet sleep contains 12–18 Hz oscillatory events lasting 1 s that resemble mammalian sleep spindles—strikingly convergent given 550 million years of separate evolution and no homologous brain structures.
- During active sleep, octopuses rapidly cycle their skin through the exact camouflage patterns they use while awake, hinting at offline 'replay' or rehearsal of skin-pattern control.
- Sleep-deprived octopuses rebound specifically by increasing active-sleep bouts, proving the REM-like stage is homeostatically defended and essential, not incidental.
- Active-sleep interval timing is temperature-dependent: a 1°C rise shortens the cycle by roughly 5 minutes.
- A preprint reports bizarre 'nightmare-like' abnormal episodes in senescing octopuses, though this is anecdotal.

## [RNA Editing and the Molecular Basis of Neural Complexity in Cephalopods](https://octopuscognition.org/sections/rna-editing-and-the-molecular-basis-of-neural-complexity-in-cephalopods/index.md)

- The majority (60%) of squid brain transcripts are recoded—versus under 1% of recoding-capable transcripts in humans—inverting the usual assumption that the genome is the master blueprint.
- Extensive editing appears to have slowed genome evolution: coleoids may have partly traded away DNA evolvability to keep their editing machinery, a rare case of a molecular mechanism constraining the genome that encodes it.
- An octopus can rewire its neural proteome to the cold within hours, reaching steady state in about four days, using RNA edits rather than new gene expression alone.
- Antarctic and tropical octopus K+ channels have nearly identical genes but diverge in function almost entirely through differential RNA editing.
- Editing that changes protein sequence is largely confined to the nervous system, tying the mechanism specifically to neural function.

## [Nociception, Pain, and Sentience in Octopuses](https://octopuscognition.org/sections/nociception-pain-and-sentience-in-octopuses/index.md)

- Crook's 2021 study is claimed to be the first demonstration of probable ongoing/tonic (spontaneous) pain in ANY non-mammalian animal, not merely reflexive nociception.
- Analgesia (lidocaine) was rewarding only to octopuses that had experienced pain—controls showed no preference—separating the affective 'suffering' dimension from mere sensation.
- Nociceptive sensitization is adaptive, not just a byproduct of damage: sensitized injured squid escaped predators better, and anesthetizing the wound removed the survival advantage (Crook et al. 2014).
- Octopus nociceptors show mammalian-style long-term sensitization and spontaneous firing, despite 500 million years of divergent evolution—a case of convergent pain machinery.
- The LSE report's recommendation against octopus farming rippled into real bans in Washington State and California and a proposed U.S. federal OCTOPUS Act (2024).
- Nautilus sentience is essentially unknown (1/8 criteria met with confidence), showing 'cephalopod' sentience is far from uniform across the class.

## [Comparative Cognition and the Convergent Evolution of Minds](https://octopuscognition.org/sections/comparative-cognition-and-the-convergent-evolution-of-minds/index.md)

- Octopuses and humans last shared an ancestor 550-600 Myr ago that had essentially no complex brain—so large brains evolved from near-scratch at least twice, making cephalopods a true independent origin of mind.
- About two-thirds of an octopus's 500 million neurons are in its arms, not its central brain; severed arms continue coordinated behavior—cognition is genuinely decentralized.
- Cuttlefish episodic-like memory does NOT decline with age (Schnell et al. 2021), the opposite of the memory decline seen in aging humans, mammals, and corvids.
- Coleoid cephalopods recode over 60% of their neural RNA transcripts via A-to-I editing (vs under 1% in humans), and appear to have traded genomic evolvability for this transcriptome plasticity.
- The 2013 cuttlefish episodic-memory paper was co-authored by Nicola Clayton—the very scientist who first demonstrated the same capacity in scrub jays—making the convergence almost poetically direct.
- The octopus genome has 168 protocadherin genes (roughly 10x other invertebrates, >2x mammals), a gene family previously thought to be a vertebrate specialty for wiring brains.
- Cephalopods break the standard theories of why intelligence evolves: they are mostly short-lived (1-2 years) and asocial, contradicting both the 'social intelligence' and 'long lifespan' hypotheses.

## [Genome, Development & Evolution of the Cephalopod Body and Brain](https://octopuscognition.org/sections/genome-development-evolution-of-the-cephalopod-body-and-brain/index.md)

- Octopuses did NOT get complex via whole-genome duplication (unlike vertebrates)—the popular hypothesis was falsified by Albertin et al. 2015.
- The Hox gene cluster is completely 'atomized'—scattered across the genome rather than clustered as in virtually every other bilaterian animal.
- Octopus has 168 protocadherin genes, 10x an oyster/limpet and roughly double a human—and squid evolved their expansion independently.
- Coleoids recode >60% of brain transcripts through RNA editing, versus <1% in humans, effectively editing proteins on the fly instead of in DNA.
- Heavy RNA editing imposes an evolutionary trade-off: it slows DNA-level evolution because editing sites require conserved surrounding sequence.
- Removing the optic gland reverses the maternal 'death spiral'—mothers abandon their eggs, resume eating, and live months longer (Wodinsky 1977).
- Octopus and human brains independently recruited active LINE retrotransposons in memory regions—molecular convergent evolution.
- Octopus intelligence is essentially non-cultural: no parental care, dispersing embryos, and death after one reproduction mean each animal learns from scratch.

## [Research Methods, Welfare in the Lab & Future Directions](https://octopuscognition.org/sections/research-methods-welfare-in-the-lab-future-directions/index.md)

- Cephalopods are the ONLY invertebrates regulated for research welfare in the EU (Directive 2010/63/EU, since 2013)—an entire animal class regulated for the first time.
- Standard behavioral tech fails on octopuses: Skinner-box lever-pressing was never reliably learned and trap-tube tasks are meaningless because flexible arms just reach around obstacles.
- Octopuses are semelparous and die within 1–2 years after a single brood, so lab animals are mostly wild-caught and cannot be aged into longitudinal studies.
- The first brain waves from freely moving octopuses were recorded with data loggers originally built to track bird brains during flight—waterproofed and implanted in the mantle cavity.
- Washington State passed the world's first legislative ban on octopus farming (2024) before any commercial farm even opened.
- Some historic 'reversal learning' results may partly reflect electric-shock reinforcement and experimenter cueing rather than pure cognition.
- The genetic future of cephalopod neuroscience may hinge not on octopuses but on a tiny transparent gene-edited bobtail squid (Euprymna berryi).

## [Vision, Eye Design, and the Perceptual World (Umwelt) of the Octopus](https://octopuscognition.org/sections/vision-eye-design-and-the-perceptual-world-umwelt-of-the-octopus/index.md)

- The octopus eye has NO blind spot — its everted retina puts photoreceptors facing the light with axons exiting the back, the opposite of the 'backwards' vertebrate retina, despite looking almost identical externally.
- Octopuses are genetically colorblind (a single 475 nm pigment) yet produce perfectly color-matched camouflage; the leading explanation is that they 'taste' color through chromatic aberration and a weird off-axis pupil rather than through color receptors.
- Polarization discrimination reaches 1.3 degrees of e-vector angle — a sensory channel humans lack entirely, effectively giving octopuses a second 'color' dimension invisible to us.
- An octopus cannot tell a horizontal bar from a vertical bar if you remove its statocysts — it never learned to mentally rotate shapes; instead it relies on gravity to physically keep its retina level.
- The octopus focuses by moving its whole lens toward the retina like a camera, not by squeezing the lens like a mammal.

## [Chromatophore Motor System, Body Patterning, and Communication as Externalized Cognition](https://octopuscognition.org/sections/chromatophore-motor-system-body-patterning-and-communication-as-externalized-cognition/index.md)

- Chromatophores are muscles, not cells — cephalopods are the only animals that drive body color by direct neural innervation of pigment organs, with no hormonal step, so the skin is effectively a live display screen wired to the brain.
- The whole color-control system apparently runs open-loop, with no visual or proprioceptive feedback, yet produces near-perfect background matching.
- Despite seemingly infinite skin output, the camouflage repertoire collapses to just three template patterns (Uniform, Mottle, Disruptive).
- The animals are essentially colorblind (usually a single retinal opsin) but produce color-matched camouflage — possibly by exploiting chromatic aberration through weird pupils.
- The skin itself contains opsins and can expand chromatophores in response to light with no input from the eyes or brain (LACE / distributed dermal photoreception).
- 'Passing cloud' displays may be driven by central pacemaker circuits analogous to locomotor pattern generators — a visible readout of an internal neural oscillation.

## [Numerical, Quantity, and Abstract-Concept Cognition in Cephalopods (with Cross-Modal and Mirror/Self Tests)](https://octopuscognition.org/sections/numerical-quantity-and-abstract-concept-cognition-in-cephalopods-with-cross-modal-and-mirror-self-tests/index.md)

- Cuttlefish reverse their numerical preference based on hunger: hungry animals pick one large prey, satiated ones pick two small—number choice is value-driven, not fixed.
- Longer delay-of-gratification tolerance predicts faster learning in cuttlefish, mirroring the human 'marshmallow test' correlation in a mollusc with no shared ancestry for such control.
- Some cuttlefish physically turn away from tempting food, resembling self-distraction strategies documented in children and apes.
- Octopus object recognition is tactile-DOMINANT: they trust touch over vision when forming representations of novel objects, the reverse of the human default.
- The 'self-directed' mark-touching octopuses do in mirror tests happens just as often with no mirror and in sham-marked animals—it's proprioception, not self-recognition.
- Numerical discrimination follows Weber's law (ratio-dependent), meaning cuttlefish use an analog approximate-number system rather than exact counting or subitizing.