Octopus Intelligence

    Posted on Monday, Apr 01, 2019
    At last! The long-awaited (by uh, one person) octopus intelligence post!

    Absolute ages ago, I wrote a bit on the frankly impressive intelligence of corvids, usually known as the group of birds containing crows, ravens, and magpies (I believe). Here's the secret. I went into biology because I wanted to know what it was like to be a cat or a dog or a dolphin or something. I'm like 80% sure that's why anyone goes into biology. You learn something like oh, birds can use tools but monkeys can use better tools and you think to yourself, dang, what's it like to be something that hasn't got any arms, and uses tools?

    Or something that doesn't?

    Octopuses are a nonsocial intelligent invertebrate - dolphins, monkeys, cats, dogs, even ravens have complex social structures that seem to loosely correlate with their own learning capability and the complexity of their behaviors. (well duh, social interactions necessitate a whole series of behaviors that you wouldn't need otherwise, which boil down to different kinds of communication - but sure, let's take this as a semi-accurate attempt at measuring intelligence for now).

    But octopuses don't have many particularly social behaviors. Also their brains are itty bitty. The pressures that drove their intellect development don't have anything to do with inter-octopodal communications. They have no need for leadership, no need for interaction outside of the brief mate-then-die.

    Another weird thing. Octopuses don't live very long. They can live for an unmeasured amount of time so long as they don't mate, but the octopus starts to experience aging, which they aren't really equipped to do. They begin to degrade and get less healthy. Once they mate, that's it. They die soon afterwards, just a couple years after being born. This is also at odds with most intelligent species - why develop all that intelligence, just to die after a little while? The benefit that the intelligence gives the octopus must be a very short-term sort of benefit, for it to make sense evolutionarily.

    So basically, here's the point of this post. Octopuses are weird. They're probably the most alien intelligence (both of those things) that we have access to.  What on earth, what in the deep blue sea is it like being an octopus? How smart are they? And how is that intelligence different from ours?
    This is probably going to be a long post with a lot of anatomy in it. If you're not interested in that stuff, just interested in knowing what it feels like to be an octopus, just skip to the second set of double bars in the article.

    See you on the flip side, meatfriends.

    With that in mind let’s get started with some facts relevant to the experience of octopusness. 


    Octopuses have olfactory chemoreceptors all over their bodies, mostly in the lips, mouth, suckers, and olfactory organ.  The organ has muscles which control its posture, allowing it to shift in order to orient the spatial gradient of cues.  Odorant molecules which the octopus can sense are water soluble (just as the odorant molecules that you can sense must be air-soluble).  They can detect salts, sugars, amino acids, amines, peptides, proteins, and some hydrocarbons. 


    The octopus’s skin is an incredible sensory organ.  It contains the standard gamut of pressure, temperature, tactile, and chemical receptive structures that your skin has.  (Remember that the octopus, because of its many arms, has a much higher surface area to volume ratio than you do). 

    Their skin is rich in opsin, the same light/color detecting protein in your eyes.  This provides visual information about their environments, which the octopus can use to camouflage itself.  The color changing properties of octopus skin are also very important for choosing a mate.


    Unusual for invertebrates, octopuses have very developed camera-like eyes.  In some ways they’re actually better than ours.  The human eye has a blind spot where the optic nerve passes in front of some photoreceptive cells of the retina.  This blind spot is removed in post-processing so you don’t notice it. 

    The octopus does not have any retinal blind spots.  The layers of its retina are organized in the opposite order from human retina layers.  They have only one type of photoreceptor, which should mean that they cannot see colors, but their pupils are off-axis on their eyes, which allows spectral information to fall onto their retinas and be interpreted by their brains at the expense of image sharpness.  Their perception of color and sight is probably very different from our own.   

    Octopuses have many other neat senses as well.  They have a vestibular system and a primitive auditory system.  The vast sensory information available to the octopus means that it requires some complex structure to deal with all of this information. 


    You can locomote because your muscles pull against your skeleton.  Your joints are a fulcrum, and your bones a lever.  The degrees of freedom you have are limited.  You can only move at your joints, and only within a small range of angles. 

    Octopuses do not have skeletons.  Their arms have infinite degrees of freedom – that is to say, there are an infinite number of ways to accomplish any motor task.  They can move their limbs in any direction at any point along their length.  They can lengthen or contract.  A large slice of the human brain along the middle is devoted to sensing and controlling the body.  Imagine the computational complexity required to control eight flexible limbs with infinite degrees of freedom. 

    In addition, they have tactile and proprioceptive sensory neurons that can detect the orientation of the arms in any configuration.  The limited degrees of freedom in a terrestrial skeleton allow for the brain to keep a “map” of the body’s position at all times.  All you have to track is the position of each joint, the stretch on each muscle.  This sounds complex and it is.  How the heck does the octopus manage it?  Maybe it doesn’t – but if it doesn’t, how does it control its tentacles so accurately?  Particularly the tentacles outside its field of view. 

    The octopus neuromuscular system has many adaptations that seem to simplify the effort of controlling the muscle fibers.  This has to do with the way the muscles are built and with the way the nerves interact with them. 

    The muscular control fibers in the motor cortex of a human are topographically arranged.  On the lateral ends, there are the foot control fibers, which are adjacent to the shins, thighs, torso, and so on.  The sensory centers for each of these locations are located adjacent to their corresponding motor control area.  This is conserved throughout many terrestrial animals. 

    The octopus does not have any such topographical map in its brain.  It does not receive any proprioceptive input from its arms, so probably it has no idea how its limbs are arranged at any point in time.  Stimulation of an octopus brain yields complex movements of multiple limbs.  There doesn’t seem to be a way for the octopus brain to activate a single arm at a time.  The complexity of the motion varies with the intensity of the stimulus.  In the octopus brain, the motor cortex appears to be organized according to complex patterns or behaviors rather than body parts.

    If you cut off an octopus limb and stimulate neurons, you will see very natural-looking limb motions.  If you try the same with a human arm, you will see jerky, forced, unnatural motions.  It appears that the motor cortex decides what sort of behavior the octopus wants to engage in, and the limbs decide themselves how this action will be completed.   

    The sensory system is also not organized in terms of body parts.  The highest order sensory neurons will fire in response to stimulus along an entire limb, or multiple limbs. 


    This is the scariest thing I’ve ever learned about the octopus. 

    In all living creatures, DNA codes in the nucleus are copied into RNA transcripts, some of which are carried out of the nucleus and into the ER, where the information is used to make proteins.  Any change in the DNA or RNA sequence changes the eventual protein structure.  RNA transcripts are usually modified in a number of predictable ways on their journey out toward the ER.  In most organisms, about 1% of transcripts are edited, and usually the transcripts that are edited do not code for proteins. 

    Octopuses edit more than 60% of RNA transcripts.  Recoding sites are highly conserved.  Octopuses have experienced an evolutionary tradeoff – slow DNA evolution in exchange for the ability to experiment with RNA transcript editing. 

    These edits occur mostly in the nervous system and largely affect proteins called protocadherins, which are involved in neural plasticity.  Editing can be transient, selective (only expressed in specific tissues), conditional (only expressed under certain temperature ranges/pressures/emotional states/etc), partial (only 10% of transcripts edited for example), or other possible permutations. 

    The octopus therefore has a highly plastic brain.  The octopus can modulate the plasticity of its brain to fit the circumstances, and they experience adult neurogenesis – they can add newborn neural cells to existing circuits.  If the octopus brain is damaged, it can regenerate in a way that the human brain cannot. 


    Octopuses have a fantastically large number of neurons for an invertebrate – slightly fewer than your average dog.  They have comparable learning and memory, but lower behavioral flexibility than dogs. 

    Their nervous systems are divided into three parts – the optic lobes, the arm motor system, and the brain.  The brain has very clear and strict divisions between its lobes – the mammalian brain is much more integrated.  Like many invertebrates, the octopus appears to have evolved a brain out of the fusion of many separated ganglia with discrete functions.   

    Mammalian neural cells have input tendrils (dendrites) and output tendrils (axons).  Octopus neurons have only one protruding tendril called a neurite. 

    The connection between the optic lobe and the brain is a relatively small number of fibers.  The optic lobes only pass highly processed information to the brain. 

    This neural plan also occurs in the motor system.  Higher order commands are sent to the arms, but the specifics of how to carry out these commands is calculated by the arms themselves. 


    An ocean-dwelling octopus in a laboratory tank hides for about a week.  Shortly after that, they begin exploring.  They react with curiosity to their environment, and eventually, if able, attempt to escape.  There are stories of aquarium octopuses making hilarious and inspired escape attempts.  I recall a news story about an aquarium that kept mysteriously losing fish – it turned out the octopus had learned to escape its tank at night, enter the tanks containing other fish, eat those fish, and then return before the aquarium opened. 

    Octopuses experience associative learning and observational learning.  An octopus who watches another octopus perform a task will be able to do that task much faster than an octopus who has not. 

    They also require constant mental stimulation.  Many octopuses have clear preferences in food or toys.  They have distinct personalities and temperaments.  They have a preferred eye and arm just like humans do. 

    Here’s a short TL:DR in case you didn’t want to go through all that. 
    • The octopus has a large amount of incoming sensory information. 
    • The octopus has a computationally intensive task controlling and coordinating its many arms. 
    • Incoming visual information is processed to a high degree by the optic lobes.  The connection between the optic lobes and the brain is quite small. 
    • The connection between the brain and the motor system is also quite small.  The brain sends very high level commands to the motor system, and the arms themselves compute how they are to complete this task.  The arms track their own position and orientation, but do not share this information with the brain. 
    • The octopus brain is very small and relatively undeveloped.  There is little integration between lobes. 
    • The octopus brain is ridiculously plastic and flexible. 
    We’re getting something of a picture here.  Octopuses experience the world through a relatively small number of fibers.  They don’t need to track all of their own arms, because their arms do whatever they need to do with the effectiveness and efficiency.  I’m working up a theory that octopuses don’t think about what their arms are doing – they think about how they want the thing they’re playing with to move.  The arms do all the calculations as to how to manage that motion themselves. 

    Their high neural plasticity means that they adapt to whatever they need to.  They learn very quickly.  But the low integration between their lobes means that they would have a harder time connecting separate concepts.  This is probably why they seem smarter than dogs in some ways, but much less behaviorally flexible. 

    Most of their intelligence exists outside of their brains and in the peripheral nervous system.  Being an octopus is probably much like wanting something and then experiencing your body bringing you that thing.  Or getting a sense for something happening, deciding what to do about it, and then your body sort of does what’s needed.  Their fine volitional control is probably worse than ours.  An octopus would likely have a very hard time learning to type or play the piano. 

    If we meet extraterrestrial intelligences that are intelligent enough, maybe we can ask them what it’s like inside their heads.  It would be very hard for them to explain and for us to understand, because this would involve experiences we have no context for.  It’s like trying to explain red to a blind man.  Octopuses are some of the most alien intelligences we have access to on earth.  Their brains are built very differently from ours, and their environment is very alien to us.  The processes that created their intelligence are also very different.  Learning what it’s like to be an octopus is very good practice for understanding aliens. 

    Also it’s fun. 

    Stay meaty, friends. 

    See it here first!  https://appliedmeatsciences.wordpress.com/2019/03/31/octopus-intelligence/

    --- Elizabeth Broadwell
    Elizabeth Broadwell is a Masters student in the College of Arts and Sciences

Search for a blog post

Defense Announcements

  • There are no announcements at this time.

News and Announcements

Upcoming Events