Jennifer A. Mather
Department of Psychology & Neuroscience, The University of Lethbridge,
Lethbridge, Alberta T1K 3M4, Canada
Roland C. Anderson
The Seattle Aquarium, 1483 Alaskan Way, Seattle, WA 98101 USA


NOTE: This article is written in the same form as an article for a scientic journal: all the literature used in writing it is carefully cited in the text, so if you want to learn more about certain aspects of  Octopus intelligence, you can go to a university library and look up the origonal article - ask the librarian to help you learn how to do this - you'll need these skills a LOT, when you get to college!!

          The same question about octopus behavior intrigued both authors, though at different places and from different backgrounds. While watching an Octopus vulgaris in Bermuda, the first author observed it sitting in its sheltering den after a foraging expedition, where it caught several crabs, took them home and ate them. Suddenly it jetted out directly to a small rock about two meters away, tucked it under its spread arms and jetted back. Going out three times more in different directions, it took up three more rocks and piled the resulting barrier in front of the entrance to its den. It held them in front with several arms and went to sleep. This didn't look like random action, but planning. The second author came in one morning to the Aquarium to find one of the giant Pacific octopuses had been busy overnight. The gravel on the tank bottom was dug up, the nylon cable ties that attached the undergravel filter to the tank had been bitten through and the detached filter had been bitten or torn into small pieces, which now floated on the water surface (experienced octopus keepers know that Murphy's Laws seem to apply especially to octopuses!). Again, this looked like a careful sequencing/planning of actions and learning put to use, though the reasons weren't at all obvious. These observations made both of us believe that octopuses could possibly be intelligent and use their intelligence for unexpected purposes.

          When humans think of intelligence, we think of ourselves. This anthropocentric viewpoint is partly because intelligence has only really been studied in vertebrates and partly because we see its evolution as leading to the pinnacle called Homo sapiens. Until recently, there hasn't been any model of how another completely separate group could show us how intelligence might evolve differently than ours. Research on the octopuses is beginning to provide that alternate model.

         An octopus is very different from a mammal. It only lives about two years. It has much less opportunity to gain and use intelligence than an elephant, which has a 50 year lifespan and three generations of a family to lead and learn from. Still, bees learn about flower locations from other bees, and they live only a few weeks as adults. However, an octopus is also not social; Humphrey (1976) suggested that intelligence has evolved to solve social dilemmas. The young octopus learns on its own with minimal contact with conspecifics and no influences of parental care or sibling rivalry. However, the octopus has a large brain with vertical and sub-frontal lobes dedicated just to storing learned information (Wells, 1978):  it has the anatomy for  a robust, built-in intelligence.

       But, it is not enough to know that the anatomy predicts an animal to be intelligent without some idea of how it uses this ability. Investigations at Naples in the 1950s and 1960s found that octopuses (or "octopi", if you want to Get Latin!!) can learn a wide array of visual patterns, encoding information mostly by comparing edges, orientations and shapes. They also learned by touch, and tactile information seemed to be stored in a different brain area than visual. Intent on  just demonstrating learning abilities at first, researchers did not follow up to find what octopuses were doing with this learning in their ocean home. As ethology's (ie, the ethical or "moral" side of science, which discourages direct experimentation on intelligent animals) emphasis on observation of natural behavior in the field began to fill the gap (see Lehner, 1998), the Naples studies ended, and no linkage was made between abstract information storage and the use of learning in daily life. Finally, this gap is being bridged by such works as Hanlon and Messenger (1996), who provide an overview of cephalopod behavior. But, even asking the right questions about octopus intelligence is difficult, since we understand so little of their minds. Watching an animal and wondering how it is organizing its world, then testing it to see if your guesses have some foundation - that is very difficult indeed! Still, we are starting to get some answers both by observing in the field and by studying areas such as prey manipulation, personality and play (yes, play!) in the octopus.

       One of the insights into how we might view octopus intelligence came for the first author when reading Neisser's (1976) definition of cognition (ie, thinking) as "all the processes by which sensory input is transformed, reduced, elaborated, stored, recovered and used."  It seemed a focal issue: what were octopuses in the ocean doing with the information that learning studies said they could acquire? One study we undertook centered on what we came to call the "Packaging Problem".  The problem posed was how an octopus could utilize a delectable clam enclosed in its hard shell,  - to get at the soft, delectable clam body.  This is the end result of what Vermeij (1993) called an "evolutionary arms race":  many predators evolved means of penetrating the hard shell the clam uses to protect itself, which is held together by powerful muscles - sea stars pull the valves apart, oyster-catcher birds pry them apart, moon snails (Naticidae) drill a hole into the shell, and gulls drop the clam from a carefully calculated height onto rocks or road pavement. But the octopus goes these predators one better: it can use several different strategies to solve this Packaging Problem, instead of just one or two!

      Octopuses come well-equipped with an arsenal of different solutions for use in feeding. They have the holding ability of hundreds of suckers and the pulling power of the eight muscular arms, flexible because they are boneless (see Mather, 1998 for arm movement capacity). Underneath, inside the mouth at the junction of the arms, they have a parrot like twin beak for biting. Also inside the mouth are two more useful structures, the radula with teeth for rasping and the extendible salivary papilla. It delivers cephalotoxin, a neuromuscular function blocker that can kill a crab in several minutes (Boyle, 1990). Fortunately for us, only the venom of Haplochlaena spp. octopuses (the famous "blue" octoupuses!) has proven fatal to humans.

      Since octopuses are well set up to "recover and use" information for solving the problem of the clam's protection, we set out to determine what the giant (up to 50 kg) Pacific octopus, Enteroctopus dofleini, would do to get at three types of bivalves. When we offered them separately or together at the Seattle Aquarium, octopuses ate many Venerupis (a Venus Clam) clams, some Mytilus (mussels) and few Protothaca clams. The prey species were each opened differently, however. The fragile mussel shells were simply broken and the stronger Venerupis were pulled apart. The thick shelled Protothaca were drilled with the octopuses' radula and salivary papilla, or chipped with the beak,  then injected with poison which weakened the adductor muscle holding the valves together. The octopuses'  strategy to penetrate into the different clams varied.
 When offered the clams opened ìon the half shell,  the octopuses changed preference and consumed both clam species, but hardly any mussels.  When they didn't have to work hard for the clam meat, they liked Protothaca. Some clue that effort might be the reason for this shift came when we measured the resistance to opening force of the adductor muscles of the bivalves : Mytilus resisted until an average of 2.2 kg,  Venerupis,  3.6 kg, and Protothaca to 4.6 kg. Octopuses could also shift their penetration strategies. When live Venerupis clams were wired shut with stainless steel wire, the octopuses couldn't pull the valves apart, so they then tried drilling and chipping as penetration techniques (given empty weighted shells glued shut, the octopuses ignored them; they were on to that trick right away!). This flexibility of strategies echoes what Wodinsky (1969) found with Octopus vulgaris drilling Strombus gastropods. These octopuses drilled through the shell apex to poison and weaken the snails' adductor muscles. When he coated this part of the shell with latex, they just pulled it off,  then drilled as before.  When he then put on aluminum, they simply drilled through the metal and shell, but when he coated it with impenetrable dental plastic they drilled elsewhere on the shell,  or pulled the snail out by sheer force. For both species of octopus, the motto might be "do whatever works to get your meal!"   They were intelligently adapting the penetration technique to the clam species presented and the situation in which they were placed.

       The first author  (Jennifer Mather) also noticed this pragmatism (ie, a "whatever it takes to get the job done" attitude!) and identified tool use by octopuses during field studies in Bermuda (Mather, 1994). Tool use does not automatically denote learning but the range of uses of one tool, water, also suggests the octopus is intelligent:  circulation of water in molluscan mantle cavities is primarily used for respiration and removal of wastes, and secondarily for locomotion in scallops and squid (Morton, 1967). Octopuses also use water jets through their flexible funnel for tertiary (ie, additional) functions such as cleaning  out their dens. They gather an armful of rocks and sand under their web, go to the den entrance and tilt the web upward, then blow the whole lot out and away with a water blast from their funnel!  Similarly, an octopus holds a crab under the web, dismembers it, eats the flesh and holds the cleaned out exoskeletal pieces.  At meal's end, it tilts up the web and blows the pieces outside, adding to a midden outside the den. Scavenging fish attend octopuses when they go hunting, and when they discard remains onto this midden. One of the techniques octopuses use to repel these "pests" is to direct strong blasts of water jet at them - like a water gun!! (Mather, 1992). On occasion, an octopus jets water to repel human observers, and octopuses in the lab have jetted into the faces of researchers or onto their delicate electrical equipment.

      In the laboratory, octopuses adapt and use this water jet in a behavior that has generally been considered exclusively of vertebrates: they play (Mather and Anderson, 1999).  We set out to prove that octopuses (Enteroctopus dofleini in particular) play, deciding that being in a non-stimulating situation except for having an item that they could manipulate, might cause such activity. A floating pill bottle, which sometimes drifted in the current from the water intake, was the item. We didn't expect social play from a solitary animal, rather that the exploration that the octopus mentioned at the start of this paper demonstrated so well by tearing apart its tank would turn the focus of its behavior, as Hutt (1966) suggested, from "what does this object do?" to "what can I do with the object?  Every octopus jetted at the floating toy at least once in the ten trials, but only two of them reached the criteria for play. These were 1) regular repetitions of 2) simple acts for 3) over 5 minutes, of pill bottle repulsion toward the water inlet jet and return by it. One octopus set up a 2 minute circuit of the bottle around the tank and a second jetted the toy straight towards the water intake, getting a return in 30 seconds. This prompted a long distance call from the more skeptical second author to the first, in the middle of one of those busy academic days, with the simple message "She's bouncing the ball!"

      Play is a difficult and sometimes controversial area, as it does not delineate a separate category of behaviors. Forms that are seen as play merge into other categories of "useful" actions (Fagen, 1981). This example appears to be a small glimpse of that continuum, change in the use of mantle water circulation from its basic molluscan function to newer situations. Play involves the detachment of actions from their primary context, and such flexibility is both a basis and a sign of intelligence, whether it be shown in a person or a fish or an octopus. It is the formation of a new combination of information input and actions.

     A third aspect of the lives of the octopuses which shows their capacity for acquiring different responses is their possession of "personalities".  The impetus for this study came from the second author's  work at the Seattle Aquarium (Anderson, 1987). Volunteers are the backbone of public institutions such as the Aquarium, and volunteers see animals a little differently than scientists. They give individual names to three species of animals in the Aquarium - the seals, the sea otters, and the octopuses. There was "Leisure Suit Larry", named for a video game character who would be cited daily for sexual harassment on the job for excessive touching. There was "Emily Dickinson", who hid behind the tank's backdrop and could barely be coaxed out.  And there was "Lucretia McEvil",  whose destructive acts are featured at the beginning of this article. Volunteers shied away from feeding her because she would try to pull them down into her tank.

      We decided to take this impression of differences between individuals and systematize it:  what would it mean to say that octopuses had personalities, and into what categories might we fit them? So we started an octopus vs  octopus study of the small Pacific red octopus Octopus rubescens. Instead of testing in a novel situation and calculating average responses, we tested three everyday situations to find variation. The situations were alerting, threat and feeding, and over three years 44 octopuses were tallied for nineteen responses. To find variation rather than averages, we did some difficult and "advanced" statistics: a Factor Analysis and then a Principal Components Analysis.  What the first does is to group behaviors into clusters of occurrence amongst individuals, called Factors, and our analysis told us there were three Factors, described below.  The Principal Components Analysis changed these factors slightly so they were not correlated with each other and could then be called Dimensions of Personality. Each octopus (and any future one) could then be placed somewhere on each of these dimensions, and could be given an Octopus Personality Profile (Mather and Anderson, 1993).

     Once the researcher has these dimensions, they can be assigned names. In the octopuses' case we chose three: Activity, Reactivity and Avoidance. So an Active octopus reacted to the threatening probe by grabbing it, a Reactive one performed a set of behaviors that put distance between itself and the threat and an Avoidant one tried to stay away from the situation in the first place. This catalog of variation is interesting by itself, but the dimensions occur in other animals as well. Fish, monkeys and people differ on some variable often called Shyness, on another called Emotionality and a third defined as Exploration or Activity. While the dimensions were of course extracted from the responses by a human brain, they are similar in phylogenetically (ie, gentically)  distant animals (see Gosling and John, 1999).

        Why does this matter to the demonstration of intelligence?  For one thing, personality overlays intelligence.  Autistic children's intelligence is often hard to measure because they don't like people well enough to cooperate with the testers.  Patterson and Linden (1981) found the gorilla Koko showed the same withdrawal in the middle of an intelligence test; he got bored and started pressing the same button over and over. One octopus in a group being tested for spatial memory "freaked out" at being put in an open tank and circled the tank for ten minutes at a time (personal observation). She never had a chance to learn the task. Was she stupid? Povinelli, et al., (1993) tested chimpanzees for self recognition and made sure to test many individuals to cover this variation. They concluded that the differences were so high that individuals' intellectual level would have been assessed as typical of quite different species, and not just the one!

        In addition,  "personality" allows individuals to show intelligence. If the sensory input is to be "transformed, reduced, elaborated, stored recovered and used" (Neisser, 1976), it has to be on the basis of individual variation. The intelligent animal can master variable environments by using all these processes, and that leads us back to the topic:  what is intelligence like?  Indeed, it may be the variable environment that selects for intelligence, in a Darwinian "survival of the fittest" sense: since many octopus species spend their early months in as plankton, drifting to all sorts of different habitat-types: the octopus that settles out of the plankton onto a rocky shoreline has to learn to find different prey and avoid different predators than the one that finds its home under the only rock on a sandy bottom. Without this ability to become different, they won't survive. Coping with a variable environment is what will demonstrate the asocial octopus's particular "take" on intelligence. Thus, the studies of Fiorito et al. (1999) on the octopus's ability to open a glass jar and Hanlon et al.ís (1999) assessment of the avoidance strategies of O. cyanea to a threatening human also open a window on the octopusís use of intelligence.

      Perhaps it is this individual sensitivity to change, honed by intelligence and variability, that has been the key to the success of both the cephalopods and the higher vertebrates. Similarities that could lead us to understand the evolution of intelligence in octopuses and humans are few, but thought-provoking: 1) neither group has the protection of exoskeleton, scales or armor,  2) both have evolved in complex environments, the octopod in the tropical coral reef and the hominid in the savanna edge,  and  3) both have considerable variability among individuals and the ability of being able to change their behaviour to help them survive. So, perhaps looking at the octopuses through their intelligence, feeding flexibility, predator avoidance, play, and personality helps us also look at aspects of ourselves, from another angle!


Literature Cited:
Anderson, R.C. 1987. Cephalopods at the Seattle Aquarium. International Zoo Yearbook. 26:41-48.

Boyle, P.R. 1990. Prey handling and salivary secretions in octopi. In: M. Barnes and R.N Gibson (eds). Trophic relationships in the marine environment. Proc. 24th Eur. Mar. Biol. Symp. Aberdeen University Press (Aberdeen, Scotland). pp. 541-552.

Fagen, R. 1981. Animal play behavior. Oxford University Press (Oxford, UK). 684 pp.

Fiorito, G., C. von Planta and P. Scotto. 1990. Problem solving ability of Octopus vulgaris Lamarck (Mollusca: Cephalopoda). Behavioral and Neural Biology. 53:217-230.

Gosling, S.D. and O.P. John. 1999. Personality dimensions in nonhuman animals, a cross-species review. Current Directions in Psychological Science. 8:69-75

Hanlon, R.T. and J.B. Messenger. 1996. Cephalopod Behaviour. Cambridge University Press. 232 pp.

Hanlon, R.T., J.W. Forsythe and D.E. Joneschild. 1999. Crypsis, conspicuousness, mimicry and polyphenism as antipredator defences of foraging octopuses on Indo-Pacific coral reefs, with a method of quantifying crypsis from video tapes. Biological Journal of the Linnean Society. 66:1-22.

Hutt, C. 1966. Exploration and play in children. Symposium of the Zoological Society of London. 18:61-81.

Humphrey, N.K. 1976. The social function of intellect. pp. In: P.P.G. Bateson and R.A. Hinde, eds. Growing points in ethology. Cambridge University Press. pp. 303-317.

Lehner, P.N. 1998. A handbook of ethological methods. Cambridge University Press. 692 pp.

Mather, J.A. 1992b. Interactions of juvenile Octopus vulgaris with scavenging and territorial fishes. Mar. Behav. Physiol. 19:175-182.

Mather, J.A. 1994. ìHomeî choice and modification by juvenile Octopus vulgaris (Mollusca: Cephalopoda): specialized intelligence and tool use? J. Zool. Lond. 233:359-368.

Mather, J.A. 1998. How do octopuses use their arms? Journal of Comparative Psychology. 112(3):306-318.

Mather, J.A. and R.C. Anderson. 1993. Personalities of octopus. Journal of Comparative Psychology. 107(3):336-340.

Mather, J.A. and R.C. Anderson. 1999. Exploration, play and habituation in octopuses (Octopus dofleini). Journal of Comparative Psychology. 113(3):333-338.

Morton, J.E. 1967. Molluscs. Hutchinson and Co. (London). 244 pp.

Neisser, U. 1967. Cognitive psychology. Appleton-Century-Crofts (NY). 351 pp.

Patterson, F. and E. Linden. 1981. The education of Koko. Holt, Rinehart and Winston (NY). 224 pp.

Povinelli, D.J., A.B. Ruff, K.R. Landau and D.T. Bierschwale. 1993. Self-recognition in chimpanzees (Pan troglodytes): distribution, ontogeny and pattern of emergence. J. Comp. Psych. 107:347-372.

Vermeij, G.J. 1993. A natural history of shells. Princeton University Press. 207 pp.

Wells, M.J. 1978. Octopus ñ physiology and behaviour of an advanced invertebrate. Chapman and Hall (London). 417 pp.

Wodinsky, J. 1969. Penetration of the shell and feeding on gastropods by octopus. American Zoologist. 9:997-1010.



This is a new counter system set up by Globel on
December 01, 2002