Script on Learning in Octopus by J. Z. Young, M.A., F. R. S.
The Octopus lives in crannies among rocks, occupying one “home” for a long time. It emerges from time to time to collect its prey of crabs of bivalves and carries several home to eat at leisure. Piles of shells thus collect outside the home.
The Octopus sits in the home with only the head and eyes protruding. If a crab comes into view the Octopus leans forwards and comes walking out on its tentacles. If the crab is some way away the Octopus swims swiftly to it by jet propulsion action of its funnel. When it arrives it open its arms and spreads them over the crab which is seized by the suckers and held inside the web and bitten by the parrot-like beak. The bite is poisonous and the crab ceases to struggle within a few seconds and is then either eaten at once or carried around while the Octopus hunts for more.
If the crab is large the Octopus will approach it more cautiously, putting out one arm to touch it and perhaps working several arms round to seize the crab from behind, avoiding its claws.
The experiment consisted in training the Octopus not to attack crab if a white square was also present. This was done by dangling the crab on a thread in front of the square to which a pair of electric wires were attached. When the Octopus came out and attacked the crab it received an electric shock, dropped the crab and retreated home.
Next time the crab and white plate were presented the Octopus makes a very cautious approach, advancing very slowly and “testing” the crab with the tip of a tentacle to discover whether it will get a shock. After receiving the shock a few times in this way the animal ceases to come out at all when the crab and white plate are shown, but continues to attack crabs put in by themselves. The memory of the painful association of the white plate is then carried for several days.
Some progress has been made in finding the part of the brain that stores this memory. The brain of the Octopus has many lobes, arranged on a plan very differently from our own. The uppermost lobe of all consists of a mass of small cells. When this was removed from one of the trained Octopuses the animal immediately lost the “memory” that the white plate was connected with a painful stimulus. Animals that for many days had remained at home when the white plate was shown came out at once after removal of this highest part of the brain. Moreover in spite of receiving shocks they provide quite unable to learn not to attack the white plate. After receiving a shock the operated Octopus would go home, but if the crab and white plate were shown again ten minutes later it would come dashing out and receive another shock.
Thus we can say that the power to learn to refrain from attacking lies in this uppermost part of the brain. This power is clearly very like the memory of higher animals or man. The interest of the experiment is that it opens some hope for finding what changes go on in the brain when animals learn. We now know where to look in the Octopus brain and we hope that by comparing this part with these regions of our own brain that are connected with learning, we may be able to find the common fact that makes memory possible.
Notes of Treatment for Exhibition
1. Living Octopus
Octopus in good condition feed regularly. This would be a very great attraction, but probably too difficult to achieve. There would have to be a good circulation of sea water at about 20 C. The Octopus could be readily obtained at Naples from March – October, less easily in the winter. They would probably fly back all right but this has not been proved.
Before deciding the problem, consult the Zoo. They will advise on the difficulty of maintaining the sea water in a suitable condition. It would mean elaborate backstage tanks. The Zoo have kept Octopuses for some months but have not found them very good subjects. Probably they do not like the strong lighting. Supplies can also be obtained from the Channel Islands at some times of year. I would very much welcome a live acquarium but must advise that the risk of failure is serious. If successful it would make a very attractive exhibit.
We have a very good set of still and moving pictures and it would be easy to make good models from these. The enclosed are only a small sample, my photographer being laid up we cannot supply further prints for some weeks but they will be good.
Models with moving parts would difficult unless they were diagrammatic. A possible treatment is by means of a series of models of the Octopus in its training tank. (1) sitting in the home, with head and eyes showing and a pile of crab shells outside, (2) emerging from the home when a crab and white plate are shown (3) swimming in the water, (4) seizing the crab and getting a shock, (5) coming out very cautiously next time, (6) not coming out at all when trained, (7) model of the brain showing the part removed at operation, (8) the Octopus attacking again after the operation.
We can of course supply any number of graphs, to show the progress of learning, number of mistakes, etc. and if you want it, much further matter about controls, operations or other parts of the brain, etc., Some of these would be desirable if the exhibit was to be thoroughly convincing scientifically.
Some of the above models could be made to move without having moving parts. Thus (1) The Octopus could lean forward and change colour when a level lowered the crab into its tank. (2) Very difficult to show in motion because shape changes quickly. (3) Would show well – the Octopus moves in the water by its jet without changing shape. The difficulty would be to know what to make it down when it arrived. (4) Difficult. (5) Possible. The Octopus near the crab could put out a cautious tentacle, touch the crab and draw it quickly back. (6) Slight changes could be shown as in (1).
A diagrammatic moving model could easily be devised to show the whole sequence of events by the movement of an ‘Octopus’ of simplified form.
I incline to think that the living show is too risky, but am sure that we should get excellent results from a combination of naturalistic and diagrammatic models.