Question for Steve Chang. Not sure if I eared it well, but I think you mention something about rhesus monkey’s performance with mirror self-recognition task… Did you say that they only past the test if we train them to do this task? I would like to ear you about it. After training, are they really recognizing themselves? Do you think that this is a limit case?
Training to rub off a spot in the mirror is not the same as being able to do it spontaneously. A student of Skinner's, Robert Epstein, had already shown it could be trained in pigeons:
Michel, P., Gold, K., & Scassellati, B. (2004, September). Motion-based robotic self-recognition. In Intelligent Robots and Systems, 2004.(IROS 2004). Proceedings. 2004 IEEE/RSJ International Conference on (Vol. 3, pp. 2763-2768). IEEE.
All it needs is to learn the correlation between self-movement and image-movement. It's more of a puzzle why all organisms with the capacity for vision and for learning don't learn it. Possibly because the sight of their species has inborn reactions that out-weigh the tendency to learn the sensorimotor correlations spontaneously -- except in those species that do learn them. (After all, that's all there is to mirror-self-recognition and self-grooming. It does not give more or less evidence of "sense-of-self" (whatever that means!) than anything else.)
One of the question relevant to this summer school is were to draw the line in terms of sentience, meaning witch animals are sentient and witch are not. Our working definition for sentience is the capacity of an organism to feel anything. A lot of the expirements that are used rely on operant conditioning, using reinforcements to train animals before they do a task. Do you think that only organism who show that they learn througth this type of conditioning are sentient?
The title of the panel, « Others in mind », can refer to the interactions of the animals between themselves (ex. the « advertising » by the skin display in squids, or the way monkeys think about other monkeys) or to the ethical preoccupation we should demonstrate toward non-humans animals. Dr ten Cate believes there have been dramatic changes in animal experiments since a few decades : according to his own experience in the field, he thinks we are becoming more and more aware about what animal can do, already knowing it is more complex than we had expected. Beyond the law, there is ethical expectations. Primates are now « respected » according to their level of complexity but other successful species, like parrots, can still be left alone in a cage… Dr Chang reminds us that the actual legal criteria must be the balance between medical advantages for humans and the other species’ sentience, but Dr Mathers sees with skepticism the so-called medical benefit. She thinks we must anyway give the best care possible to all animals, and that hopefully in 25 years from now we willn’t need anymore of this kind of studies on animals. Dr Harnad thinks the worst isn’t in the labs, but in the plate (ie the farms). Sentience, ie ability to feel pain, should be the only criteria.
Question for Steve Chang. Not sure if I eared it well, but I think you mention something about rhesus monkey’s performance with mirror self-recognition task… Did you say that they only past the test if we train them to do this task? I would like to ear you about it. After training, are they really recognizing themselves? Do you think that this is a limit case?
ReplyDeleteAt least, I think it raised the possibility of many false negative results.
DeleteTraining to rub off a spot in the mirror is not the same as being able to do it spontaneously. A student of Skinner's, Robert Epstein, had already shown it could be trained in pigeons:
ReplyDeleteEpstein, R., Lanza, R. P., & Skinner, B. F. (1981). "Self-awareness" in the pigeon. Science, 212(4495), 695-696.
It can also be trained in robots:
Michel, P., Gold, K., & Scassellati, B. (2004, September). Motion-based robotic self-recognition. In Intelligent Robots and Systems, 2004.(IROS 2004). Proceedings. 2004 IEEE/RSJ International Conference on (Vol. 3, pp. 2763-2768). IEEE.
All it needs is to learn the correlation between self-movement and image-movement. It's more of a puzzle why all organisms with the capacity for vision and for learning don't learn it. Possibly because the sight of their species has inborn reactions that out-weigh the tendency to learn the sensorimotor correlations spontaneously -- except in those species that do learn them. (After all, that's all there is to mirror-self-recognition and self-grooming. It does not give more or less evidence of "sense-of-self" (whatever that means!) than anything else.)
Thank you for your comment. My aim was to raise the possibility of many false negative results, but I understand now why It doesn't.
DeleteOne of the question relevant to this summer school is were to draw the line in terms of sentience, meaning witch animals are sentient and witch are not. Our working definition for sentience is the capacity of an organism to feel anything. A lot of the expirements that are used rely on operant conditioning, using reinforcements to train animals before they do a task. Do you think that only organism who show that they learn througth this type of conditioning are sentient?
ReplyDeleteThe title of the panel, « Others in mind », can refer to the interactions of the animals between themselves (ex. the « advertising » by the skin display in squids, or the way monkeys think about other monkeys) or to the ethical preoccupation we should demonstrate toward non-humans animals. Dr ten Cate believes there have been dramatic changes in animal experiments since a few decades : according to his own experience in the field, he thinks we are becoming more and more aware about what animal can do, already knowing it is more complex than we had expected. Beyond the law, there is ethical expectations. Primates are now « respected » according to their level of complexity but other successful species, like parrots, can still be left alone in a cage… Dr Chang reminds us that the actual legal criteria must be the balance between medical advantages for humans and the other species’ sentience, but Dr Mathers sees with skepticism the so-called medical benefit. She thinks we must anyway give the best care possible to all animals, and that hopefully in 25 years from now we willn’t need anymore of this kind of studies on animals. Dr Harnad thinks the worst isn’t in the labs, but in the plate (ie the farms). Sentience, ie ability to feel pain, should be the only criteria.
ReplyDelete