zoon wrote:Mr.Samsa wrote:Is it possible that you're interpreting "consciousness" as "awareness"?
What distinction would you make between the two? The Oxford dictionary defines “conscious” as, first, “aware of and responding to one’s surroundings”.
The definition changes depending on what question you're asking, but generally all forms of the definition reduce down to "the ability to experience" - which is, of course, synonymous with 'sentient'. Only more advanced definitions of consciousness require awareness (since obviously it's possible to experience things without being aware of them).
zoon wrote:Mr.Samsa wrote:Firstly, I think I might have been unclear when I described the basic behavioral law that controls behavior. This is not a claim or belief of behaviorism, it is a conclusion from behavioral research. That is, Skinner didn't say "All behavior must or should be controlled by basic pain and pleasure principles". Instead, it just turns out that all behaviors that have been studied and understood have all been reducible to things like the matching law, or the contingency discriminability model. Essentially, all behaviors are reduced to choice theories because all behavior is choice behavior; so whenever you choose to do something (e.g. wave your arm around), you necessarily choose to refrain from doing something else. And all the currently available choice theories rely on "pain" and "pleasure" parameters.
A thermostat can be programmed to close a switch in response to a sensed temperature above a certain level (a heating mechanism is then closed and the thermostat cools). It would also be possible to program the thermostat to learn to close the switch, for example, if a light flash was followed on two occasions by a rise in temperature above the set level, the thermostat could be programmed to close the switch after any further light flashes, instead of waiting to sense the temperature going up.
If a thermostat was programmed to learn in that way, then presumably the thermostat, after learning from two or more light flashes followed by a rise in temperature, chooses the behaviour of closing the switch when the light flashes.
I'm not sure what you mean by "light flashes". Do you mean literally turning a flashlight on and off at a thermostat with some light sensor? If so, I don't see how this differs from kicking a rock and saying it has "learnt" to move away from a boot the moment it is touched.
zoon wrote:This would be because the thermostat has pleasure in a cool temperature and pain when the temperature gets too high? Someone who did not know anything about programming or thermostats might reach that conclusion, after observing its behaviour. And that method of predicting the thermostat would actually be correct, it would work – the thermostat acts to reduce temperatures above the set level, or when it’s learnt that the temperature is likely to go above the set level. Pragmatically, the thermostat would be correctly described as reacting to painful or pleasurable stimuli, and as learning.
However, for anyone who did know about thermostats and programming, the available theories would not be limited to reliance on “pain” and “pleasure”. There would be an alternative explanation for the thermostat’s behaviour in terms of electrical circuits and the properties of switches and heating mechanisms.
I don't think it would be possible to describe the thermostat's behavior in terms of "pleasure" and "pain". Obviously, as I mentioned earlier, I'm using those terms loosely to make conversation easier and they refer to more complicated variables, and importantly, the "pain" and "pleasure" are objectively defined and not inferred. For example, the thermostat's behavior would not be enough to attribute the label "pain" to it, as it does not exhibit the fundamental properties of what we define as a pain response - specifically, there are no avoidance or escape responses when presented with the aversive stimulus.
For example, if a significant rise in temperature was preceded by a slight rise in temperature 5 minutes earlier, then the thermostat should learn the first rise as a discriminative stimulus. As such, it should significantly reduce the temperature to counteract the future effects of a temperature rise.
If a thermostat could be designed to learn in this way, then not only have there been huge advances in AI, but I'd probably consider such a thing conscious.
zoon wrote:It seems to me that “pain” and “pleasure” are useful markers for ignorance, when we don’t know the physical mechanism for an animal’s behaviour (the mechanisms in even the simplest creatures like bacteria are fantastically complex and barely beginning to be understood).
When we use “pain” and “pleasure” as explanations, we are using Theory of Mind, which humans use more than other animals. We probably evolved it primarily to predict fellow humans, but it also works for predicting many non-human animals. If I say an animal, or a thermostat, acts in response to pain, then I’m putting myself in the place of the animal, or thermostat, and predicting what I would do. This all goes on unconsciously, I’m not deliberately pretending to be a thermostat, I’m just using my own action circuitry to predict what the thermostat will do, and it works.
I understand the point you're trying to make, but I'm using the scientific definitions which are necessarily true or not, with no need to infer mental states or invoke ToM.
rEvolutionist wrote:Mr.Samsa wrote:You seem to be describing "sapience" here.
Probably. As I said, more "sapient" beings will have greater sense of potential loss, will have a greater understanding of what is happening to their fellow animals that are being killed for consumption. They're more likely to be able hypothesise about the future and potentially suffer more due to that. Would you agree that "suffering" would be greater for more sapient creatures?
Indeed - as I mentioned earlier, I think organisms with more advanced cognitive functions experience qualitatively and quantitatively more suffering.
rEvolutionist wrote: There is nothing necessarily wrong with this, but it does fall into the trap Hugh describes above of valuing things which are human-like.
Why is that necessarily a "trap"?
As I said, I see no reason to attribute more or less importance or significance to this value system over another. And in a way it makes more sense, as it is the most relevant system for us to utilise as humans.
The trap is that it's essentially arbitrary - like defining alien life as having to be humanoid in shape, and then concluding that alien life doesn't exist even if we found a planet full of flora.
rEvolutionist wrote:EDIT: I forgot to mention this, but plants don't just respond in a 'chemical/physiological' way (at least no more than animals). They actively employ methods to avoid pain, even without the painful stimulus present, etc. There is more complex learning and behavior involved in plants, and not just a simple cause-and-effect type response.
I didn't read Trewaras' (sp?) big paper, but read his shorter one in
Nature. I can't say I was convinced by anything he said. Perhaps there was something extra in his big paper that wasn't in the shorter one? He didn't seem to give many examples in the shorter one, other than direct physiological reactions to stimulus (like growing towards light, or away from adverse soil conditions).
What about the ability to navigate a maze? If you track the responses of a rat with some kind of dye, and then watched the behavior of a plant on some time-series camera, you find that the routes taken are indistinguishable. (Also, his points of growing towards the light and aversion to soil conditions are more complex than how you seem to be presenting them - I do recommend reading his full article some time. I'm not saying he's right, and I don't agree with everything he says, but it's certainly an interesting perspective and I think at the very least he forces us to reconsider some of our understanding of plant behavior).
rEvolutionist wrote:Behaviorism isn't a theory, it's a methodology. It can't be right or wrong in the same way "biology" can't be right or wrong. The only way it can be ignored or rejected is if it can be shown to be useless or unable to predict the subject matter it seeks to explain.
It's tenets could be wrong in the same way that biology's tenets could be wrong. But the underlying tenet of behaviourism - that behavioural organisms seek pleasure and avoid pain, is unchallengeable. There's no scenario that can be presented (well, that I can think of anyway) which couldn't be addressed by this tenet. Someone could naively say that those choosing suicide over life would be seeking pain and avoiding pleasure. But the obvious answer to this is that the person is seeking pleasure by avoiding the pain of life.
But as I mentioned below, that isn't a tenet of behaviorism at all (nevermind an underlying tenet). The basic claims of behaviorism are simply: a science of behavior is possible, mentalistic and anthropomorphic explanations should be avoided, circular explanations and explanatory fictions should be avoided, and cognitions can be studied and understood.
There is no specific requirement to understand behavior in terms of conditioning, or pleasure and pain, or anything. The fact that behavior IS understood in these terms is a result of overwhelming scientific evidence. That is, we wouldn't just say "the suicidal person is seeking pleasure by avoiding the pain of life", instead we'd experimentally test their responses to a number of conditions. Obviously this example is difficult since it's probably unethical to test a suicidal person in such a way that could encourage their behaviors, but what we'd do is put them in some setup which tested their response to "choosing life" and responses to "killing themselves" (I'm not sure how exactly this would be done), and then we'd take a couple of models of behavior to see which described the behavior the most accurately.
If a model that didn't include "pain" and "pleasure" explained it better, then we would reject all of those models as it has been disproved. (And, to clarify, I'm using "pain" and "pleasure" as very simple labels to describe more complex phenomena, but they're accurate enough I think to get the point across).
rEvolutionist wrote:I'm not saying anything about the veracity of Behaviourism here. I find it absolutely fascinating. Before I read what you had to say about it, I knew next to nothing about it. In fact, it seems the beautiful simplicity of it is its strongest point (i.e. it can effectively answer all questions about the act of living - at least at a general level).
I can't disagree with that
rEvolutionist wrote:It's arguable whether it specifically comes from evolution. The behavioral law could be a natural consequence of having an organism that avoids bad things and approaches good things.
I don't get it? Surely "bad things" would have an evolutionary selection pressure?
Yes, I'm pretty sure the ability to experience "pleasure" and "pain" are a result of evolutionary pressures, however, this does not necessarily mean that the behavioral law that controls our behavior through these processes was specifically selected for. That is, the behavioral law could be a necessary derivation of these simple experiences.
As an analogy, imagine a river. The fixed parameters of the path it will take are: 1) gravity, and 2) the solidity of the surrounding ground (just keeping at these two factors for simplicity's sake). Now, just because those two factors are fixed does not mean that the resulting river was selected or designed by the mountain - it was a necessary consequence of the interaction of the two variables.
rEvolutionist wrote:You mean on a neurological level? Long-term potentiation and in-vitro reinforcement appear to the mechanisms which control classical and operant conditioning.
I guess I'm asking how we scale up from say a tree root, or a simple animal, steering away from a negative stimulus, right up to humans abstracting and hypothesising about the future and behaving based on that seemingly immaterial stimulus.
It depends what question we're asking, i.e. whether we're asking if there's an absolute difference between humans and plants, or a relative one. On an absolute level, there is no difference between our responses and theirs (if we accept some of the examples given by Trewavas). But on a relative level, of course we differ. To paraphrase Darwin, the difference between animal and plant is one of degree, not kind.
rEvolutionist wrote:How'd I do?
Not good enough.
I want more info. I still don't fully appreciate what it is Behaviourism does and tells us.
Well behaviorism doesn't "tell us" anything, since it's just a methodology. But the results of sciences based on behaviorism tell us that the behavior of humans and animals can be described, understood and predicted using very simple laws. One of my entries to the science writing competition outlined in detail what behaviorism is:
Misunderstanding Behaviorism, if you were interested.
rEvolutionist wrote:Is it possible that you're interpreting "consciousness" as "awareness"?
I don't know. To continue my reasoning from above, I guess I think that "suffering" requires more than just simple "experience". It requires a deeper experience that implies some sort of understanding of sorts. i.e. Sapience, as you say above.
As I mentioned at the beginning of this thread, "suffering" (in science at least) is defined as experiencing a state of pain or aversion that is either high in intensity, or occurs over a prolonged period of time. Therefore, if something can experience pain, then it can suffer.
The question that I suppose you'd want to ask next would be: "Is this suffering worth worrying about, or should it be considered comparable to the suffering of other animals?".
rEvolutionist wrote:Firstly, I think I might have been unclear when I described the basic behavioral law that controls behavior. This is not a claim or belief of behaviorism, it is a conclusion from behavioral research. That is, Skinner didn't say "All behavior must or should be controlled by basic pain and pleasure principles". Instead, it just turns out that all behaviors that have been studied and understood have all been reducible to things like the matching law, or the contingency discriminability model. Essentially, all behaviors are reduced to choice theories because all behavior is choice behavior; so whenever you choose to do something (e.g. wave your arm around), you necessarily choose to refrain from doing something else. And all the currently available choice theories rely on "pain" and "pleasure" parameters.
That's fine, I wasn't implying otherwise. It also doesn't change what I was saying.
You described it as an "underlying tenet of behaviorism" though? The point I was just trying to make is that describing it in such a way, is like saying an underlying tenet of biology was evolution. Evolution is obviously important to biology, but only because it's such an important finding backed by scientific data, not because it's an assumption of the field.
I appreciate that this is not what you intended to imply, but I just wanted to clarify it just in case.
rEvolutionist wrote:Secondly, you're defining "variability" too narrowly. The "variable" in "adaptively variable behavior" simply means that the topography of a behavior can change in response to the same task or challenge. So if we put a rock in a maze, it's unlikely that it will find the cheese at the end. If we put a wind-up car in a maze, it's unlikely it will find the cheese at the end, etc, but if we put a rat in a maze, it will display a number of behaviors (e.g. 'go straight', 'turn left', 'turn right' etc) and this variability in behavior is adaptive (i.e. incorrect responses are extinguished, and correct ones are strengthened).
Yeah, I get what you are saying, but I guess it depends on the ability to separate variable movement out from an invariable "drive". As an example of what I am trying to say, think of a tree growing in part shade under other trees. There is an invariable goal "driving" its growth, and that is the need to get more sunlight on its leaves. But depending on other variables - herbivores, disease, wind, humidity etc. - it might display a different "behaviour" by growing towards a different area of higher light. But I wouldn't say this plant is showing variably adaptive learning. Perhaps this isn't the best analogy, as you might very well say that the plant is displaying this. Meh, i've agonised over this paragraph for a while now and I just can't nail what I am trying to say. I'll leave it half out there, in case it prompts me at a later point to get to the crux of my problem.
I see what you're trying to get at in your 'erased' paragraph. You're trying to get at some notion of 'agency', in that whilst the plants behavior might have 'variability', you think this is just because other forces are pulling it around, rather than it 'choosing' one option over the other.
rEvolutionist wrote:Bloody pigeons. I'll never look at them the same ever again!
Blip wrote:Fascinating stuff, thank you. I'm predictably perturbed that the birds are kept markedly underweight in order to motivate them to take part, but that's for another discussion.
Don't worry, this is a common misreading of experimental procedure. The pigeons aren't kept underweight, they are on restricted feeding where they are kept at around 85% of their
freefeeding weight. In other words, for a period of time they are given access to as much food as they like, and they pig out until they essentially reach their maximum size and weight. So it's like being taking to an all-you-can-eat for every meal of every day for about a month. Then we calculate 85% of this weight.
The end result is that they are within the correct weight range for what vets consider to be a healthy bird, and some approach the heavier end of this recommended weight. This means that they are far heavier than any wild bird, and generally fatter than any pet bird. In addition, the training sessions involve receiving food rewards, so they get plenty of food. So you don't have to worry about birds being "starved", as restricting their food actually means keeping them within the recommended weight range.
Blip wrote:I am not sure that, in the final analysis, these experiments don't simply reveal that pigeons can see and hear while fish can hear, which I knew already.
I'm not sure what you mean Blip? Seeing and hearing are not enough to solve these discrimination tasks. To discriminate between two classes of objects, the organism needs to develop a concept of what each class represents. Since they were tested using stimuli that they had never seen or heard before, then the only way to correctly complete the task is to identify elements of consistency between the two.
Blip wrote:I'm now wondering if my choice of the word 'sentient' was the best one: I'm used to choosing my words with care, but not in a scientific context. I suppose in the end it's the propensity for
suffering that I'm seeking to establish, which takes us into the realms of consciousness, awareness and self-awareness (zoon and rEv speak of this above).
My own research in this area led me long ago to abjure intensively-reared animals (including fish - the sight of salmon trying to escape their captivity in fish farms would distress the hardest of hearts) but my current ethical dilemma is around eating wild fish.
Anyway, it's a fascinating discussion, that notwithstanding.
I suppose we can ignore all the scientific terminology for the moment and I'll just ask you one question: Do you think fish can feel pain? If so, then they must be able to suffer.
For me personally, it doesn't matter whether they have the cognitive functions necessary to contemplate their pain or awareness of their surroundings, but simply if it can experience pain then I'd want to avoid causing it unnecessary pain.
cavarka9 wrote:hello, sorry for interrupting,is it possible to have a condensed version of where people stand now. Do fish have sentience or not,
I still stand by the position that fish are sentient.
cavarka9 wrote:also we do have pigeons living in our balcony to raise their young-lings, they keep coming into our home, they probably know that we are not threatening.
They will be habituated to your presence, in the same way someone living in a big city becomes habituated to the sound of cars driving past and alarms going off.