Moderators: kiore, Blip, The_Metatron
HughMcB wrote:Holy crap that was amazing. Thats some amazingly complicated behaviour.
natselrox wrote:Life is amazing, Hugh!!! And at moments like this, I feel happy that I quit Physics for Biology!
chairman bill wrote:I've tried making injured gazelle calls & threatening the wife with lions. Might work better if we lived in Africa
natselrox wrote:HughMcB wrote:Holy crap that was amazing. Thats some amazingly complicated behaviour.
Think of the risks, the gains, the complicated calculations required to evolve such a complicated behaviour. All done over an evolutionary timescale by the mindless trial-and-error process of natural selection.
Mr.Samsa wrote:natselrox wrote:HughMcB wrote:Holy crap that was amazing. Thats some amazingly complicated behaviour.
Think of the risks, the gains, the complicated calculations required to evolve such a complicated behaviour. All done over an evolutionary timescale by the mindless trial-and-error process of natural selection.
How did they establish that it was an evolutionary behavior? I haven't had a chance to read the paper yet..
natselrox wrote:
Oh! I haven't read the actual paper either.
Either these bugs are calculating the risk-vs.-gain of such a behaviour in their head (seems unlikely) or they are hardwired to behave as such. Both the options seem to be an evolutionary behaviour to me.
Mr.Samsa wrote:Keep in mind that the processes which underpin how we learn ("we" as in "all living things") are selectionist principles; in other words, behaviors are culled or strengthened according to their consequences. Evolution on the time-scale of the individual. So if a behavior is able to be produced through evolution, then there's a good chance that they can be learnt as well.
*off to read the article now*
natselrox wrote:
"Evolution on the time-scale of the individual!" I have been searching for that phrase for quite some time now. Glad that you mentioned it.
natselrox wrote:I know what you are talking about, but it seems almost impossible in a bug of that size. My intuitions are not to be taken too seriously though...
Mr.Samsa wrote:natselrox wrote:
"Evolution on the time-scale of the individual!" I have been searching for that phrase for quite some time now. Glad that you mentioned it.
You'll probably find more stuff on it by searching for ontogenic learning, which is the development of behaviors over a lifetime, compared to phylogenic "learning" which is the evolutionary process that produces genetic behaviors. Specifically you can look for papers containing the phrase "selection by consequences", like this one: A Computational Model of Selection by Consequences - McDowell. (It's a theoretical paper, but it's an interesting read especially for background).
natselrox wrote:I know what you are talking about, but it seems almost impossible in a bug of that size. My intuitions are not to be taken too seriously though...
I'll have to try to find some papers on insect behavior. They aren't often used in experiments due to their limited lifespan and the difficulties involved in measuring their behavior (you can't exactly get them to press a lever). They still learn things though, there was an experiment on woodlice that used sunlight as an aversive stimulus and they managed to train them to perform some pretty complex behaviors in order to avoid being exposed to sunlight. I can't remember any of the specific details though
Thinking of behaviors as a "risk/benefit analysis" is a bit misguided though, no animal really does this - even humans. Instead the processes rely on past experiences which have shaped specific patterns of behaviors. In such a case, it may be that the female waterstrider has evolved to avoid excess surface disturbance, and when the males are rewarded by less resistance as a result of their "tapping" (during the struggle to mount the female) then they will be more likely to engage in this behavior in the future. This doesn't mean that they are weighing up their odds of attracting a predator, it just means that they know if they continue with a certain behavior then they're likely to score. (Still reading the paper though, so I'm not sure if that is valid speculation or not).
natselrox wrote:
I've read bits and bobs of this but never really did the hard-work. But I am slightly wary of allowing too much plasticity in any animal's behaviour.
natselrox wrote:Like all complex behaviours, this can surely be broken into simple steps each consisting of a phylogenic behaviour.
natselrox wrote:But I somehow doubt that this would be the case in an insect. More complex circuits tend to be transmitted via the hereditary route in insects (Bee dance, for example. It would be very hard to assume that they learn to dance by a positive reinforcement mechanism. Surely they can perfect the dance by such a mechanism but I'm not sure if any research has been done to show that they even do this. The dance seems to be inherited as a whole.). There's not much that you program in a computer that doesn't have the required hardware.
Am I making sense?
natselrox wrote:I'd say that we all do risk-benefit analyses, usually subconsciously. Our whole brain represents a structure which is based on the hierarchy of needs. The result is that most of the physiological functions essential for staying alive are involuntary. This in itself represents a very complex risk-benefit analysis in me. And it has been done over the evolutionary timescale.
natselrox wrote:Your mechanism seems plausible but it still does not rule out the c/b analysis.
natselrox wrote:If I had been a student of Neuroethology, you'd be a great person to debate against! I'm already starting to enjoy our different perspectives of looking at things!
EDIT: And it's kinda fitting that you have a Skinner quote as your sig!
natselrox wrote:Like all complex behaviours, this can surely be broken into simple steps each consisting of a phylogenic behaviour.
Uh... How so? Take riding a bike for instance. We obviously need genetically programmed things like breathing regulation, balance-checks, etc - all the essential things we need to run 'automatically' so we don't die when we forget to breathe or whatever. But learning to balance yourself on two wheels, adjusting to the forward and backward effects of peddling, using the brakes effectively, etc are all mostly (if not entirely) learnt behaviors. Learning mechanisms are so powerful that they can actually override genetic behaviors.
So if you mean "underneath it all" there are phylogenic behaviors, then obviously I agree but at that level we're no longer discussing the same behavior. It's like pointing out that at the base of all flocking behavior of birds there is gravity - technically true, but practically useless when trying to predict and understand their behavior.
I can't recall the stuff I've read on the bee dance, but I'm pretty sure it's agreed that it's a genetic thing - although learning does modify it to some degree, I think?
Hierarchy of needs? That sounds very Maslow-ish. Unfortunately, the idea that our behaviors are developed as a result of some underlying force which requires us to meet certain needs or desires is not what is accepted in science - it's known as Drive Theory (or Drive-Reduction Theory, depending on which angle you look at it). The basic concept is that we have things like a Hunger Drive, and a Sex Drive, Thirst Drive, etc. But researchers very quickly discovered that to try to understand behavior in this way, we need to create an almost infinite number of drives and, as such, the whole enterprise became unfalsifiable. Anything could be explained by calling it a drive. It also failed miserably at explaining things like conditional reinforcement, where organisms could act in ways that contradict the drive that it is supposed to be adhering to.
There were a couple of other attempts to explain behavior in a similar way, but they all eventually failed. Now we focus on underlying rules or laws that seem to govern all behavior, no matter the species. Genetic imperatives are still important to consider, for example, a rat responds very well to sugar, but pouring sugar into a fishtank creates unhappy behaviors in the fish. So there are certainly physical requirements that need to be considered for individual species, but they are essentially just a variable, rather than being the driving force behind behavior.
natselrox wrote:
Yes. There's only a limited amount of complex behaviours that you can form by permutation and combination of the simpler parts.
natselrox wrote:All the components of bike-riding can indeed be viewed as learnt behaviours. But there are subtler components to it that are not learnt. The instinctual correcting of one's body-position on processing the data from the vestibulocochlear nerve, for example, is not something that is learnt.
natselrox wrote:And I think it is somewhat myopic to say that learnt behaviours can override genetic instincts. Of course, the statement is valid. And it's validity varies across the animal kingdom, but given that we are in a thread discussing insect-behaviour, I'd say this looks a bit odd.
natselrox wrote:So if you mean "underneath it all" there are phylogenic behaviors, then obviously I agree but at that level we're no longer discussing the same behavior. It's like pointing out that at the base of all flocking behavior of birds there is gravity - technically true, but practically useless when trying to predict and understand their behavior.
Come on! Now you are stretching this!
natselrox wrote:I can't recall the stuff I've read on the bee dance, but I'm pretty sure it's agreed that it's a genetic thing - although learning does modify it to some degree, I think?
Hmm... Would be interesting to see that.
natselrox wrote:Maslow must've been in my subconscious when I said that. I don't know about the drive-theories, but I know that some parts of the brain can indeed 'override' other parts. The more primitive parts like the limbic system, the amygdala etc. can definitely over-ride the cerebral cortex. Call it 'hunger-drive' or whatever, but it'd be very hard to make a hungry animal do things that drives it away from food. You can't teach an animal to stop breathing indefinitely. Learning has a limited influence over a small part of the physiology of animals. This involves a cost-benefit analysis IMO.
natselrox wrote:I hold the exact opposite view. I think, learning is an important variable behind behaviour, but physical 'requirements' are the driving force. (Oh! How I have assumed the extremist position as well!)
natselrox wrote:Skinner was a smart guy but he's outdated now.
Mr.Samsa wrote:natselrox wrote:
Yes. There's only a limited amount of complex behaviours that you can form by permutation and combination of the simpler parts.
You reckon? Don't you think you're forgetting about emergence? When we combine a few simple parts, we have the sum of the parts as well as the novel behavior which results as the interaction between them all. Then when you combine these emergent behaviors with other emergent behaviors, you produce even higher level behaviors. Theoretically, this could continue on recursively forever - and most likely, this is the basis for human thought.
natselrox wrote:All the components of bike-riding can indeed be viewed as learnt behaviours. But there are subtler components to it that are not learnt. The instinctual correcting of one's body-position on processing the data from the vestibulocochlear nerve, for example, is not something that is learnt.
Not quite, the vestibulocochlear nerve is vital for maintaining balance, but the process of correcting yourself as a result of the information gained from the nerve is something that requires learning. This is why newborns fall over a lot, and why when we're put in weird situations (like zero-gravity etc) we are initially "off-balance" but gradually learn how to understand the information they are receiving.
This is the problem with trying to separate fundamental behaviors out into nature and nurture.
(Also, don't get me started on the useless concept of "instinct".. )
natselrox wrote:And I think it is somewhat myopic to say that learnt behaviours can override genetic instincts. Of course, the statement is valid. And it's validity varies across the animal kingdom, but given that we are in a thread discussing insect-behaviour, I'd say this looks a bit odd.
How so? Insect behavior can be changed as a result of learning. Fixed-action patterns, which govern most insect behavior, aren't impossible to change despite them being hugely important behaviors in insects (and animals).
natselrox wrote:So if you mean "underneath it all" there are phylogenic behaviors, then obviously I agree but at that level we're no longer discussing the same behavior. It's like pointing out that at the base of all flocking behavior of birds there is gravity - technically true, but practically useless when trying to predict and understand their behavior.
Come on! Now you are stretching this!
Not at all! If we're going to trace learning to ride a bike with the basic bodily processes that are necessary for survival, then surely we've gone back too far? I mean, I know that to be able to learn how to throw a basketball we need muscles and tendons in our arms, but having muscles and tendons doesn't tell us a whole lot about throwing basketballs.
Of course we can teach animals to stop breathing indefinitely! They just pass out though and start breathing again. But whatever kind of "need" or drive that you can think of, there is a way to learn how to override it.
Food? Self-control, people and animals often forgo immediate food for something later - even when they're hungry.
Sex? We have celibate nuns and priests (or so they say).
Preservation of life? Suicide.
As mentioned above though, changing fixed-action patterns is the best example of this. Herring gull chicks, for example, respond to a red dot on their mother's beak and this is how they feed. But, with training or specific environmental conditions, this can be changed - you can make them respond to different colours, simply stop responding to red dots, make them act like a chicken when they see red dots instead of trying to feed, etc.
natselrox wrote:Skinner was a smart guy but he's outdated now.
Oh, the gloves are off now, bitch!
Nah, obviously you're right, in the same way that Darwin is outdated. Specific ideas he had have been adjusted, or rejected, and generally developed according to the evidence that has accumulated over time, but his general idea that behavior can be modified as a result of consequences to behavior is undeniably true. (Oh, and just to be ultra clear, Skinner argued that to understand the behavior of an organism then you have to understand the genetic factors as well as the environmental factors - often he's viewed as someone who rejected the "nature" side of the debate, but in reality he was firmly placed in both camps).
natselrox wrote:Fuck it! The Green Vagina is more interesting than Suppurative Otitis Media.
natselrox wrote:
Now you're moving into territories of human thought and other higher cognitive functions. Remember that we started with a mating strategy of a simple organism. Although I'd like to think of the higher cognitive functions as being much less complex as you'd like to say. But that is another discussion.
natselrox wrote:Funny how we are looking at the same thing and coming to completely different conclusions! Where you see this activity-mediated influence on the developing brain in the critical time period as an proof of the importance of 'learning', I see it as the irrefutable evidence of inherited 'instincts'. The correcting of the body-position by the infant is totally dependent on a feedback loop that feeds the information about the current position and the final one, thereby modifying the synapses that regulate this behaviour. This shows that the minority of physiological functions controlled by neurons susceptible to Hebbs-ian modification are indeed dependent on a complex inherited circuit and external influence can only slightly modify them only within a small time frame.
natselrox wrote:This is the problem with trying to separate fundamental behaviors out into nature and nurture.
To put it bluntly, nurture only influences a tiny fraction of our behaviour and is probably insignificant in studying most of ethology.
natselrox wrote:(Also, don't get me started on the useless concept of "instinct".. )
Instincts are the most important things in an animal's life. In the Dale-Purves text, it is written on 'Phylogenetic Memory', that this is the "most important form of stored information in determining whether the individual survives long enough to reproduce." (Chapter 30. Box A, Page 735) it's an interesting read about how hatchlings hide after seeing specific patterns in shadows without ever 'learning' it.
natselrox wrote:
How so? Insect behavior can be changed as a result of learning. Fixed-action patterns, which govern most insect behavior, aren't impossible to change despite them being hugely important behaviors in insects (and animals).
Even if most of the neuronal circuits do not have the necessary plasticity? Plasticity is a trivial weapon in the behavioural arsenal of an animal. Try teaching a monkey not to be scared of snakes without cutting the amygdala!
natselrox wrote:Really? When someone throws a basketball at you, you subconsciously calculate the trajectory, assume the final position, gauge the dimensions of the ball and move your hands in a way so they just catch them (depending on your athletic ability, of course.) Tell me, how many of these components does one need to 'learn'?
natselrox wrote:Of course we can teach animals to stop breathing indefinitely! They just pass out though and start breathing again. But whatever kind of "need" or drive that you can think of, there is a way to learn how to override it.
Now this is too much!
natselrox wrote:Food? Self-control, people and animals often forgo immediate food for something later - even when they're hungry.
The enteric nervous system has 100 million neurons almost the same as that of the spinal cord. And they have almost zero plasticity. How are you going to modify something that is fixed.
natselrox wrote:Forgoing of immediate food for something later is often an evolutionary feature.
natselrox wrote:Sex? We have celibate nuns and priests (or so they say).
But no celibate dogs/bacteria/mouse. You are thinking human, mate. All too human.Preservation of life? Suicide.
Show me one example where an animal commits suicide before attaining the reproductive age.
natselrox wrote:As mentioned above though, changing fixed-action patterns is the best example of this. Herring gull chicks, for example, respond to a red dot on their mother's beak and this is how they feed. But, with training or specific environmental conditions, this can be changed - you can make them respond to different colours, simply stop responding to red dots, make them act like a chicken when they see red dots instead of trying to feed, etc.
Of course, you can do that. Remember that pic where the gooslings follow Tinbergen? But that means that a few of our behaviours have been left plastic enough for a small time-window by evolution. Nothing more.
natselrox wrote:Of course, Darwin is outdated. His role may be valued in History books but Science is a cold and heartless bitch without any place for sentimental attachment or historical relevance.
natselrox wrote:HughMcB wrote:Holy crap that was amazing. Thats some amazingly complicated behaviour.
Think of the risks, the gains, the complicated calculations required to evolve such a complicated behaviour. All done over an evolutionary timescale by the mindless trial-and-error process of natural selection.
Life is amazing, Hugh!!! And at moments like this, I feel happy that I quit Physics for Biology!
Users viewing this topic: No registered users and 1 guest