Richard Dawkins answers your questions about evolution in honor of Darwin Day 2015.
“How does learned behavior evolve into inherited instinct?”
Edited by Stephanie Renee Guttormson
Copyright 2015 Richard Dawkins Foundation for Reason & Science
Sign In
If this is your first time visiting our website since January 2021, please reset your password to regain access.
Your account now works on all of our websites.
73 comments on “Darwin Day 2015 Questions: #6 How does learned behavior evolve into inherited instinct?”
I didn´t know nothing about it, but I suspected that was an important issue in Biology (after some lecture given actually by the psychiatrist/anthropologist that wrote the critical preface of the 1st edition of The Selfish Gene. The subject is quite interesting and I am really ignorant about it.
If my understanding is correct, rather than the original thrush in the Baldwin chain thinking ‘Ah look, I’ve found a snail I must look for an ‘anvil’, the event went as follows: Thrush finds snail, thrush thrashes snail about on everything including thin air and grass, snail randomly hits hard object and shatters and snail is smart enough to understand cause.
The specific learned behaviour does not evolve into a inherited instinct, thrushes do not have a inherited instinct to smash snails on anvils, they still have to learn this behaviour from other thrushes..
What can be selected is an aptitude for learning , thrushs better at learning , in this case snail bashing, have a reproductive advantage. That a better ability to learn confers selective advantage seems fairly obvious.
That the Social behaviour of populations involving learned behaviour, in this instance snail bashing, can have evolutionary effects also seems self-evident.
That the frequency of snail bashing through generations increases has a small evolutionary component, in that the new generations are slightly better at learning, but is mainly due to the fact that the passing on of learned abilities is geometric not arithmetical.
Beautiful. I have enjoyed this video so much. Like maria melo I am also very interesting in our instincts and what ever is left of them in us humans. Beautiful explanation.
I might be missing something but do not see that the video has anything to do with instincts. The learned behaviour does NOT become an instinct,. In this example no thrush instinctively bashes snails, it still has to learn it anew from other thrushes.
It is the aptitude of learning which is evolving, it is not learned behaviour evolving into an instinct. As he says ( to paraphrase) , “so that snail bashing SEEMS like an instinct.”
I sometimes wonder about the benefit/deficit implications of instincts. As with the thrush example the snails with a thicker shell might become predominant and withstand the onslaught by the birds but in the long run their genetic variability is reduced.
I also wonder about migratory birds like geese that might suffer from magnetic disruptions or an outright reversal of the poles, all the while further pressured by climate change effects. Species that have inherent instincts benefit in some ways but become potential victims in another light and might suffer from an inability to rapidly change ingrained behavior.
Steve I do not claim that I know much about instinct, but I think Dawkins have just explained in this video how something we humans call instinct evolved. Perhaps you definition of instinct is different (You have not explained). I agree with Dawkins explanation that at one point ability to learn becomes so fast that it looks as if it was not learned at all, but it comes naturally, becomes as something immanent.
Me to. But considering instinct I think that birds would not suffer, but they would simply change their paths,… they would still follow (have) magnetic pole as compass wherever they are.
🙂
Hi Modesti,
“Any behavior is instinctive if it is performed without being based upon prior experience (that is, in the absence of learning), and is therefore an expression of innate biological factors. Sea turtles, newly hatched on a beach, will automatically move toward the ocean.”
“The book Instinct (1961) established a number of criteria which distinguish instinctual from other kinds of behavior. To be considered instinctual, a behavior must: a) be automatic, b) be irresistible, c) occur at some point in development, d) be triggered by some event in the environment, e) occur in every member of the species, f) be unmodifiable, and g) govern behavior for which the organism needs no training (although the organism may profit from experience and to that degree the behavior is modifiable).[4]”
This behaviour does not meet any of these criteria. For this example thrushes do not a)automatically bash snails when they see them, b) do not HAVE to bash a snail everytime they see one, d) they do it when they want not in a trigger response, , e) not all do it, f) individual thrushes can modify their bashing technique and different populations can have different techniques ( i can remember reading somewhere that the milk bottle pecking had different techniques in different areas, e.g in one area all the birds pecked straight through, in others they pecked and lifted to force the top off) ,g ) they have to learn it by copying
In the main it is behaviour that requires no learning such as imprinting or a joey climbing towards its mother’s pouch , building a nest etc
To correctly say “that the learning happens so quickly that it seems like it was not learned at all” is STILL saying the behaviour IS learned, so it cannot be instinctive.
I can immediately and without thought answer the question 2+2=?, but that is not instinctive, I was not born innately knowing this, I was taught it
The book Instinct (1961) established a number of criteria which distinguish instinctual from other kinds of behavior. To be considered instinctual, a behavior must: a) be automatic, b) be irresistible, c) occur at some point in development, d) be triggered by some event in the environment, e) occur in every member of the species, f) be unmodifiable, and g) govern behavior for which the organism needs no training (although the organism may profit from experience and to that degree the behavior is modifiable).[4]
if it is only snail bashing!
I got timed out! More time Richard, I am an old man!
I was going to say that what is more amazing is that, as it is not instinctive, this learned behaviour shows thrushes etc have a culture, even if that culture does consist of snail bashing!
How did nest-building arise? Birds were not born with the skill, it must have been learnt before it became ‘instinctive’, surely? Infinitely more complex than merely bashing a snail…
Much has been written in recent times about epigenetics which seems to support the hypothesis that learned traits can be inherited, reviving the debate about Lamarckism. I’m surprised that RD dismisses this, since the great strength of scientific thought is the ability for self-correction…
The generality of Lamarckism is refuted, but epigenetics is the exception rather than the rule. It seems to be an incomplete function of the normal genetic “reset” which takes place in sperm and eggs.
With all due respect I don’t think the difference between learning and instinct are explained at all. As has been pointed out, building a bird’s nest seems instinctive but the hatched bird was not there to see the nest built, so how could it have learnt it? Similarly the hatched cuckoo knows it has to throw out the young or eggs of the host species. It obviously cannot be copying what it has seen because it has only just been born and vever saw its mother.
In the video Richard Dawkins tries to imply that the Baldwin effect means that animals learn so quickly that appears to be instinct but as these examples, and thousands of others, show, that is not the case. The Baldwin effect may well exist for things that are learnt but it cannot explain complex actions which the animal could not have seen.
When I commented on this site previously my suggestion that Lamarckism has a role as far as behaviour is concerned it was ridiculed but I think even Richard Dawkins is less dismissive of it than he used to be and evidence is accumulating that it exists.
Assuming the first snail could outrun the Thrush – but was unaware of this ability, he’d still be killed and unable to pass on his “speed” to further generations. And now I hope your understanding has improved.
You say “This behaviour does not meet any of these criteria.”. It seems to me that it meets ALL of the criteria. Don’t understand why you can’t see it!!
As far as I understand, learned behaviour does not evolve into inherited instict at all.
Either that, or Lamarck, not Darwin, was right.
http://phenomena.nationalgeographic.com/2013/12/01/mice-inherit-specific-memories-because-epigenetics/
Richard Dawkins is correct because what is being inherited is not some specific learned behavior, but the capacity to learn and that has a genetic base and can be selected for in the evolution of a population. To think that a specific behavior or memory can be inherited is a Lamarkian theory (the inheritance of acquired characteristics) and there is little or no evidence to support such a theory.
A very entertaining movie was produced I believe in 2005 called Aeon Flux and starred Charlize Theron. It was an excellent movie… good acting, good directing, good photography, and an excellent story line. However, the entire story revolved around a misconception that memories could be carried on DNA. That is Lamarkian and is entirely incorrect. Still it is fun to watch in spite of a central flaw. Memories would never be carried along in the clones of people from one generation to the next.
Please post the transcripts of remarks and responses, either in addition to a video or instead of. Reading is a better way to understand, absorb and consider a logical argument, and I am saddened by the trend on so many websites to post videos only.
I find that I can read much more quickly than I can watch a video, and I remember it better.
Agreed, but his wording is somewhat ambiguous, as is the question , both open to the (mis)interpretation that the learned behaviout itself becomes instinctual.
To be pedantic one would could say that “leaning” is an evolved aptitude or capability rather than an instinct, as instinct is a specific set of beaviours which learning is not.
The “problem ” being more grammatical than real as the term “instinct” is a rather fuzzy and imprecise biological term.
How?
But to learn to build a nest someone, Bird Einstein, must have already built a nest! Chicken and egg!
A more sound evolutionary hypothesis would be that birds with a hard wired propensity to “peck and throw” ,perhaps due to hormonal effects of about to lay, would accidentally “build” a proto-nest giving selective advantage, so that later generations would be hard wired better to ” peck and throw” until gradually it becomes a hard wired “instinct” to build a nest. A straight forward evolutionary explanation. Random hard wired behavior giving selection advantage so that it spreads to the whole population and becomes honed.
No need for learned behaviour to become instinctual, just a ordinary case of random mutation causing random behaviour which giving selective advantage is fixed into the genome. No mystery.
Darwin himself believed that there is a Lamarckian element to evolution so it is unfair to say he was wrong in that respect. It is those who came after Darwin, sometimes called neo Darwinists, who made a dogma of denying that characteristics or behaviour acquired in one generation could not be inherited by those following.
The example linked by John below is one example of recent discoveries leaning in a Lamarckian direction. Cbrown, also below, says it is the capacity to learn which is inherited. No one denies that the capacity to learn is inherited but the whole point of that discussion is that much of what we call instinct must have been learnt in previous generations and passed on as it could not possibly have been learnt (see my previous examples). Whether that is through the genes (which I admit seems unlikely) or in some other way of which we are not yet aware, cannot be said but it certainly cannot be accounted for by the Baldwin effect as Richard Dawkins suggests..
I had the same reaction. The fact is, unless you adhere to the Blank Slate theory that Pinker pretty much demolishes in his book of the same name, there ARE actual inherited behaviors but Prof. Dawkins didn’t seem to address that question at all here.
For example, the dance of bees used to communicate sources of new food is very complex behavior and almost certainly (at least mostly) in the genes of bees not learned by a general learning mechanism. Or language in humans. The rate that young children learn language and the kinds of mistakes that they do and don’t make (this is called the Poverty of Stimulus argument by linguists) indicate that there is a genetic predisposition for language in humans.
Maybe I’m missing something as well but Dawkins didn’t seem to address these kinds of actual “instinctual” behaviors at all which to me is a much more interesting question than how behaviors get passed down to successive generations via general learning.
Well, I was nodding my head in agreement until you got to the part about “Lamarckism”. Just because Prof. Dawkins qualified his dismissal of Lamarckism a bit doesn’t indicate to me he is “less dismissive” of it. He was just being a good scientist and essentially saying that of course virtually nothing is certain in science. What exactly is this evidence you are talking about that is accumulating for Lamarckism? I haven’t heard of any such evidence.
But I agree with you that this video really glossed over the distinction between learning and instinct. Although I don’t like the word instinct it has a lot of baggage from the past but if by instinct you mean behaviors that are encoded for in the genes then OK. As Pinker illustrates in The Blank Slate there are many forms of behavior that are almost certainly not primarily learned but that are part of the genetic code for the organism. Nest building in birds, damn building in beavers, web building in spiders, the dance of bees, and language in humans are all examples. Dawkins didn’t seem to address these kinds of genetic behaviors at all in this short video.
Surely one of the main thrusts of evolutionary theory is that the complex can arise out of the random through selection.
As in my explanation of “nesting evolution” above there is no need to invoke learned behaviour becoming instinct or Lamarckian possibilities.
If a genetic mutation causes a random hard wired behaviour which confers selective advantage then that behaviour can get fixed into the genome and we call it an instinct.
The first “nest” was a randomly assembled mess caused by a birds random hard wired propensity to “peck and throw” giving reproductive advantage leading to the complex instinctual nest building of today. Exactly the same explanation as the evolution of the eye and everything else!
I think that is absolutely wrong. Yes, there is a general learning ability but the idea that there is ONLY a general learning ability has been very soundly refuted by people like Steven Pinker. Pinker showed in The Blank Slate that there is overwhelming evidence for genetic predispositions for all sorts of complex behaviors in humans and other animals. As I’ve said in other comments here, web building in spiders, the dance of the bees, and language in humans are all examples where there is overwhelming evidence that they can’t be learned via a general learning ability but must have a strong genetic predisposition in the organism.
Just to be clear, learning and environment are also important and it is difficult if not impossible to completely and unambiguously demarcate where genetics stops and environment takes over. But that is no different than for many other kinds of genetic traits. If you deprive cats of the ability to see for the first year or so of their life they will never gain a mature vision capability. That doesn’t mean that vision is “learned” in cats. In the same sense if you deprive humans of verbal interaction for the first few years of their life they also won’t develop a complete language capability but again that doesn’t prove language is only learned and not significantly genetic.
And evolutionary theory explains how these inherited behaviours, bee dancing, arise.
Random mutation causing random behaviour which if advantageous gets the mutation fixed in the genome with the inherited behaviour being an ” hard wired instinct”
In reality behaviours are usually a combination of instinct, learning and aptitudes and each component has to be unpicked to fully explain and understand the behaviour.
I think he was only refering to learned behaviours not becoming inherited behaviours, not to instinctive inherited behaviours , such as bee dancing, which are explained by standard evolutionary theory.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1973965/
This is some recent evidence of Lamarckism as is the link posted by John above.
Apart from the bee dance, is there any proof that dam building (I bet the beavers say damn sometimes), spiders’ webs, human language etc have a genetic basis? I can’t speculate on what the mechanism for inheritance is but I find it hard to believe that genes can pass on the detailed plan as to how an archer fish can bring down an insect by blowing water or for the male bower bird to create such elaborate structures.
Genes pass on an elaborate sequence of events which build bodies with complex working mechanisms, so building nests and shooting a jet of water, should be a doddle!
Pecking and throwing would not build a nest however many millions of years passed and however good the progeny got at it . And even if against the odds of trillions to one, one was built by accident, the next generation would not know genetically how to do it again so it would not be passed on.
It is not like the evolution of the eye where tiny advantages can be built over time. A nest must be built according to a plan and if any part is omitted it will not work. The famous monkeys who type the works of Shakespeare by accident will only do it once and they will not pass on the skill to the following monkeys by random mutations.
I think it is exactly like that. Nests have been built at least since the time of the dinosaurs, and vary from scraped hollows in the ground, to holes in the ground, a few twigs on a cliff ledge or tree, right across a range to the very complex nests of weaver birds.
Like the eye, there is a full range of the stages of complexity represented in present day examples.
In complex nests such as those of weaver birds, the more experienced birds build better nests.
http://people.eku.edu/ritchisong/birdnests.html
Few studies have focused on the question of whether nest building by birds in an innate behavior or is, at least to some extent, learned. Two lines of evidence suggest that nest building has a large instinctive component: (1) the structure and composition of nests exhibit little intraspecific variation, and (2) birds raised in isolation tend to build species-specific nests. However, particularly for birds that build more complex nests, evidence indicates that birds learn by experience and, as a result, the quality of nests built by older birds may exceed that of nests built by younger, less experienced birds. For example, the first nests of young male Village Weavers (Ploceus cucullatus) are more loosely and crudely constructed than nests built by more experienced males.
Peck and throw is just a suggestion of the first behaviour for sake of the argument, it could have been peck-peck-drop, it is irrelevent to the logic of the argument.
They did not start with a complete nest just as a complete eye did not suddenly appear
One twig would give a tiny advantage, two twigs a further tiny, two twigs interleaved more……………….swan’s nest.
No plan needed, neither Plato or God dunnit.
Birds also have intelligence so they use that as well as instinct. As Alan4discussion says they learn to build better nests as they get older.
I be,ieve there is some geezer called Dawkins who apparently writes at length on these matters, maybe you should read him as he gives fuller and better explanations than me
This series of videos was very interesting. Thank you for making it 🙂 Dawkins has a remarkable capacity to explain complex theories in a simple, accessible way, but without taking away their profound explanatory power. It is a pleasure to watch videos like this one.
I agree withSteve. I was referring to only that behavior that is learned, for learned behavior is acquired, not something in the genome. The dance of the honey bees as Steve said IS inherited, thus is an expression of the genome. No bee must learn how to do a round dance or a waggle dance, that comes as an inherited behavior. I think I was not specific enough in my first statement.
Well said
If Steve is right that the information that two twigs is better than one is passed on genetically (and who I am I to say he is wrong – although, unlike the eye, there seems to have been no proof of it) then it must be a form of Lamarckism.
The contrary suggestion must be that a particular random mutation dictates two twigs, and that is advantageous. Unlike a mutation which allows an animal greater use of light, the next generation would have no advantage. Two twigs are no better than one and the mutation would not be fixed.
The evidence that is linked to one of my previous postings indicate that one generation has learnt something from a previous one in many examples be it plant, insects etc. It may be possibly done genetically but I cannot believe it is totally random.
Two twigs are very much better than one, if the second twigs stops eggs rolling off a cliff ledge or a tree branch or hides them better from predators.
The improved survival of offspring of birds with a greater twig collecting urge should be obvious.
Indeed, presenting twigs as presents to potential partners is often part of bird courtship, so the selection begins even before mating!
The progression of complexity was illustrated on my earlier link.
http://people.eku.edu/ritchisong/birdnests.html
Bird nests vary from a simple accumulation of materials on the ground to elaborate refuges in or on secluded & elevated substrates. Dial (2003) observed that nest construction and placement are correlated with other features such as flight ability. For example, basal avian taxa (ratites & many Galliformes) create a simple depression in the ground to harbor their incubating eggs, like those of nonavian dinosaurs. The progression of nest complexity moves from cryptic ground nests of some galliforms to simple elevated nests (e.g., Columbiformes, Cuculiformes, & Ciconiiformes). Taxa that construct elevated nests in a bush or tree or on a cliff or rock ledge tend to be better fliers than simple ground nesters.Young raised in elevated & cavity nests, including primary (Psittaciformes, Piciformes, & Coraciformes) and secondary (many Passeriformes) cavity nesters, have a robust forelimb flight apparatus, and less hindlimb mass, which is consistent with increasing flight capacity. As nest placement (e.g. invisibility, inaccessibility), construction (e.g. impregnability, camouflage), and attendance (e.g. feeding, protection, incubation) increase in complexity, a concomitant enhancement of flight styles is observed, including maneuverability and acceleration. The most complex nests are associated with some Passeriformes, particularly swallows, oropendolas, and weaver finches. Weaver finches (Ploceidae, Passeridae) and oropendolas (Icteridae) build intricately woven chambered, pendant nests hung from the resilient thin branches of bushes and trees in predator-rich environments. Perhaps the most predator-proof nests are those of swallows (Hirundinidae) and swifts (Apodidae) that often construct mud encasements secured to the most remote overhanging feature within their habitat (e.g. cliffs and human-made structures) (Dial 2003)
Predation, seems to be the selection pressure leading to the most complex defensive nests – eliminating those birds without the protective instincts, or ability to enhance these, by learning from experience.
I think you might have missed some of the point. A trigger need not be one single event or circumstance but can be multiple circumstances. It may be the trigger of seeing a snail isn’t activated unless the thrush is also hungry or the sun is overhead as well, that is, multiple triggers might be required.
On Kangaroo Island (Australia) there live yellow wattle birds. They attack snakes, possums, goannas (monitor lizards), other birds or any of the native fauna. They don’t bother with dogs, cats, humans or koalas. They attack anything native to the island. Koalas are not native there and neither are dingoes (wild dogs). It was also uninhabited by humans when Europeans arrived there. As an aside, the first European settlement in South Australia was on Kangaroo Island because the settlers didn’t want to disturb the aborigines on the mainland.
On the other hand, populations of the same wattle bird on the mainland (they don’t fly back and forth or migrate) attack dogs and swoop on humans.
The instinct is to attack things they recognise and that recognition must be a part of that otherwise those on the island would attack everything like those on the mainland.
THE DOUBLE TRIGGER – they only do it when they are nesting.
David,
The information that two twigs are better than one is not passed on genetically, or in fact any other way. No information whatsover is passed on. The birds have no conception of two twigs being better than one, or even of a complex nest being better than none.
What is passed on genetically is the hard wired propensity to ” peck and throw” twigs together, which by reproductive advantage and natural selection leads by tiny steps into more complex “peck and throw” behaviour and nests.
Steve disputes that any information is passed between the generations and maintains that what is passed on is a hard wired propensity. Apart from the fact that no one has found any such hard wires, as far as I know, this seems like a semantic distinction. The hard wired propensity, if that is what you want to call it, causes the bird to act as if information about how to build in nest had been passed on. The result is a Lamarckian one that the progeny has inherited a characteristic of pecking and throwing.
But this does not explain how a propensity results in highly a complex structure. I might have a propensity to put one brick on top of another (unfortunately I don’t) but this does not mean I will be able to build a house.
The nest building is instinctive, moulded by millions of years of selection and evolution.
If it was Lamarckian, young inexperienced weaver birds would not build roughly thrown together nests, and have to learn to improve them with experience.
The evidence of weaver-birds, is that the size of the entrance hole in the nest is evolved according to the size of the predator species.
Nests from a range of sizes, with a large enough hole to admit predators get predated so those instinctive traits don’t reproduce. Those which resist predators are more successful.
It is not surprising that Weaver birds (or anything else) get better at nest building the more they do it. As Gary Player said about golf “the more I practice the luckier I get”. So we can agree that there is a learning component.
This issue is how they have the innate knowledge to build their first one. Alan4 says it is instinctive, built up over thousands of generations (Richard Dawkins might say it was the Baldwin effect from watching others build them). But if it is instinctive does it arise from random mutations?
In a further contribution below Alan points out nests are designed according to the threats faced by that particular bird but that would be the case if Lamarckian evolution was taking place. Igloos are not built with a big door that allows polar bears to get in so if one generation is learning from previous ones they would be increasingly predator proofed.
The subject of nests, the effects of predators, and some of the symbiotic relationships make for complex ecosystems with lots of interactions.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4242575/
You really do not understand the logic of evolutionary theory at all if you cannot see that instincts can evolve with absolutely no neccessity for learned behaviour itself to be passed on in any way , aside from the next generation learning anew. ( the hard wired aptitude to perform the learned behaviour being the thing which evolves and increases in a bog standard evolutionary manner)
I think this covers it:-
Extended Phenotype – http://www.sciencedaily.com/releases/2009/01/090119081333.htm
The EP states that the genes of an organism can be expressed beyond its immediate biological boundary, extending for example to birds’ nests, or the behaviour of hosts infected by parasites. The key point is that the EP embraces constructions, such as the dams built by beavers, whose quality is correlated with variations, or alleles, in certain of the organism’s genes, so that natural selection can act upon them. A new allele that leads to better dam constructions will in turn benefit the beaver expressing it. Similarly a parasite allele that causes the behaviour of its host to change in a way that increases the parasite’s chance of survival will be favoured.
The distinction between behaviour based on learned information and instinctive behaviour is not semantic, it is real.
The evolution of peck and throw, or anything else, is not Lamarckian as the pecking and throwing does not alter genes in any way whatsoever, instead already existent peck and throw genes are selected and their frequency, stength and complexity evolve as long as they give reproductive advantage.
If you can comprehend how an eye evolves by tiny increments from a light sensitive patch of cells into the complex eyes of today I am baffled that you cannot see how, in exactly the same way, nests ( and the instinctive behaviour to build them) etc have evolved from the simple to the complex.
David,
I feel it is you who appear grammatically confused about the difference semantic applications of the term ” information”
We say genes pass “information” and also we learn”information” but the genetic “information” is totally different in form and content to learnt ” information” .
To think the 2 “informations” are equivalent is a error of grammar, of the correct application of semantics.
First a disclaimer for being completely ignorant of the science but interested at a layman’s level I wonder about the effects of EP as translates to brain structure and size over time. I ran across some information that only served to confuse me but I will include it for clarity. Hopefully (others) can give me an idea of the principles related to both concepts that might might paint a picture of harmony between the two. The following is taken from the article/paper from the first paragraph. Should my inquiry fall into the complete nonsense and blather category please disregard and ignore my ramblings.
“Human brain development is created through continuing complex interactions of genetic and environmental influences. The challenge of linking specific genetic or environmental risk factors to typical or atypical behaviors has led to interest in using brain structural features as an intermediate phenotype.”
Additionally in the third paragraph of the article/paper.
“Genes do not code for behaviors, but for the building blocks of the cells whose interactions eventually give rise to those behaviors; conversely, the translation of environmental input into persistent behavioral changes occurs through alterations in brain systems and even structures.”
My interest as stated, is whether external factors outside of ourselves related to our continuing evolution, as absorbed through higher cognizance, might ultimately affect our brain structure beyond the internal reorganization possibility. I had heard somewhere on Richards site that our brains are actually shrinking and this peeked my interest as to the evolution principles and factors being discussed in this thread.
Here is a link to the information I pulled the quotes from.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2892674/
Please ignore my inquiry. It’s total rubbish and doesn’t need any consideration or an answer. Sorry about that. Need to keep my musings of the absurd to myself.
A bird must learn to fly, and a cat must learn to hunt. But I think we can say that flying and hunting are instincts in these species.
Is this speculation/informed guess or the result of some experiments?
Bite into a slice of lemon; your eyes squint shut, and you make the “lemon face.” Get hit with a shot of capsaicin; your eyes water, you cough, drool, etc. Give catnip to a cat; the cat begins to roll around wildly, or act frantic or “high.” (And let us not forget the extreme case of Dicrocoelium dendriticum). These are physical and physiological responses to natural chemical stimuli. Maybe snail-bashing on an “anvil” is a physiological reaction by thrushes to a chemical present in certain snails.
Imagine six thrushes tasting six snails. Then suppose they have the following reactions: a) fly to the highest branch of the nearest tree; b) hop around in a circle; c) crouch down onto the ground in a “nesting” posture; d) flail beak against the nearest rock; e) begin to sing loudly; f) fly full-speed into the trunk of a tree.
Certainly these reactions would be the result of genetic coding in the thrushes’ brains, triggered by the chemical stimulus of the snail. Some of the behaviors (f, for example) would be fatal for the bird, preventing the allele for that behavior from being passed on to future generations. Most of the behaviors (a, b, c, e) might have no net effect upon the bird’s survival or its propensity to pass on those alleles. But SOME behaviors (d, flail beak against a rock) would have the effect of smashing the snail held in that beak, delivering a nutritious meal to the bird, which then gives that bird a better chance of surviving long enough to pass on its genes, including the allele that codes for reaction (d) upon tasting a snail. Reactions a, b, c, and e might still exist in the gene pool, but over evolutionary time they would fade away, outcompeted by reaction d, unless circumstances changed which gave one of those reactions a survival advantage comparable to reaction d.
This scenario gives us a completely “instinctive” behavior with a completely genetic cause—no learning is required. The other reactions, while we might notice them, seem insignificant to us, because we don’t see them resulting in any survival advantage. Therefore we give special notice to the one reaction that does, and label it as “learned behavior.” We thereby overlay our interpretation onto a perfectly natural set of circumstances, and spend hours debating it on the RDF website. Plausible?
Steve D,
Nope. We do not pay attention just to d) snail bashing and label it “learned”
The example involved specifically learnt behaviour, learning to take a snail to a specific anvil rock and bash it against that rock, it being experimentally proven that birds copy(learn) this behaviour from others. Which is not the same as in your example, where there is no learned behaviour but just instinctive thrashing
Is this the last of the videos? Or will there be more?
I like these videos. They are informative and they give the feeling that Richard is present on his own website. I miss the days when he would pop in with the occasional comment here and there. A video is not the same effect but better than nothing.
Dr. Dawkins traces how “clever”individual thrushes discovered how to peck off the shells covering snails setting in motion a long evolutionary process which favored the survival of thrushes who acquired the genetic-based ability to learn the procedure faster and faster until the behavior seemed instinctive.
He also mentioned, in an aside, that blackbirds were more than happy to feast on “naked” snails, but had no idea how to remove the shell. I wonder why blackbirds did not follow in the evolutionary footsteps of thrushes? The example Dr. Dawkins describes involves learning a simple task of operant conditioning. It would seem likely, almost inevitable that most bird species as a matter of course – like Skinner’s pigeons- would learn a simple task acquired by incessant pecking punctuated with regular rewards.
I agree. An instinct such as a bird intuitively beginning a migratory flight of several thousand miles the same time each year, . . . how is that explained by the Baldwin Effect? Is leaving at this exact same time every year somehow enhancing the fitness of the bird/species? What about the nest building capability that seemingly comes natural to a bird? Why, with random fragments of twigs and other material are most nests constructed so perfectly round in the middle? I doubt birds study the geometrical perfection of other birds’ nests. And there are many examples of behavior in species (i.e., bees homing instincts, the life-long hierarchy of ants, etc., etc.) that are not easily explained (actually not explained at all) by the Baldwin Effect. While the Lamarkian Theory of Acquired Characteristics obviously cannot account for a three-legged dog siring a three-legged pup, the innate characteristics of so many behavioral traits seem much more attuned to a precariously preset mechanism devoid of any learning curve.
In this example the birds do not learn how to peck the shells of snails, so it is not a question of incessent pecking punctuated with regular rewards.
The learnt behaviour is to smash the snails against a anvil stone., not something which would occur by random pecking.
Many thanks, Steve, for the correction. I should have paid closer attention.
Operant conditioning may be a bit of a stretch given the cognitive skill for learning the behavior. But a related question persists. How did did the thrush species learn the adaptive behavior of smashing snail shells against rocks to feed on the contents while blackbirds living in the same environment failed to do so? Dr. Dawkins says that other thrushes learned by watching a primordial “clever” ancestor discover then demonstrate the technique. Why didn’t blackbirds also observe the demonstration or otherwise procreate a clever ancestor who chanced upon the same nourishing practice coincident with the “thrush discovery?”
The problem may lie partially in the necessity of using narrative language to explain the Baldwin affect in evolution forever lost in the veils of the past to scientific observation and experiment. When Dawkins speaks of an ancestral thrush he seems to designate a single protagonist acting out an evolutionay story that goes, “once upon a time, there was a very clever little thrush who came across a snail crawling on the grass…” We reflect that there could have been no such lone bird to set the practice in motion. He or she would have had to teach a critical number of co-specimens to copy the behavior in too short a lifespan. Unless there had been many co-innovaters of the snail smashing technique, the behavior would have been extinguished within a generation.
Perhaps a more likely scenario is that ancestral thrushes slowly evolved a special cognitive ability to learn the behavior long before the species started to demonstrate it in a snail-rich food environment. Similarly human language, may have required hundreds of thousands of years of genetic-neurological development before the ability to acquire and use language slowly began to manifest itself in human behavior. Perhaps, there is no hard line between learned and instinctive (genetic) behavior in “higher” animals but rather a species-specific genetic predisposition to learn those behaviors. What we see in the manifestation of those behaviors is a feed-back loop, a seamless interaction between learning and the species-specific genetic ability for learning them.
Another classic example of a species using rocks to obtain food is displayed by sea otters swimming in shallow waters off my California coast. The animal dives to the bottom and picks up a rock that it uses for feeding on shellfish catches. Always feeding on the surface. the otter floats on its back and places the rock on its furry chest. Holding the shellfish with paws and claws the animal smashes the shell open and eats the exposed meat. (Very cute brief footage is available on Youtube).
Nest building that also involves cognitive ability has become “instinctive” in birds. Curiously the much more complex skills displayed by beavers for building dams across shallow streams has ambiguous causes. Only beavers living near running water build dams, the apparently “engineered” lodges, for habitation and reproduction. ( Ironically human observers have stigmatized beavers overall as pretty “stupid” animals.) In other environments, non-dam-building beavers burrow in mud or soft ground. One experiment took beaver offspring, “kits” never exposed to water and placed them in a dry setting with dam building materials. Nothing happened… until the experiment exposed the kits to the recorded sound of running water. Upon hearing the sound, the kits immediately set out to build a dam without ever seeing a stream.
The relationship between learned and instinctive, genetic-triggered behaviors, is highly complex. For reasons we can probably never trace, species probably evolved special cognitive abilities through incipient dispositions to behave in specialized adaptive ways. At some point those cognitive abilities specific to the species emerged in overt practices across the species in relation to environments they evolved in. I would suspect that “clever individuals,” an anthropomorphic attribution, played an elusive role in the dissemination of the behavior.
Should also take into account that species take especial note of what their own species are doing, i.e blackbirds are smart enough to learn this behaviour but more likely or not pay little attention to what thrushes are doing so never pick up on the behaviour.. Thrushes on the other hand instinctively pay more attention to what other thrushes are doing and are more likely to learn from other thrushes than a blackbird is.
This does not mean that birds cannot learn from other birds, the milk top opening of Tits was eventually learnt by some sparrows as well.
species take especial note of what their own species are doing, i.e blackbirds are smart enough to learn this behaviour but more likely or not pay little attention to what thrushes are doing so never pick up on the behaviour.. Thrushes on the other hand instinctively pay more attention to what other thrushes are doing and are more likely to learn from other thrushes than a blackbird is.
Maybe so… I’ve considered the point. What their own species is doing embodies the phenomenon of species-specific behavior. We are both saying much the same thing under different descriptions. “Not paying attention” also elides the question of why birds which are extremely vigilant in their surroundings would not ‘pay attention” to another bird of a different species getting yummy food. If the blackbird is “as smart as” the thrush we are also eliding the question of why blackbirds did not independently discover the behavior even in the absence of thrushes in their environment. (The absence of the snail eating technique neither promoted nor discouraged the survival of blackbirds.) I believe the problem derives form our linguistic minds anthropomorphizing animal behavior by talking about evolution “as if” animals thought with human language. Once again, perhaps using the beaver dam-building example, we are observing a more holistic process of genetically disposed learning manifested in behavior specific to the species and therefore absent in related but different species. The evolution of each species produces a unique genome that generates unique features and capacities, including unique capacities/abilities to learn specific practices.
I might be missing something but do not see that the video has anything to do with instincts. The learned behaviour does NOT become an instinct,. In this example no thrush instinctively bashes snails, it still has to learn it anew from other thrushes.
It is the aptitude of learning which is evolving, it is not learned behaviour evolving into an instinct. As he says ( to paraphrase) , “so that snail bashing SEEMS like an instinct.”
Dr. Dawkins certainly needs to clarify this point. Perhaps I can distill my thinking by moving back to the title topic. Without putting words in his mouth, I believe Dawkins is trying to describe how learned behavior might evolve into instinctive behavior. ( TITLE: “How does learned behavior evolve into inherited instinct?”) Effectively he is describing behavior as an interactive manifestation of an organism’s genetic capacities, including mutations, with selective pressures of environments resulting over time in adaptive evolution as we observe them on a spectrum. We observe birds building nests “instinctively” because the behavior fits the description or definition of “instinctive” -automatic, involuntary, unlearned- behavior. What we cannot observe are distant ancestors perhaps “learning” nest building copying what their “clever” peers were doing to protect their young and avoid predators. Perhaps we insist too vehemently on hardline distinctions when we think we are observing radically “different” phenomena while we are actually observing, in the larger evolutionary story, a dynamic process on a spectrum. The process, when viewed in the current stage of snail-shell bashing techniques practiced by most thrushes can be classified as “learned” behavior that must be “copied” by offspring. On the evolutionary spectrum thousands of years from now, our successors could observe the same behavior as “instinctive” -automatic, involuntary, unlearned , genetic. The behavior displayed by beavers in dam building is far more difficult to define. It fits the definition of instinctive as shown by the experiment cited above, but the complexity of the behavior suggests a far more ambiguous process of evolution. It is difficult to imagine this complex “engineering” never passing through a learning phase in the beaver ancestral past.
I suspect our language has not caught up with the magnitude of the diverse mechanisms of evolution that lie beyond current discoveries.
Our language is quite precise, and if not precise enough we can quite easily make it more precise. This is the power and beauty of language.
The distinction between instinct and learned behaviour is linguistically and semantically clear, so the problem is not grammatical, any confusion lies in the fact that in reality behaviour usually involves a mixture of both.
Holy Carp! My wife and I were just talking about this yesterday. Putting those organisms aside that actively fight their genes at every opportunity (sentient?), it’s slow, random evolution. Now I have to watch RD answer to see if I’m wrong.
Even where the language accurately describes current knowledge and discoveries, most people outside the specialist professionals, do not understand the concepts, or the terminology which describes them.
Our language is quite precise, and if not precise enough we can quite easily make it more precise. This is the power and beauty of language.
The distinction between instinct and learned behaviour is linguistically and semantically clear, so the problem is not grammatical, any confusion lies in the fact that in reality behaviour usually involves a mixture of both.
I do not know what “quite precise or precise enough” means to different people situated in different times and places. Your proposition that, “if not precise enough we can quite easily make it more precise” confirms the point of view that language is a tool which will always be subject to modification as we gather more evidence which refines our understanding of how things work. I stand by my judgement that in all sciences but particularly in evolutionary science where so much remains to be clarified: ” I suspect our language has not caught up with the magnitude of the diverse mechanisms of evolution that lie beyond current discoveries.”
Even where the language accurately describes current knowledge and discoveries, most people outside the specialist professionals, do not understand the concepts, or the terminology which describes them.
Claudius Ptolemy used language to accurately to describe a geocentric model of the universe consistent with current knowledge and discoveries and sophisticated mathematics. When Copernicus came along he was “confused” because he could not understand Ptolemy’s concepts and terminology.
Even where the language accurately describes current knowledge and discoveries, most people outside the specialist professionals, do not understand the concepts, or the terminology which describes them.
You make a good point about the long-standing history of specialist language – particularly in science and mathematics.
Not really! It is not necessary to understand all the terminology of flawed concepts to recognise basic errors! (Similarly evolutionary biologists don’t need to understand the invented terminology of “barminology”!)
Copernicus, like many scientists refuting earlier misconceptions, understood Ptolemy well enough to recognise the flaws, and correct the basic errors, even though his new model was still inaccurate!
It was Ptolemy’s basic premises which were wrong, not his mathematics.
http://en.wikipedia.org/wiki/Geocentric_model
Johannes Kepler analysed Tycho Brahe’s famously accurate observations and afterwards constructed his three laws in 1609 and 1619, based on a heliocentric view where the planets move in elliptical paths. Using these laws, he was the first astronomer to successfully predict a transit of Venus (for the year 1631). The change from circular orbits to elliptical planetary paths dramatically improved the accuracy of celestial observations and predictions. Because the heliocentric model by Copernicus was no more accurate than Ptolemy’s system, new observations were needed to persuade those who still held on to the geocentric model. However, Kepler’s laws based on Brahe’s data became a problem which geocentrists could not easily overcome.
It was team-work which refuted geocentricism. (gravity!!!)
Johannes Kepler analysed Tycho Brahe’s famously accurate observations and afterwards constructed his three laws in 1609 and 1619, based on a heliocentric view where the planets move in elliptical paths.
It was Ptolemy’s basic premises which were wrong, not his mathematics.
There’s no disagreement on this judgment (and thanks for the detailed info on Copernican genius). Steve seems to make a similar point substituting “grammar and semantics” for “mathematics” in your example: The distinction between instinct and learned behaviour is linguistically and semantically clear, so the problem is not grammatical, any confusion lies in the fact that in reality behaviour usually involves a mixture of both.
Perhaps we can concur that as new evidence emerges from ongoing research into evolutionary processes we will more accurately understand the dynamic relationship between learned and instinctive behavior along some holistic spectrum. (Or not.)
To illustrate by way of example: About 70 to 80 years ago, scientists believed that the “universal’ force of gravity would first slow then reverse the expansion of the universe. At some point, the galaxies would all pull each other inward toward a Big Crunch, reversing of the Big Bang expansion back into a singularity of infinitely dense matter. The discovery that the universe was not only expanding, but expanding at an accelerating rate lead to the theory of dark matter and dark energy working against gravity from the great and exponentially increasing expanse of the “empty” vacuum of space. Within this vacuum which constitutes about 96% of the cosmos, a repellant force, caused by virtual particles popping in and out of existence, pushed against gravity, propelling the galaxies apart into an infinite cosmic darkness one from the other. Scientists had to “invent” a new language (dark matter, dark energy, etc.) to describe “logical” phenomena almost in reverse terminology. Today no responsible scientist will tell you he understands this radical process completely. Our language must catch up to the awesome new realities we’re just beginning to intimate. I recommend a documentary called “From Here to Infinity” that I caught on Netflix that traces the discovery. (Easy to understand.)
“Everything that can be thought at all can be thought clearly. Everything that can be put into words can be put clearly.”
― Ludwig Wittgenstein, Tractatus Logico Philosophicus
I enjoyed this video and found it very interesting. But I still feel a bit confused abut this. Some animals perform actions that are completely irrelevant and unnecessary, but do so apparently because it is an inherited instinct. We notice this particularly in domestic dogs. For example, spinning around and around before laying down, or an urge to bury things. Apparently random acts with no benefit to their life or current situation, but acts which are common to all dogs and were of benefit to their wild ancestors. So in this sense, how does learned behaviour evolve into inherited instinct, when the described theory of learning quickly does not apply?
How is the “information” passed down for those who as you say no longer need to learn (to open milk bottles example). it is not like you can be born able to do complex tasks without learning UNLESS there is a transmission of information to the offspring. This sounds like you still haven’t resolved the information problem. This doesn’t seem at all like science, just a big guess.
Comments are closed.