Talk:Collective animal behavior

Wiki Education Foundation-supported course assignment

This article was the subject of a Wiki Education Foundation-supported course assignment, between 9 September 2021 and 3 December 2021. Further details are available on the course page. Student editor(s): Wormygal. Peer reviewers: TV2424, Alwayh.

Above undated message substituted from Template:Dashboard.wikiedu.org assignment by PrimeBOT (talk) 19:17, 17 January 2022 (UTC)

Wiki Education Foundation-supported course assignment

This article is or was the subject of a Wiki Education Foundation-supported course assignment. Further details are available on the course page. Student editor(s): Chiaki167.

Above undated message substituted from Template:Dashboard.wikiedu.org assignment by PrimeBOT (talk) 18:01, 16 January 2022 (UTC)

Collective animal behavior and brain behavior

please see this page

http://en.wikipedia.org/wiki/Collective_animal_behavior_and_brain_behavior

it is yet not very articulated article — Preceding unsigned comment added by Scienficreal (talk • contribs) 21:50, 20 March 2011 (UTC) --Scienficreal (talk) 22:02, 20 March 2011 (UTC)

Should this article have a criticism section (or should it have no criticism section)?

Let me start with a very careful apology, as my previous edit and talk contribution were deleted. Just the talk entry above (by someone whom I do not know, and someone with whom I am completely unconnected) leads to a deletion log with harsh words, some of them nearly obscene, I won't repeat them here. Similarly, in events that are unrelated except conceptually, one of the originators of the subject of this article, the author of some of the reference materials cited in the article, had met with criticism in the past which was deemed to have been bullying and illegitimate.

The editor who deleted my attempt to create a stub of a 'criticism section' makes intelligent edits and comments elsewhere in Wikipedia, tending to a suggestion that insights from nature might extend thought into domains where ideas in computer science, or even science generally, are considered to be, or proven to be, paradoxical and inadequate.

But I must insist on the strength of my own hope that my request here on the talk page for an editor to look into sourcing a criticism section should not be deleted, for the same reason that justified the deletion of my incompetent attempt edit to the article proper, that it genuinely did constitute incompetence; and should be left to someone with familiarity with the subject.

I have no previous involvement and am completely neutral; and yet I also have no intellectual background in the subject; my first and only attempt to find criticism led only, via Google, to a vapid debate about university administration etc etc; though also searching here led me to what seems to me to be a fairly deep consensus among editors, unspoken, reflecting the depth of collective animal behaviour beyond choosing a leader to follow, according to any mechanical voting scheme, or even using any leader at all.Createangelos (talk) 23:37, 27 June 2015 (UTC)

I must admit, I'd utterly baffled by this. What supposed controversy or criticism is there? Nobody doubts that animals engage in collective behavior, and all the rest is just various mechanics, evolutionary forces, and benefits. How is there controversy? You've yet to really articulate any actual content or concepts for this supposed section, beyond linking to an article that nobody here can read. HCA (talk) 15:49, 28 June 2015 (UTC)

As the editor who deleted the criticism section, I have been slow responding because I cannot work out what Createangelos is talking about (both here and on their and my respective talk pages). Has it something to do with this exchange? Or has it something to do with the many reverts of algorithms whose status as "swarm intelligence" has been disputed (these seem related more to Indian mathematicians than Swedish ones). There is some legitimate material critical of the proliferation of swarm algorithms, for example, this paper, and it might be appropriate to add something about this. But memos concerning the internal politics within a university over the appointment of a staff member does not seem like useful material in this context. --Epipelagic (talk) 18:55, 28 June 2015 (UTC)

Hi, thx for your thoughtful replies, and HCA's questions about what to think about. About the first one, how can the entire article have a criticism section, one possible answer is that maybe it shouldn't (I'm not dogmatically saying that it should). The wikipedia article Ocean starts, "An ocean (from Ancient Greek Ὠκεανός, transc. Okeanós, the sea of classical antiquity) is a body of saline water that composes much of a planet's hydrosphere." That article has no criticism section. Whereas if we choose a controversial topic like Intelligent design we find that the article begins "Intelligent design (ID) is the pseudoscientific view that 'certain features of the universe and of living things are best explained by an intelligent cause, not an undirected process such as natural selection,' " and has four criticism sections. I'm actually confused a little about something here: if the Intelligent design article had started out saying "Intelligent design refers to the planning or designing of living or non-living entities or algorithms by the action of a conscious intelligence," then we'd have to conclude that all four criticism sections are misplaced. It would be nonsense to criticize doing things intelligently in general.

Here, the first sentence is general, like the first sentence of the Oceans article is, then the second sentence says "Facets of this topic include the costs and benefits...." and so what is being discussed in sentence 2 is an academic topic *about* collective animal behaviour, rather than the way animals behave collectively in nature independently of how they are discussed or analyzed by people. The fact that the article title is verbatim equal to the title of one of the reference sources might possibly suggest that the article really is explaining one particular school of thought. Someone who already has some understanding of the deeper issues would be better placed than me to try to balance the whole picture. Also, not to be annoying and bring up endless irrelevant objections, but doesn't a concept of costs or benefits rely on an economic premise somehow? Regarding evolutionary connections, is evolution supposed to be like a scale of success where there are rich and poor organisms, which either are successful at surviving or not; or is it like a phylogenetic tree, or even some other thing involving deep interdependency even among species? Does 'collective' refer to one species only, like sort of an interest group, like a team of similar organisms all wearing the same color jerseys? I'm not trying to be intentionally stupid here. But does the second sentence of the article about costs and benefits mean that collective behaviour is a way of understanding organisms like how a football coach plans a successful team strategy (you pass the ball to joe and when he sees the blocker he passes it back again) against the background of a scoreboard someone had set up ahead of time, measuring costs and benefits numerically? Is it supposed to be obvious to any observer which species or flocks or schools in nature are the successful ones (like humans obviously are)?

I haven't finished reading Epipelagic's reply, because it is technical and requires background reading. I hope to reply before a week has passed.Createangelos (talk) 21:34, 28 June 2015 (UTC)

Some of this will clear up as you read more, but generally speaking, "criticism" sections are for concepts where there's a clear and substantial controversy. ID has such a section because it has zero evidence for it and a mountain against it. Oceans doesn't have a criticism section because nobody seriously doubts the existence of oceans, or will argue that all oceans are bad. Sliding_filament_theory is more illustrative for our purposes - it is widely/totally accepted, and while there are modifications of the theory suggested in the scientific community, these are minor, of largely technical interest, and do not fundamentally contradict it (but rather simply add 'extra' mechanisms on top of it). Thus, there is no criticism section.

For collective animal behavior, there's no doubt that it's real, so that's not an issue. I'm not really aware of substantial divisions or schools of thought within the field - most sensible people hold that Species A may perform collective behavior for a different reason than Species B (due to differing ecologies, life histories, etc.). As I understand it, the current state of the field is largely concerned with description, understanding the mechanisms (how to monitor and maintain distance from other animals), and evolutionary causes of this behavior in different species.

The evolutionary causes are, ultimately, at the level of the individual, not about comparisons between species. Sardines make large schools, so presumably there is some evolutionary reason for this, some advantage over solitary life (such as being able to more effectively spot predators via many eyes, being able to confuse and evade predators better, etc.). Starlings form large flocks, which also presumably provide some advantage relative to solitary life, but it may be a different advantage from the sardines. Neither is "better", they just represent independent convergences on the same solution to various problems. Flocks / schools / herds may contain related individuals (in which case helping your kin helps you, due to shared genes), or may not (in which case there has to be some advantage for the individual, such as early warning about predators). HCA (talk) 14:53, 29 June 2015 (UTC)

OK, I haven't had any thoughts about whether the article really does or doesn't after all need a criticism section, versus, for instance, putting in other references which would be relevant if the article really is about the subject of collective animal behaviour per se (which could not have been criticized), rather than a particular school of thought about it (which has, but in an indirect and incorrect ad hominem manner).

However, here is a sort of high-school 'book report' summarizing the article and the other references you've both mentioend. To start with the Sliding Filament Theory article, I do understand that it represents a model of good Wikipedia practise (as well as good scientific practise), even without any criticism section, maybe for the same reason that the Ocean article does.

Re the article at hand, the final section raises some questions, "how are these decisions made? " and suggests only two possibilities (leader and consensus). In the previous section it says that Hilderbrandt's Starling model is "convincing" but not to whom it is convincing. The actual Hildenbrandt reference does not say that their model is convincing. They say that Starling flock behaviour is "still a mystery" but they do imply that it could be un-mystified by a model (i.e. the Wikipedia article is wrong to imply that the article says that individual human ingelligence of Hildebrandt has already encompassed collective Starling ingelligence, but not wrong to imply that the references say that it soon will -- here is where I do not know if there are any references making any opposing point).

The profound survey on swarm intelligence by Mahant et al is also consistent with the notion that collective animal intelligence is something cognitively trivial from an individual human perpsective, e.g. "Another reason PSO is attractive is that there are few parameters [that the programmer needs] to adjust." As applications the Wikipedia article faithfully represents things in the Mahant article such as designs of fleets of military drones, presumablyh farms and highway systems, and the idea seems to be that there is a structure to collective animal behaviour (such as a voting scheme) and when this is understood then human inventions will be somehow on a par with natural phenomena.

This is consistent with the notion in the Wikipedia article that collective behaviours have what the Wikipedia article calls a 'function.' I'm not clear in my mind about what the various references say about the notion of a 'function,' it might be based on a notion of a cost/benefit interpretation of evolution in explaining where the sort of trivial inter-linkings like number of nearest neighbours had 'originated,' and it might be based on the notion of applying the model in subsequent human actions in nature where presumably there is a known goal which might be optimized.

The article Metaheuristics – the Metaphor Exposed by Kenneth Sorensen is all about optimization, while making no connection between a notion of optimization and either a cost/benefit interpretation of evolution or an analysis of what it is in human experience that is meant to be optimized. The article appears competent and worthwhile somehow, though it seems to hold quantum mechanics as an unassailable model of good practise, and modifying our understanding of it by analogies with nature is criticized.

The exactly opposite position which Sorensen takes, is taken by the article, Using river formation dynamics to design heuristic algorithms, by Rabanal. He says "the method is inspired by the way rivers are created". Unlike Wikipedia, Hilderbrandt, or Mahant, this notion of inspiration does not limit the amount of structure which is conveyed to the computer programmer by experiencing nature, to a finite amount of information (such as two numbers for each individual in a species). They are in different domains of computer science. Sorensen is working in optimization, and Rabanal working in algorithms (P=NP).

I have not included any artificial intelligence links, for instance about neural networks. In that subject there are indeed easy calculations with a small amount of structure analagous to the calculations made in the Wikipedia article, but there is also a notion that there is a large or possibly infinite amount of underlying information that has to be acquired by 'tuning' the neural network, essentially by letting it observe nature directly. That is, it is not assumed that the human programmer could directly set the multitude of weights in a neural network by performing experiments in nature and writing code. In that sense any of the relevant neural network references would be similar to the Rabanal reference, though Rabanal advocates that the human observes nature (such as river formation) and then uses the `inspiration' to write code. That is, in a neural network a sort of trivial structure as the one ascribed to collective animal behaviour overlays a second, deeper, and nearly continuous structure which is understood to arrive by direct contact with nature, without a human being having the ability to describe it in words or computer code. Also the Rabanal article attributes unplumbable depth of nature as even occurring in river formation. Unlike the Wikipedia article, although the depth of even natural phenomena involving particles of sand and water is considered unplumbable, this depth is not applied to notions of how hummans might design military drones, roads, etc, but it is applied to how humans might understand knowledge of abstract algorithms.

I have not thought about any of these references, just tried to summarize what they say. I'd like to return to what the Sorensen article says, for a minute: it is about 'optimization' which in that article is an undefined concept, and it accepts quantum mechanics as a fixed authoritative model, presented as perhaps an example of how nature should apply to human understanding (like how HCA presented Sliding Filament Theory as a model of good practise here perhaps). It is clear that Sorenen versus Rabanal present two diametrically opposing views. Both are focussing on algorithms in computer science, neither is getting involved in the big questions evolution, applications, human behaviour, man versus nature etc etc. Sorensen is directly critical of the approach that people like Rabanal take, while Rabanal does not criticize anyone. Sorensen takes what might be considered a fundamentalist interpretation of quantum mechanics, and is critical of Rabanal's type of science as though Rabanal is trying to be a chef in a kitchen, invoking inexplicable mysteries of good taste, perhaps, when, in Sorensen's view, these issues can be nailed down numerically once and for all.Createangelos (talk) 09:42, 7 July 2015 (UTC)

(...actually it was below the belt for me to call Sorensen's approach 'fundamentalist.' He criticizes analogies with nature which change our intuitive impression of scientific theories while not actually making any change in the theories themselves, to be a bit more precise and fair.)Createangelos (talk) 10:34, 7 July 2015 (UTC)

I think the problem only arises once you move beyond animals and into computer science, which is not within the scope of this article. Which author is right about how to design algorithms is irrelevant - the article is about collective behavior in biological animals, and how that arises and evolves. How and why animals arrive at a particular state is often very, very different from human design, or even from what's optimal. Your references might be very relevant to the articles Swarm_intelligence or Collective_behavior, but this article has a narrower scope, restricted to non-human animals in nature. HCA (talk) 14:20, 7 July 2015 (UTC)

I hope you realize how passive (lazy?) I'm being.... you're right that the article (after word 20) says "Facets of this topic include the costs and benefits of group membership, the transfer of information across the group, the group decision-making process, and group locomotion and synchronization. Studying the principles of collective animal behavior has relevance to human engineering problems."

So it is talking about a `topic' rather than about collective animal behaviour per se. And right away focuses on two things you flagged up: cost/benefit and engineering problems. Now you say "How and why animals arrive at a particular state is often very, very different from human design, or even from what's optimal." So can I use you as an authoritative reference for that, if the debates about computer science applications are too narrow? Sorry, I'm not exactly feeling like I've been lazy but admit that my attempt to improve the article is rapidly approaching failure.Createangelos (talk) 17:37, 10 July 2015 (UTC)

When it says "cost/benefit", it's talking in an evolutionary senses as well as mechanical, and the two could be the same (reduced cost of flight in flocking birds, for example), or utterly unrelated (sacrificing one's own fitness for a greater fitness benefit for kin, known as kin selection). And yes, it is relevant to computer methods, and there may be some exchange of ideas, but this article, as given by the title, is about "collective animal behaviour", so unless you can show how the disputes you cite are directly relevant to our understanding of how and why natural animals engage in collective behavior, it's irrelevant.

Consider an analogy - animal locomotion and robotics. There is significant intellectual interchange between these fields, and ideas of one can influence the other, but there are also areas in each that have no relevance to the other - for example, muscle physiology is of little relevance to a machine powered by electric motors and vice versa, so the controversies in either area aren't relevant to the other.

In the case of the authors you cite above, I fail to see how this controversy has anything to do with collective animal behavior - it's all coding. Who's right or wrong may affect coding, but won't change how we study animal collective behavior in the slightest. HCA (talk) 15:45, 11 July 2015 (UTC)

This is going to be a wandering reply because I really am on the verge of giving up on this. Let me start with the abstract of one of the references for the 'Evolution' Wikipedia article, Fellenstein (1978) "When a polygenic character is exposed to natural selection in which the curve giving fitness as a function of phenotype is a mixture of two Gaussian (normal) curves, the population may respond either by evolving to a specialized phenotype near one of the two optimum phenotypes, or by evolving to a generalized phenotype between them. Using approximate multivariate normal distribution methods, it is demonstrated that the condition for selection to result in a specialized phenotype is that the curve of fitness as a function of breeding value be bimodal. This implies that a specialized phenotype is more likely to result the higher is the heritability of the character. Numerical iterations of four-locus models and algebraic analysis of a symmetric two-locus model generally support the conclusions of the normal approximation."

This is the reference for one of what are said in the article to be three types of natural selection: directional, disruptive, or stabilizing. The articles are old, and predate some of the debate that was popular later. The articles referred to there, and also the Wikipedia article, model evolution as a response to a 'fitness' function which is presumed to be pre-existing.

When you contrasted one case when notions of cost/benefit in the evolutionary or mechanical sense are the same "reduced cost of flight in flocking birds, for example" or different "sacrificing one's own fitness for a greater fitness benefit for kin, known as kin selection" in both cases the -- essentially -- justification for the action of evolution is that it ends up being aligned with such a fitness function, as you say "greater fitness benefit for kin."

I am grateful to Epipelagic for showing me the active and intelligent debate within the computer science community, to me and allowing us to consider whether any of the main disagreements could be interpreted as criticism of the one approach to 'collective animal behaviour' in this article.

On questions of applying observations of collective animal behaviour for the sake of acquiring engineering wisdom, the reference Hildebrandt in this 'Collective Animal Behaviour' article considers only the question of whether to go with 'leader' or whether to go with 'consensus' and, although misquoted by the article, says nearly as much, that the entirety of Starling behaviour is 'still' a mystery.

The comp sci debate didn't even consider such shallow notions of 'leader' or 'consensus.' It dealt with questions about whether meaningful observations of nature are of a type where a human being even has the capability at all to make make observations of nature and faithfully record them in a computer language or algorithm of any kind. It would not be a far stretch to find articles criticising the notion of the acquisition of scientific knowledge on an analagous basis. But for our purposes here we could restict ourselves to engineering since the topic of the article seems to be in large part engineering applications of observing and codifying collective animal behaviour, even while it might be simpler to bring in criticism from the wider domain of scientific method.

Debates joined by Pinker, Dawkins, etc perhaps do cover some of the same ground in areas of human cognition. Here one becomes distracted by issues of morality, consciousness, intention, and questions of logic which had been outside the computer science domain, or even the domain of philosophy of science.

In evolutionary theory, debates we hear all the time are just between people who tend to be non-academically oriented, and whose ideas are rejected out-of-hand (as you said, "ID has such a section because it has zero evidence for it and a mountain against it"), versus people, who found themselves needing to explain what they mean by 'evidence' or 'science.'

Contrasting a mechanistic cost/benefit function with an evolutionary one, giving examples where they coincide or are different, as you did a few paragraphs above, implicitly raises the issue whether the premise of a single underlying real-valued 'fitness function' for every purpose and interaction even for a single species in a natural setting has ever even been explicitly articulated anywhere. In Descent of Man, as far as I can remember, issues were about whether speciation exists, why it is irreversible, and there were notions of desireability, but these were culture specific. It seemed more like a notion that nature knows how to use the carrot and stick, not that nature is at the behest of some huge carrot and stick, and we thinkers duty bound to break the code so we can get the carrot sooner, or first.Createangelos (talk) 22:23, 11 July 2015 (UTC)

What does any of that have to do with this page? Seriously, stop rambling and give me a suggested edit that directly pertains to this page. As in "I want to add the following 5 sentences", not vague ruminations. Nothing you've said yet actually pertains to the topic on the page in any but the vaguest and most indirect ways. HCA (talk) 11:23, 12 July 2015 (UTC)

The simplest thing would be to take it to be an article about collective animal behavour per se. The references then need to be broadened, perhaps without increasing the article length, by adding references, including the things initially documented by Darwin and Wallace; also there must surely exist many references from farming literature describing the value of dealing with the collective behaviour of animals on farms and in slaughterhouses; references somewhere surely must discuss how collective behaviour is affected by decreased genetic diversity or habitat fragmented by development; and the references Epipelagic found represent a debate about which aspects of collective animal behaviour can survive human codification (by inspiration vs. algorithms vs. continuous networks), perhaps also we might try to find a reference about poetic writing about collective animal behavour if the article isn't only about scientific perceptions. I'll take maybe 1 week to try to finalize the list of additional references.Createangelos (talk) 08:32, 20 July 2015 (UTC)

Social behaviour

Is this the same as Social behavior? If not, what is the difference? ghouston (talk) 22:59, 9 November 2017 (UTC)

Read the lead sentences of the respective articles. Social behaviour is defined as a "behavior among two or more organisms, typically from the same species". Collective animal behaviour is defined as "the coordinated behavior of large groups of similar animals...". The emphasis is on large groups, where all the animals are simultaneously behaving in a coordinated manner. A lone peacock courting a lone peahen would be a social behaviour, but would not be a collective behaviour. A large number of herrings schooling, that is, moving together in a coordinated manner, would be a collective behaviour and also a social behaviour. Collective behaviours are a subset of social behaviours, that is, all collective behaviours are social behaviours but only some social behaviours are collective behaviours. --Epipelagic (talk) 05:45, 10 November 2017 (UTC)

OK, thanks. ghouston (talk) 07:57, 14 November 2017 (UTC)