Animal Cognition: concept formation

[DarkWaters]

I am presently reading through OPAR for the first time. Incidently, I am having trouble understanding how we know for certain that animals cannot form even very simplistic concepts.

To my understanding, some mammals can do some pretty sophisticated things. Some examples that I have heard through various sources:

• A few apes (I believe gorillas and orangutans included) have been able to learn a decent number of words and expressions in sign language.
  
• Apes are particularly good at using tools to accomplish goals, such as using a broom to knock down a bunch of bananas from an otherwise inaccessible place.
  
• Some dogs have demonstrated the ability to use process of elimination. Given three different kinds of toys, exactly two of which have been identified with the animals with names, several dogs fetched the correct (to them, unidentified) toy when asked for it by the name which they had not been taught.
  
• It is widely believed that chimpanzees have casual conversations in the wild on matters such as good places to get food, the whereabouts of other chimps and so forth.
  

I do not understand how animals can engage in some relatively complex behavior and demonstrate learning abilities without having any basic (and surely vastly inferior to a human's) concept formation. Perhaps I am not correctly grasping what concept formation truely is? Perhaps all of the above examples are just faulty conclusions of bad science or bad science writers?

Needless to say, my intent is to learn and not to be contentious or to promote the philosophy of David Hume.

[ggdwill]

Since the purpose of this forum is philosophical rather than scientific inquiry, I'm going to risk changing the focus of this topic.

While the examples you list are compelling, I don't think they're compelling enough to dislodge man from his position as sole possessor of a conceptual conciousness - at least not in any context relevant to the less-fundamental branches of philosophy (which is what you will discover the rest of OPAR addresses).

While it may be true that some animals are capable of forming primitive concepts - I don't know, I'm not a scientist - it is glaringly obvious that nothing any animal has ever done approaches the cognitive level necessary to include animals in moral or political affairs.

With that in mind, be careful not to misconstrue the observation made in OPAR that because man can form any concept, he possesses rights. While the ability to form concepts is certainly the basis, it is the ability to form certain specific concepts - as well as concious adhearance to those concepts - that endows him with rights. These would be the concepts of personal sovereignity as well as private ownership of property.

Assuming the examples you listed were true, Dr. Peikoff would have done well to point out that it is not solely the capacity to think conceptually, but rather a much greater capacity to think conceptually, that makes humans unique.

- Grant

[DarkWaters]

While the examples you list are compelling, I don't think they're compelling enough to dislodge man from his position as sole possessor of a conceptual conciousness - at least not in any context relevant to the less-fundamental branches of philosophy (which is what you will discover the rest of OPAR addresses).

While it may be true that some animals are capable of forming primitive concepts - I don't know, I'm not a scientist - it is glaringly obvious that nothing any animal has ever done approaches the cognitive level necessary to include animals in moral or political affairs.

With that in mind, be careful not to misconstrue the observation made in OPAR that because man can form any concept, he possesses rights. While the ability to form concepts is certainly the basis, it is the ability to form certain specific concepts - as well as concious adhearance to those concepts - that endows him with rights. These would be the concepts of personal sovereignity as well as private ownership of property.

Assuming the examples you listed were true, Dr. Peikoff would have done well to point out that it is not solely the capacity to think conceptually, but rather a much greater capacity to think conceptually, that makes humans unique.

I agree with all that is said here. The more I think about this, the less consequence I see in whether or not animals have the ability to (at best) form very basic concepts. Dr. Peikoff mentioned, as an example, that a dog cannot form the concept of "bone". This makes sense to me, but I always wondered if dogs could form a broader concept such as "food". It seems pretty evident, that even if a dog could form this could, a dog could not nearly exploit this concept well enough to foment intellectual growth in anyway comparable to how a human could. I think we would all agree that even if a dog does have the concept of food, they cannot really apply it well at all, as many dogs will avoid eating some things like a cactus but might still attempt to consume unappetizing objects that would induce nausea.

Needless to say, I never for once suspected that we could teach a manatee the philosophical shortcomings of a sophisticated school of thought, such as utilitarianism.

Thank you for your response.

[DavidOdden]

To my understanding, some mammals can do some pretty sophisticated things. Some examples that I have heard through various sources:

• A few apes (I believe gorillas and orangutans included) have been able to learn a decent number of words and expressions in sign language.
  
• Apes are particularly good at using tools to accomplish goals, such as using a broom to knock down a bunch of bananas from an otherwise inaccessible place.
  
• Some dogs have demonstrated the ability to use process of elimination. Given three different kinds of toys, exactly two of which have been identified with the animals with names, several dogs fetched the correct (to them, unidentified) toy when asked for it by the name which they had not been taught.
  
• It is widely believed that chimpanzees have casual conversations in the wild on matters such as good places to get food, the whereabouts of other chimps and so forth.
  

We ought to have a FAQ on this. Regarding point 1, some apes have been trained to engage in manual-gesture behavior which some researchers interpret as indicating the use of "sign language". As long as the claim is reduced to saying that they have learned some association between gesture and consequence (thus, they have not learned a language, which involves much more than just random words), this seems to be true. Point 2 is correct; I would like to see the evidence that point 3 is correct (I suspect that the experimental design wasn't rigorous enough). Point 4 is a statement about beliefs (of whom?): I know of no evidence that it is a fact.

I do not understand how animals can engage in some relatively complex behavior and demonstrate learning abilities without having any basic (and surely vastly inferior to a human's) concept formation. Perhaps I am not correctly grasping what concept formation truely is? Perhaps all of the above examples are just faulty conclusions of bad science or bad science writers?

Probably the latter is true. Popular science writers usually do not have the training to handle hard science plus behavioral science, and unfortunately the monkey researchers are, well, simply bad scientists. This has been a long-standing problem in the field.

I don't know what exactly you know about concept formation, but keep a firm grip on the fact that complex behavior is not the same as concept formation. To form a concept, you must be able to abstract from what you perceive, to differentiate the things you perceive (you have to be able to tell that X and Y are actually two different things), and you have to be able to integrate those things into one mental unit. In addition, and you may consider this to be an unfair requirement on concepts (it knocks critters out of the competition totally), you have to be able to assign that mental unification to one symbol of some sort.

One of the important facts about human cognition that makes concept formation possible is the fact that we can abstract things discretely. It is likely that animals process things gradiently, so they don't really have an idea of "cat", they have some architype and then other things are like or unlike the "cat" architype to varying degrees. The dog "concept" of food is more a post hoc analysis, namely whether they can eat it and get away with it (they never seem to remember the answer to the latter question), so "food" could encompass leaves, string, rocks, paper towels, cactus (yes, cactus, though not the prickly ones), socks but, oddly, not a new brand of kibble. What hasn't ever been demonstrated in animals is the ability to divide the world up into discrete categories, something that human can easily do. If that ability could be demonstrated, there might be some point in debating whether some animal has a rudimentary form of concepts; until that happens, I think the discussion of the complexity of animal behavior is pointless -- it still isn't conceptual.

[JASKN]

The dog "concept" of food is more a post hoc analysis, namely whether they can eat it and get away with it (they never seem to remember the answer to the latter question), so "food" could encompass leaves, string, rocks, paper towels, cactus (yes, cactus, though not the prickly ones), socks but, oddly, not a new brand of kibble.

I laughed out loud at this. Very funny.

[DarkWaters]

I would like to see the evidence that point 3 is correct (I suspect that the experimental design wasn't rigorous enough). Point 4 is a statement about beliefs (of whom?):

Your responses were very helpful. I did some brief research on points three and four, and it appears that I have not portrayed them accurately. Point three seems to refer to just a specific dog, a border collie named Rico. Two links on this dog can be found here and here. Of course, these are non-scientific resources, so we should take these for what they are.

With regards to point 4, I believe the beliefs I recited, which I see in zoos throughout the country, ultimately stems from people like Jane Goodall and possibly Dian Fossey. With regards to the Jane Goodall's claims on chimpanzee communication, it seems that they are limited to announcing basic information (i.e. Food is here. Threatening predator is there. I want food. I want attention.) and not so much an exchange of ideas or opinions (i.e. these bananas are sure better than those peaches. Chimp A sure is being rude today.) This being said, we could hardly qualify the former as conversing when it seems to be just the ability to communicate a decent amount of information to other chimps.

It is likely that animals process things gradiently, so they don't really have an idea of "cat", they have some architype and then other things are like or unlike the "cat" architype to varying degrees.

What is your definition of "archetype" here? I am interpreting this as a (temporary) base for comparison. For example, a dog observes cats A, B and C. A dog might later observe cat D and think, "D is similar to A, B and C" but probably will not integrate all of his "cat experiences" into the concept cat. I am not even sure if a dog has the ability to compare cat D to previous cats he has sensed but might instead think more generally: "animal D is another small animal that I like to harass." Would this not presuppose that dogs have a concept of "living creatures" that includes most small mammals. I think that most dogs should be able to differentiate between a living creature and a toy that possesses some mechanism for locomotion. But the real question seems to be if they integrate all living creatures into a concept.

Thank you for your responses.

[DavidOdden]

With regards to the Jane Goodall's claims on chimpanzee communication, it seems that they are limited to announcing basic information

Okay, that's referring to an actual communication system of wild chimps, and not the over-inflated chimp sign language claims. We do know that apes have a number of diffferent facial gestures and vocal calls for conveying information. Though one source inficated that the "gesture" for conveying the notion of nausea was "vomiting". Duh. Actually, the guys at MPI in Leipzig are doing interesting research on ape systems, and they seem to have reasonable evidence that some of the "signs" are learned, which is an interesting breakthrough in ape research. But nothing close to concepts.

What is your definition of "archetype" here?

This is informal, but I think "average" would be the right basic idea. It took my hounds a fair amount of experience to figure out that the little Pekinese the next street over was a dog and not a cat -- initial, they had a cat-reaction, because they had no previous experience with pocket dogs. A Pekinese "averages" in better with cats than dogs (if your other dog experiences are with large dogs).

[Ifat Glassman]

Found an adorable video about a monkey doing Karate. It seems like the monkey is able to connect between certain words and required actions: Monkey Ninja

Another video that shows a monkey playing around with some stuff: Orangutan does housework.

I also discovered that some monkeys enjoy teasing other animals:

Monkey teasing dog. Notice how he comes back after the first time to pull on the dog's leg.

And... (video quality rather bad, and edited, but still funny): Monkey with a death wish

[DarkWaters]

Found an adorable video about a monkey doing Karate. It seems like the monkey is able to connect between certain words and required actions: Monkey Ninja

Another video that shows a monkey playing around with some stuff: Orangutan does housework.

These are apes and not monkeys!

Looks like the creators of these videos gets the concept: "hilarious apes".

[xavier]

Edwin Locke describes "a conclusive test of a chimp's conceptual ability" on page seven of his article:

http://www.fireflysun.com/ObjectivistPersp...nPsychology.pdf

"...Make a large pile of geometric objects differing in size, ... color, ...and shape...ask the chimp to bring the experimenter ten green triangles..."

(I can't copy and paste)

Edit: quote

Edited September 8, 2006 by xavier

[Ifat Glassman]

These are apes and not monkeys!

Sorry, didn't mean to dishonor your species, Gorilla person . In Hebrew we have just one word for "apes" and "monkeys", hence my mistake.

Now, about animals and concepts:

A concept is a mental identification of a set of two or more existents, which share particular characteristics. Concepts abstract away from the specific properties of concrete examples. Man creates concepts by understanding relationships of similarity and difference observed between existents

Taken from Objectivism wiki.

I was thinking about Squirrels.

One well-known trait of some species of squirrel is the gathering and storing of nuts for the winter. These squirrels are scatter-hoarders; they will gather nuts and store them in any accessible hiding place, usually by burying them.

(Wikipedia).

Squirrels collect more than one type of nut. They deal with a lot of Objects in their life, but they treat nuts (all the different types of nuts) the same. Wouldn't this be like a child able to sort objects according to shape or color? Moreover, if a squirrel will be introduced to a different type of food, I believe they will be able to learn that this is "food" and treat it like they treat other objects in the group (Like they learn to eat bread and other man-made stuff).

Another example, much more basic, is animals' ability to identify a member of their species of the opposite sex. We know they are able to do that since animals reproduce. The concept "member of the same species of the opposite sex" unites a group of entities with some common characteristics. Wouldn't this show that animals sort entities in groups as well? And if so, would that be conceptual thinking?

[KendallJ]

Now, about animals and concepts:

Taken from Objectivism wiki.

I was thinking about Squirrels.

(Wikipedia).

Squirrels collect more than one type of nut. They deal with a lot of Objects in their life, but they treat nuts (all the different types of nuts) the same. Wouldn't this be like a child able to sort objects according to shape or color? Moreover, if a squirrel will be introduced to a different type of food, I believe they will be able to learn that this is "food" and treat it like they treat other objects in the group (Like they learn to eat bread and other man-made stuff).

Another example, much more basic, is animals' ability to identify a member of their species of the opposite sex. We know they are able to do that since animals reproduce. The concept "member of the same species of the opposite sex" unites a group of entities with some common characteristics. Wouldn't this show that animals sort entities in groups as well? And if so, would that be conceptual thinking?

Ifat,

The definitions you gave were describing mental processes, but the examples you gave were of behaviors. What clearly shows that certain types of behaviors are ONLY the province of the process of conceptualization? in this case, generalization can be shown to follow from other processes besides just conceptualiation. It would seem that there is the possibility that because you generalize by conceptualization that you believe that animals generalize from conceptualization. In fact, you would have to show that generalization behaviors come ONLY from conceptualization mental processes.

Sexual identification is highly biochemical. It could be hardwired into the body chemistry, and in fact is.

Ants have been shown to be capable of seemingly complex foraging behavior, but in fact, when broken down this behavior is the result of a very simply algorithm based upon pheromone decay, coupled to the random meanderings of ants, and the simple instinct to follow a trail. This is hardly conceptualization.

I do a whole lot of animal training (my dog and I run agility). While he is capable of some very seemingly complex behaviors, once you start working with over the long term there are some indicators that say to me that he is not conceptual.

1. Their ability to generalize is very limited. To get a strongly trained behavior like "sit" or "stay" one has to repeat this behavior in all sorts of circumstances, over and over. Sit Inside, is not the same as Sit on the Deck, is not the same as Sit when there's a Squirrel Around, is not the same as Sit.

2. New behaviors start this process all over again. Even the most complex behaviors are built up from simple ones. there is no starting at a higher level because they have conceptualized it.

Take agility for example. This looks complex, but I know my dog is not conceptualizing. It is a series of trained behaviors that I direct, and that I string together. The dog watches all sorts of physical cues (arms, voice, body position, speed, etc) and responds. If he sees something in a cue that he's not seen before or is even a little slightly different, he doesn't know what to do, and I send him off course. It's a lot like dancing, which is not conceptual either. You can't think about dancing and improve, you have to do and practice to improve. Yes, you teach it conceptually, but you do by automatizing motions in response to cues (leads).

Yes, my dog has a rudimentary ability to learn and to generalize, but it doesn't operate the same way that my ability does.

I think the most convincing case for non-conceptualization in animals is that the fundamental mental process that Rand considered conceptualization allows for successive levels of conceptualization. That is, the process to abstract from abstraction is the same as the process to abstract from concretes. This means that conceptualization is theoretically infinite. Animals don't do this - AT ALL. No matter how well trained Moxie is, we always start over again for new behaviors, and he doesn't generalize one behavior from another.

What that would say to me is that the ability to generalize is not inherently conceptual. Neither is the ability to learn by positive reinforcement. Seemingly complex behaviors can result from just these capacities, but it is not at all convincing evidence for conceptualization.

[Ifat Glassman]

Ifat,

The definitions you gave were describing mental processes, but the examples you gave were of behaviors.

How else would you suggest to learn about mental processes of another living creature than watching it's behavior?

What clearly shows that certain types of behaviors are ONLY the province of the process of conceptualization?

Similar brain activity of an animal that watches concretes that have similar characteristics. Recording such brain activity has been done, BTW.

And observing an animal identify a group of objects that is based on more than one characteristic of that object.

in this case, generalization can be shown to follow from other processes besides just conceptualization.

Sexual identification is highly biochemical. It could be hardwired into the body chemistry, and in fact is.

Ants have been shown to be capable of seemingly complex foraging behavior, but in fact, when broken down this behavior is the result of a very simply algorithm based upon pheromone decay, coupled to the random meanderings of ants, and the simple instinct to follow a trail. This is hardly conceptualization.

You have to be careful here... Every thought every animal has is "highly biochemical" and it does have representation in the brain as a neuronal-circuit. If it doesn't, it cannot exist (the thought).

What you meant was that the process of identification is based on certain chemicals, but (and here is the main difference) this identification involves no mental connection at all between the chemical (smell) and other features of that animals such as shape, color etc' in ants.

I agree that if an animal is following a chemical, without being able to connect between the chemical and other features of that object, then it did not identify "the object".

But if an animals is able to connect a chemical with other (at least one) feature of the object, then it identifies "the object". Just like we humans are able to connect between smell, texture, and visual data of an object. Be sure that there are other ways to measure that object, that are not accessible to you as a human being (such as perceiving a 3D image created by a sonar, like some animals have).

Just because some animals have more limited senses than human beings, does not mean that by relying only on their limited resources they are not identifying "the object". If rats use their moustache and smell to identify an object then they identify the object. Just because you can also see the object very clearly does not mean that you are identifying the object and they are not. (I know you didn't say anything about this, I'm just expressing my thought here, the last part is not an answer to something you said).

The way mammals identify females of their own species is not pheromone based and nothing more. Animals also identify members of their own species based on visual data. How else would they stay as a herd? And there is a lot of difference between members of the species: some are young, some are mature, some have horns, some do not, some have white spots, etc'. Animals still distinguish members of their own species from other species, despite of individual differences. that is a fact.

Another example is how animals identify human beings (despite the fact that human beings come in different shapes). Animals learn either to be afraid of human beings, or to trust them. Human beings are a group of different concretes.

Same things for enemies of the species. The concretes are different from one another, but an animal will still identify it and call "danger" (or "food").

the process to abstract from abstraction is the same as the process to abstract from concretes. This means that conceptualization is theoretically infinite. Animals don't do this - AT ALL.

Yes! Thank you for saying that! This is indeed the difference between human beings and animals. Animals (except maybe for smart species, like apes) do not conceptualize abstractions.

However, animals do seem to think in concepts when grasping concretes. They are able to form a concept from a group of concretes, but not from a group of abstractions.

[DavidOdden]

How else would you suggest to learn about mental processes of another living creature than watching it's behavior?

One suggestion is to try to experimentally manipulate it. Let us hypothetically suppose that thee is some animal out there, other than man, which is has a conceptual consciousness. We would have to determine that they had a conceptual consciousness by observing their behavior, for sure, but unlike man, who we know does have a conceptual consciousness, we would have to design experiments that do not assume the conclusion that this animal is conceptual, and instead we would test whether it does. Lousy experimental design is where the ape-language fans mess up, and as far as I know there is also no decent experimental work to show that any animals can form concepts. The fact that a behavior is complex is no evidence that the animal engaging in it is a conceptual being. I assume that you understand what "concept formation" refers to -- then what you should do is design an experiment that would definitely test the hypothesis that some animal can actually form concepts.

(At this point, I should interject that it might be best to perform these experiments on dogs. Whatever the evolutionary precursors are to a conceptual faculty, I suspect that man has been selecting for that trait in dogs for millenia).

Similar brain activity of an animal that watches concretes that have similar characteristics. Recording such brain activity has been done, BTW.

I don't see how that has any bearing on conceptualization. I don't even believe that there is, yet, any physiological evidence that diagnoses a concept for humans. Remember that "concept" and "proposition" are different things; and in English, snakes, worms and millipedes are not the same concept, so if brain reactions were diagnostic of concepts rather than physical similarity, you should not get the "concept" reaction (presuming that there were such a thing in the first place) from seeing snakes, worms and millipedes. In other words, get into the details of these experiments and ask critically, "Does this really give evidence for concepts, or does it give evidence for something else?".

One issue which is always ignored in the animal cognition literature (by those of the "animals are just like people" ilk) is that you have to go through a long and directed training process to create a behavioral association. People form concepts freely, and by focusing on one aspect versus another, you can easily form multiple concepts where a particular concrete is part of this concept versus that one. As far as animals go, man forms the concepts and trains the animal to behave in a manner that might convince certain experimenters that the animal has learned a concept, because it behaves in such-and-such way.

Finally, concepts are not just floating abstractions, they are unified via a symbolic representation. Where is the evidence for a symbolic representation in animals? With people it is very easy to see what concepts we have -- we have words, like "rock", "dog", "circle", "blue", "justice". What is the mental object in the minds of these supposed conceptual animals that ties this concept together?

[Ifat Glassman]

DavidOdden, I don't see how we can have a discussion about this if you just ignore every example I supply.

If humans ability to identify species of animals is a conceptual ability, how is it not the same for animals who can also identify species (based on the same data that humans do, like visual data, smell, sound etc')?

[DavidOdden]

DavidOdden, I don't see how we can have a discussion about this if you just ignore every example I supply.

I suggest that you provide the scientific evidence, first, that some animal can in fact identify a species and, second, that it has learned to do so. Where is your scientific evidence?

[Ifat Glassman]

It is a known, very easily observed fact, that animals can identify species.

Animals hang around in herds, they make "danger" noises when encountering a certain animal (regardless of any differences it may have from other animals of it's species), and they identify types of animals that they eat (regardless of age, colors, gender).

What would you consider a "scientific proof"? In this case a proof that animals move in herds is obtained by observing reality.

Do you agree that animals hang around in herds?

And if you do, how would you explain their ability to stick to members of their own species, if not by identification that this is "this type"?

This is so obvious, that it makes me feel like the kid who is yelling "the king is naked!".

[DavidOdden]

What would you consider a "scientific proof"?

How about if we reduce this to journal citations? Give me citations of articles in peer-reviewed scientific journals to provide (defensible) experimental or observational evidence for this putative ability, and then we can examine the evidence for the claim, especially the experimental design. Also, reexamine my previous statement and your reply. What animals hang out in herds, and which ones have to learn what animals are part of "their species"? I'm suggesting treating the claim with respect: accept it if the evidence shows that it is a fact, and not otherwise. Focus on the evidence, and not the conclusion which you would like to be true. Show me why the ability to hang out with members of the same species proves that an animal has a concept "species", or any other concept.

BTW that reasoning would lead us to the remarkable discovery that connochaetes taurinus and equus quagga are the same species despite being in different mammalian orders. Do you now understand why your claim about herds proves nothing?

[Ifat Glassman]

What animals hang out in herds, and which ones have to learn what animals are part of "their species"?

Lions are predatory carnivores who manifest two types of social organization. Some are residents, living in groups, called prides. The pride consists of related females, their cubs of both sexes, and a group of one to nine males known as a coalition who mate with the adult females.

...

Both males and females will defend the pride against intruders. Typically, males will not tolerate outside males, and females will not tolerate outside females. Males are expelled from the pride or leave on their own when they reach maturity.

(Wikipedia)

I don't think I can find any scientific magazine that would discuss the question of whether or not some animals move in herds as a serious one.

Show me why the ability to hang out with members of the same species proves that an animal has a concept "species", or any other concept.

We need to agree on what demonstrates "conceptual thinking": if someone is able to sort objects according to similar characteristic, which means to differentiate them from other objects, and integrate them as belonging to the same group, and he would be able to do that based on more than one aspect of those objects (both smell and sight for example) then they would be thinking in terms of concepts.

In the case of these lions they are able to tell the difference between members of their group and outsiders. According to wikipedia they also differentiate between females that belong, and males that don't belong (implied). The female that might appear in their area might be of different color and size of other lioness, but they would classify it in a certain group. This classification is visible since they give the members of that group the same treatment.

The way they treat other animals is sorted according to groups: they react to snakes in a certain way, to outside lions in another way (according to gender), to prey in another manner.

If they had to learn how to handle each individual animal they meet each time, you would not see a lioness storming on a young deer spotted from a distance, since it would be a new concrete for her. "But she recognizes that it is similar to others like it" you say? well what do you know, isn't that conceptual thinking.

I see no difference between a squirrel sorting nuts from stones and a child sorting cubic shapes from pyramidal ones. Further burden of disproof is on you.

BTW that reasoning would lead us to the remarkable discovery that connochaetes taurinus and equus quagga are the same species despite being in different mammalian orders. Do you now understand why your claim about herds proves nothing?

I don't understand your last example, please explain.

[DavidOdden]

I don't understand your last example, please explain.

Zebras and wildebeests hang out in herds. They are not the same species, by any means. So "hang out in herds" is entirely irrelevant to having a conceptual faculty.

If you have some credible scientific evidence to contribute to the debate, which shows that some animals can form concepts, you should bring it out. For example, the ability of a squirrel to correctly sort nuts (of all types) from similar objects -- this experiment has not been performed, but it would be doable. Your evidence will only be credible if it shows evidence that the author does actually understand what a concept is. Something that I myself have been wondering, in this debate, is whether you actually understand the concept "concept". Can you summarize what a concept is, to the best of your understanding?

[Ifat Glassman]

Zebras and wildebeests hang out in herds. They are not the same species, by any means. So "hang out in herds" is entirely irrelevant to having a conceptual faculty.

It means both animals are able to identify objects as belonging to a certain group, and separate them from other objects.

This is different from the case of ants picking up a trail of a chemical since the identification has to be done (in the examples I provided, and in this one) by different criterions about that object.

If it is not conceptual ability to relate visual data + smell + other available characteristics of concretes to a group then I have no idea how anyone can conceptualize at all.

When a child is sorting objects according to their shape - is that a demonstration of conceptual ability? and if so, why is it that when a squirrel is doing the exact same thing (it does not attempt to store stones for winter, and I am NOT going to waist my time searching for a proof for that) how is that different?

And I supplied the definition of a concept already. I am working according to that definition.

[KendallJ]

Yes! Thank you for saying that! This is indeed the difference between human beings and animals. Animals (except maybe for smart species, like apes) do not conceptualize abstractions.

However, animals do seem to think in concepts when grasping concretes. They are able to form a concept from a group of concretes, but not from a group of abstractions.

Ifat,

You misunderstood my main point. If conceputalization is inherently the same process wether conceptualization from concretes or from abstractions, then animal don't conceptualize. They may be able to generalize (and it is highly limited) by some other mental mechanism than conceptualization but because they can't do both, what they are doing cannot be conceptualization.

David has taken up the fight with me, but the basic premise that "generalization behavior MUST come from a conceptual process" has to be proven. You are imlicitly using this premise, and walking around pointing at examples of "generalization" behavior and saying right away that this must be conceptualization.

It is not at all clear to me that recognition behavior requires a conceputal process. Pointing to generalization behavior is insufficient. You would have to show that the specific mental process of conceptualization is occuring. Otherwise the behavior you are pointing to is simple enough as to be brought about by "less than conceptualization" mechanisms which we already know exist in nature.

Conceptualization leads to broad, non-perceptual generalizations. Animals can't do that.

Conceptualization leads to conceptualization from BOTH concretes and abstractions. Animals can't do that.

It is not a more simple form of conceptualization that animals do but rather, it is a non-conceptual process altogether.

[KendallJ]

It means both animals are able to identify objects as belonging to a certain group, and separate them from other objects.

This is different from the case of ants picking up a trail of a chemical since the identification has to be done (in the examples I provided, and in this one) by different criterions about that object.

If it is not conceptual ability to relate visual data + smell + other available characteristics of concretes to a group then I have no idea how anyone can conceptualize at all.

If an ant can recognize a pheromone without conceptualization, then why can't another animal recognize 3, 4 or 100 factors about a certain existent without conceptualization? Even complex recognition does not imply conceptualization. measuring aspects of an entity (which is what recognition is) does not mean that one can abstract away from the measurements of that entity (which is what conceptualization is) to form a concept.

If I teach my dog to walk over to and lie down by a lamp, and then I change the lamp to a different one, there is no guarantee that he will understand what I mean or recognize the object. The best shot is if it looks almost identical. At best he has a "this concrete looks like that concrete" sort of behavior, or maybe even "this concrete IS that concrete".

[DavidOdden]

Ifatart,

Lions do not have a concept of "pride": it is not a concept, because it is closed, and concepts are open-ended. Savannah herd animals have an ability to identify "their group" which is part of their genome, and concepts are not hard-wired. Squirrels have no concept of "nut", instead they bury anything that is "somewhat like" the things that they were raised to bury -- there is no sense of abstraction or measurement (and without measurement, there are no concepts). This is entirely unlike humans, who can apply a concept to substantially new concretes and identify a Brazil nut as also being a nut. And none of these supposed animal "concepts" is unified under a symbol of any kind -- animals cannot think symbolically.

I asked you if you would summarize what you understand a concept to be, because I had a suspicion, based on the aspects of the argument that you have been persistently ignoring, that you simply did not understand what a concept is (and therefore why your arguments for animals as conceptual beings are non-probative). Your persistent failure to point to anything that supports the notion of an animal conceptual faculty and to squarely address these criticisms confirms my suspicion.

You continuously return to the argument "If I can't tell from behavior X what the nature of reality is, then I have no idea what reality is" is, no doubt, true for you, but I suggest that you could repair that problem. Rather than assuming that animals necessarily have a conceptual faculty, you should start by taking what you know (or should know). Man has a conceptual faculty; to understand what concepts are, look at human cognition; do not confuse behavior with causation and do not assume that superficially similar behaviors are caused by the same mechanism. And finally, do not let ignorance because a source of "knowledge" for you. If you do not know how to explain animal behavior, simply say that you do not know what causes a particular behavior, and do not attribute a conceptual faculty to animals because you can't think of anything better to do.

I will let Kendall continue this discussion with you on his own, if he is interested. The question of animal cognitive abilities is a very complicated scientific issue, and you are trivializing the issue in exactly the same way that Fouts, the Gardners and Patterson trivialised the concept of "meaning" and "language" in their ape projects. I was mistaken in my assumption about your willingness to engage in a serious scientific debate over this serious scientific question.

[Ifat Glassman]

If conceptualization is inherently the same process weather conceptualization from concretes or from abstractions, then animal don't conceptualize.

If you can't lift a 175kg weigh, and it is essentially the same process as lifting a box of milk, then you can't lift at all.

Some people have a hard time (up to no ability) to understand certain concepts in mathematics. Does that mean that they can't conceptualize?

They may be able to generalize (and it is highly limited) by some other mental mechanism than conceptualization but because they can't do both, what they are doing cannot be conceptualization.

...

the basic premise that "generalization behavior MUST come from a conceptual process" has to be proven. You are implicitly using this premise, and walking around pointing at examples of "generalization" behavior and saying right away that this must be conceptualization.

Here are several methods of identification (or selectivity):

1. Having a mental symbol of similar characteristics of a certain group of concretes. Comparing concretes to that symbol as a method of identification.
   
2. Having a mental symbol of some shape/color/smell and reacting to that shape/color/smell in the same way whether it is attached to the object at hand or not.
   
3. Having an automatic behavior pattern that is triggered by a direct, invasive physical contact between a certain chemical and the organism, like a bacteria that changes the way it swims when a molecule of glucose is physically attached to it's receptors.
   

A concept is composed of several mental images of characteristics (like shape, smell, and combination of colors) that are related in our mind to the same entity.

The essential difference between the first method and the second is that in the first one, the characteristics, those mental symbols are related to some entity, while in the second they just stand for themselves.

When you identify a cat you are doing this because the cat makes a certain sound, has a certain shape, way of movement etc', and you relate all those things to the entity cat.

Only other option that produces the result of identifying a cat is if the characteristics of a cat would not be attached to that entity, but would be reacted to separately.

An example of an experiment to check this: Record sounds of a cat, and play them as if they are coming from your mouth, and see if the dog will choose to attack you. If not, then the only explanation is that the dog has learned that the shape of a cat and the sound a cat makes has to come together, to actually be the annoying entity that they would rejoice in attacking. This connection between more than one stimuli to the entity that has them, indicates conceptual thinking.

*Note: all dogs mentioned in my post are cat haters.

It is not a more simple form of conceptualization that animals do but rather, it is a non-conceptual process altogether.

Interesting how you think that one thing should be proven, but not the other.

If an ant can recognize a pheromone without conceptualization, then why can't another animal recognize 3, 4 or 100 factors about a certain existent without conceptualization?

You are confusing method 3 with methods 2 and 1 here. A pheromone creates a physical connection that triggers automatic mechanisms. Light waves by themselves, do not trigger a mechanism of hunting.

When several factors has to appear together for the animal to recognize them as an entity, yet the animal is able to recognize each element individually (and search for a cat when hearing the sound), then they obviously have made a connection between those elements to the entity, and this is exactly how humans do it.

Edited September 10, 2006 by ifatart