Given that my family has two border collies back home (that I get to visit in a few days), I think the stuff done with Chaser is really cool. To me, the most interesting part is the demonstration of deduction by picking out the novel object among familiar ones. The nature of the training ultimately encouraged Chaser to pick out an object rather than come back empty-handed (which would have been a reasonable response in this task). And in fact, she started out doing just that: just standing among the objects without retrieving anything. Once the "encouragement" made it clear that the trainers wanted Chaser to pick something, she was seemingly able to reason out that it was the novel object. I also found it cool that they did not use food, but instead play and praise, as reinforcement. This may be a good example of how a non-human animal like a dog can find certain things to be intrinsically reinforcing outside of pure survival factors. Very cool stuff.
This is all great if the goal is to test the limits of dog cognition, but personally I would be cautious about comparing this to how children learn.
And I'll just close out my last blog post of the semester with a completely unscientific anecdote relating to the discussion from last week about whether certain animals can display intent. The two border collies we have back at home are sisters from the same litter, with one of them being the clear "alpha dog" while the other tends to be more submissive. Whenever we give each of them identical bones to chew on. the dominant one will often abandon her own bone to steal the other's bone, seemingly as a means of exercising dominance. Although this would not qualify as communicative intent, when this discussion came up it made me think of how my dogs seem to (intentionally) steal from each other :)
These experiments are certainly interesting, and I believe that they are important for challenging the assumption that animals are comparatively mindless. However, by an enactivist description of the mind these results are almost to be expected.
An enactivist description? What is that? What is enactivism? Maybe the presenter will tell us before addressing the questions. =)
I thought these experiments were really interesting. While reading about Rico and Chaser I kept looking at my dog trying to picture him doing this. I have a Pomeranian that while adorable, is not the brightest crayon in the box. I wonder if anyone has looked at the range of cognitive abilities among different types of dogs. For example, border collies and german shepherds are typically thought of as very intelligent dogs. So could it be possible that these two breeds are analogous to gifted/genius people? I wonder if you ran the same experiment with different breeds of dogs how the results would vary. Border collies and german shepherds are dogs that have been used more as helpers to humans (like other breeds from the herding group, working group, or sporting group), therefore needing to understand human communications (whether verbal or something else). Whereas toy dogs are bred simply for companionship. So I wonder if toy dogs would have a lower cognitive capacity or possibly a different area of cognitive abilities are more developed.
The one questioning that dominates my brain after reading these articles is, "if these animals are capable of such feats of lexical and numerical memory, what separates our abilities from theirs?". Their abilities and similarities to our abilities fascinates me, but I am more fascinated by how human lexical and numerical understanding can become self-generative. Could an animal be trained to begin discrimination without guidance? Could they be taught to manipulate letters to make their own words/sounds? (I want to work with animals now.) Could it be a matter of exposure? A momentum within exposure over time? Is it the constant stimuli humans are exposed to that enables our abilities? If we set a group of dogs/birds into motion, could they eventually gather their own understanding? What about putting two birds together to see if they could communicate using human language? That would be awesome to see, and would likely turn viral on YouTube in a matter of seconds. :)
I thought these papers were interesting. We seem to be building up a stronger and stronger argument for mental representations in animals (or something like mental representations, for the sticklers out there). As Blair and a short part of one conclusion point out, these capabilities may be the result of selection in some animals, which means that there may be variations in mental abilities within a single species. But, we see this representation-like ability in a variety of species as well, from dolphins to dogs to parrots. I wonder if there's a way to identify the pressures that result in representation abilities being advantageous for an animal. Let's say, for instance, that cats do not have representations and that some dogs do (who knows if this is true; cats would probably not let on if they did as part of their secret world domination plot). Why might this be? And what about horses, which do take verbal commands from humans and are bred in a way similar to dogs. In nature, why would a Grey parrot find representations advantageous vs a pigeon? I feel like if we identified enough animals with some of these capabilities, we might understand why representation abilities exist in some animals and not in others. Somehow, this might make these abilities easier to understand in terms of pervasiveness and degree. Perhaps also this would also settle the "degree vs kind" argument that seems at the core of these papers.
I also thought it was interesting that at the beginning of the parrot article, the author states that parrots can mentally represent objects, but that training them to communicate this has been inappropriate to the species (presumably, until his experiment). Perhaps this problem is pervasive - if we knew how to test each animal in an appropriate way, maybe we would see more animals displaying this ability. If different species evolved this ability due to different circumstances in the environment, would it not follow that getting them to produce the associated behaviors would be somewhat unique to the species?
Basically, can my cat represent his pom poms differently than his silly squids? I must know!
11-21-2011 One critical point: Pilley and Reid (2010) made it clear that the experiments they designed did not have the intent to test Chaser’s innate abilities independent of training but discover what lg accomplishments Chaser could achieve (p.3). On the contrary, Pepperberg (2002) claimed that the intensive and skillful training imposed on the grey parrot is a direct and simple means for testing animal’s lg intelligence (p. 55). I definitely go by Pilley and Reid’s point of view. However, it is not saying that I don’t think animals are mindless and do not have the capacity to learn. Animals do have some learning capability, but only under specific and intense training condition could they expand or develop that kind of capability. Pepperberg was not completely wrong about animal cognition; she recognized the possibility that grey parrot and other nonhuman primates can gain communicative abilities through social interaction with conspecifics or other organisms. Only through interacting with either the environment or their own species friends can some specific kinds of animals be triggered to learn human lgs. This also demonstrated why mother-infant interaction is that essential to children's lg learning. Still, even animals have certain innate capability to develop lg, it is not so important how much they can learn, but why they can but cannot if without training? Similarly, what kind of selection pressure that makes them capable of learning so well the motor skills but not vocal ones?
People got to play with their dogs and get published in Science and Behavioral Processes. Well, Whoooopedy-doo! No, really, the Chaser study was actually well managed (but still, I'm not a big fan of case studies, especially when it was their dog.). It was the only one of the 3 readings that made me not think it was totally behaviorism (everyone's favorite topic!). The other two studies just seemed like training. We have a parrot, it has a personality and definitely has preferences. It says stuff. But can it ever hold a conversation that I can understand? I don't think so (although, you wouldn't know that. I keep talking, he keeps talking, we whistle and everybody's happy).
I will say that I found the parrot and Rico studies just good examples of behaviorism (and more susceptible to the clever Hans effect). But really, maybe what we humans think and do is too? Yeah you might be able to argue we have extra reasoning abilities, but all the heuristics that we use (that I can think of) were first used with another category. Then we categorize our categories. We were also reinforced at some point. We then can make choices and suppositions about unknown possibilities based on categorization and previous outcomes. And then ho-and-be-lowed, we look like we have reasoning. :D
I had mixed reactions to these articles. One thing that bothered me (as Yuna brought up) was that the Pepperberg article was claiming that training the parrot was a measure of its intelligence. I just don't think that's the case. I think there is a difference between stimuli-response actions and intelligence. However, one could argue that most of what humans do that we consider to be signs of intelligence, is just a stimulus-response situation. Which might be what Rick is saying. I also started thinking about what it is that separates our abilities from those of animals. And I thought, is it possible that there is no real difference, just that our brain is more developed and we thus have hightened capabilities? Are we all doing the same thing, just to a different extent? Do we now want to admit that because we want to think we're special and to compare our intelligence to that of a dog seems degrading to us?
And yes, I realize that I've expressed kind of opposite opinions from stat to finish here.
Once I learned one of the obvious differences between human beings and animals is the language: human uses the language, but animals don’t. These two weeks’ reading seemed to challenge this concept. I agree animals have their own unique ways to communicate and to learning, or even to acquire the meaning representation. However, the characteristics of the language like displacement, arbitrariness, productivity, cultural transmission, and so on only are only shown in human beings. The different animals perhaps embody some of these characteristics during their evolution for the survival, but it is appropriate to compare animals with human infants. Does some research study whether infants learn when they are still in mother’s wombs? Why is the antenatal training so popular in the world? Before infants start to speak, have they already acquired some rules for the meaning representation? Does this prior acquisition lead to their “language explosion” when they speak at the age of one and the “vocabulary explosion” at the age of about five or six? I am not sure whether this deduction is correct when we see some peripheral phenomena occur?
I am also not very fond of case studies. However, if you are to make certain cases that animals CAN have certain human abilities, they are valuable research. So it depends on what your research question is. If the research question is: "DO animals have the same kind of abilities that we human have?" In that case, since you are generalizing, case studies would not be very useful. But if your question is: "CAN animals have some human abilities?" Then case studies would be useful However, I want to emphasize this point that case studies cannot be used for generalization. It is a known FACT that you can have certain genetic mutations that can result in extraordinary abilities. We have case studies of savants, geniuses, prodigies in human beings, so why can't we have savants, geniuses, prodigies in non-human animals as well? In previous discussions people have pointed out this erroneous idea of point-referencing humans. So if an alien came down to earth, and someone like Mozart and Einstein would be their level of intelligence at their equivalence of age 3 to 5 let's say, would it be correct for them to ask the question: "CAN human beings have XXX(Alien) abilities?" And would they be correct in making case studies on Stephen Hawking and then conclude that human beings CAN. I guess from all the research I am convinced that Povenalli is wrong in trying to emphasize dichotomy and not continuity. But then again, what exactly is he and others trying to "disprove" Darwin (by using so many strawmen arguments) is still unclear to me.
Yay dogs! I wish that my dog was this smart. Unfortunately, he is quite the opposite.
Soooo my question is going off of Blair and Whitney's comments. I think it makes sense that different species have developed "functional" cognitive abilities. My bichon has a much different purpose than a border collie might have. I wonder if people have tried to train dogs like poodles or Pomeranians to do the things that say a border collie might do. Can poodles hunt?
Next thing. Along these same lines, can we look at humans in any way that is analogous to this? It seems like it would be stereotypical and offensive to do so, but I'm just curious.
Back to the article: I'm fine with case studies. I think they are fun to read. I am curious though if more research is warranted here that will use different methods. For example to make some of the comparisons that have been discussed?
These were interesting readings. I was contaminated last week by the cuteness of voles but nonetheless I still am attracted to studies on animal learning abilities in the context of human language, even when those animals are not especially small and fluffy . I don’t see a clear path between the title “Animal Communication” and the articles for this week, maybe something like “Abilities of Animals related to Human Language”, or a title about animal learning would have been more appropriate. I believe that Blair’s mention of differences between breeds is very smart. It would be one of the hottest topics (for me, at least) inside the field of animal/human cognition. How do you obtain differences between populations (breeds) in terms of intelligence? We already know how to do it for size, speed, aggressiveness, etc. I agree that intelligence is more abstract and complicated but isn’t it worth experimenting with, using the scientific method, and not only the subjective measurements of dog breeders along the centuries? I also like Whitney’s interspecies approach and her touch on the "degree vs kind" argument. I agree with Shi’s approach on the value of a case study. You can disprove an “ALL”-type hypothesis with only one counter-example and that makes it valuable research, unless you suspect that it is completely unreasonable to formulate that hypothesis with a straight face, but I don’t think it is the case here.
sorry this is late... i forgot... :( As human beings, we all too often view non-human communication from a language-based standpoint instead of from a functional standpoint. Following Occam's razor, the authors should first aim to determine if the results can first be explained from a functional standpoint whereby these animals are simply signaling to elicit some response on the part of the listener and not actually using referential signals.
A greater concern is that these specific animals had so much exposure and training with human communication. Furthermore, they were often rewarded for communicating and thus we would expect that they would initiate and respond to communication more often. Why this is more than simply a learned behavior?
These authors claim that animals can understand and recombine individual label units to create new signals. I basically am concerned that the authors are assuming referential signal use when a functional explanation might better (and more simply) explain the findings. I would suggest the author explain why this is not the case.
Sterling's comment convinced me to address the following question: is there anyone of you guys actually believing that these animals are using referential signals or that they are showing something more than just learned behavior?
Given that my family has two border collies back home (that I get to visit in a few days), I think the stuff done with Chaser is really cool. To me, the most interesting part is the demonstration of deduction by picking out the novel object among familiar ones. The nature of the training ultimately encouraged Chaser to pick out an object rather than come back empty-handed (which would have been a reasonable response in this task). And in fact, she started out doing just that: just standing among the objects without retrieving anything. Once the "encouragement" made it clear that the trainers wanted Chaser to pick something, she was seemingly able to reason out that it was the novel object. I also found it cool that they did not use food, but instead play and praise, as reinforcement. This may be a good example of how a non-human animal like a dog can find certain things to be intrinsically reinforcing outside of pure survival factors. Very cool stuff.
ReplyDeleteThis is all great if the goal is to test the limits of dog cognition, but personally I would be cautious about comparing this to how children learn.
And I'll just close out my last blog post of the semester with a completely unscientific anecdote relating to the discussion from last week about whether certain animals can display intent. The two border collies we have back at home are sisters from the same litter, with one of them being the clear "alpha dog" while the other tends to be more submissive. Whenever we give each of them identical bones to chew on. the dominant one will often abandon her own bone to steal the other's bone, seemingly as a means of exercising dominance. Although this would not qualify as communicative intent, when this discussion came up it made me think of how my dogs seem to (intentionally) steal from each other :)
These experiments are certainly interesting, and I believe that they are important for challenging the assumption that animals are comparatively mindless. However, by an enactivist description of the mind these results are almost to be expected.
ReplyDeleteAn enactivist description? What is that? What is enactivism? Maybe the presenter will tell us before addressing the questions. =)
I thought these experiments were really interesting. While reading about Rico and Chaser I kept looking at my dog trying to picture him doing this. I have a Pomeranian that while adorable, is not the brightest crayon in the box. I wonder if anyone has looked at the range of cognitive abilities among different types of dogs. For example, border collies and german shepherds are typically thought of as very intelligent dogs. So could it be possible that these two breeds are analogous to gifted/genius people? I wonder if you ran the same experiment with different breeds of dogs how the results would vary. Border collies and german shepherds are dogs that have been used more as helpers to humans (like other breeds from the herding group, working group, or sporting group), therefore needing to understand human communications (whether verbal or something else). Whereas toy dogs are bred simply for companionship. So I wonder if toy dogs would have a lower cognitive capacity or possibly a different area of cognitive abilities are more developed.
ReplyDelete-Blair
The one questioning that dominates my brain after reading these articles is, "if these animals are capable of such feats of lexical and numerical memory, what separates our abilities from theirs?". Their abilities and similarities to our abilities fascinates me, but I am more fascinated by how human lexical and numerical understanding can become self-generative. Could an animal be trained to begin discrimination without guidance? Could they be taught to manipulate letters to make their own words/sounds? (I want to work with animals now.) Could it be a matter of exposure? A momentum within exposure over time? Is it the constant stimuli humans are exposed to that enables our abilities? If we set a group of dogs/birds into motion, could they eventually gather their own understanding? What about putting two birds together to see if they could communicate using human language? That would be awesome to see, and would likely turn viral on YouTube in a matter of seconds. :)
ReplyDeleteI thought these papers were interesting. We seem to be building up a stronger and stronger argument for mental representations in animals (or something like mental representations, for the sticklers out there). As Blair and a short part of one conclusion point out, these capabilities may be the result of selection in some animals, which means that there may be variations in mental abilities within a single species. But, we see this representation-like ability in a variety of species as well, from dolphins to dogs to parrots. I wonder if there's a way to identify the pressures that result in representation abilities being advantageous for an animal. Let's say, for instance, that cats do not have representations and that some dogs do (who knows if this is true; cats would probably not let on if they did as part of their secret world domination plot). Why might this be? And what about horses, which do take verbal commands from humans and are bred in a way similar to dogs. In nature, why would a Grey parrot find representations advantageous vs a pigeon? I feel like if we identified enough animals with some of these capabilities, we might understand why representation abilities exist in some animals and not in others. Somehow, this might make these abilities easier to understand in terms of pervasiveness and degree. Perhaps also this would also settle the "degree vs kind" argument that seems at the core of these papers.
ReplyDeleteI also thought it was interesting that at the beginning of the parrot article, the author states that parrots can mentally represent objects, but that training them to communicate this has been inappropriate to the species (presumably, until his experiment). Perhaps this problem is pervasive - if we knew how to test each animal in an appropriate way, maybe we would see more animals displaying this ability. If different species evolved this ability due to different circumstances in the environment, would it not follow that getting them to produce the associated behaviors would be somewhat unique to the species?
Basically, can my cat represent his pom poms differently than his silly squids? I must know!
11-21-2011
ReplyDeleteOne critical point: Pilley and Reid (2010) made it clear that the experiments they designed did not have the intent to test Chaser’s innate abilities independent of training but discover what lg accomplishments Chaser could achieve (p.3). On the contrary, Pepperberg (2002) claimed that the intensive and skillful training imposed on the grey parrot is a direct and simple means for testing animal’s lg intelligence (p. 55). I definitely go by Pilley and Reid’s point of view. However, it is not saying that I don’t think animals are mindless and do not have the capacity to learn. Animals do have some learning capability, but only under specific and intense training condition could they expand or develop that kind of capability. Pepperberg was not completely wrong about animal cognition; she recognized the possibility that grey parrot and other nonhuman primates can gain communicative abilities through social interaction with conspecifics or other organisms. Only through interacting with either the environment or their own species friends can some specific kinds of animals be triggered to learn human lgs. This also demonstrated why mother-infant interaction is that essential to children's lg learning. Still, even animals have certain innate capability to develop lg, it is not so important how much they can learn, but why they can but cannot if without training? Similarly, what kind of selection pressure that makes them capable of learning so well the motor skills but not vocal ones?
People got to play with their dogs and get published in Science and Behavioral Processes. Well, Whoooopedy-doo! No, really, the Chaser study was actually well managed (but still, I'm not a big fan of case studies, especially when it was their dog.). It was the only one of the 3 readings that made me not think it was totally behaviorism (everyone's favorite topic!). The other two studies just seemed like training. We have a parrot, it has a personality and definitely has preferences. It says stuff. But can it ever hold a conversation that I can understand? I don't think so (although, you wouldn't know that. I keep talking, he keeps talking, we whistle and everybody's happy).
ReplyDeleteI will say that I found the parrot and Rico studies just good examples of behaviorism (and more susceptible to the clever Hans effect). But really, maybe what we humans think and do is too? Yeah you might be able to argue we have extra reasoning abilities, but all the heuristics that we use (that I can think of) were first used with another category. Then we categorize our categories. We were also reinforced at some point. We then can make choices and suppositions about unknown possibilities based on categorization and previous outcomes. And then ho-and-be-lowed, we look like we have reasoning. :D
I had mixed reactions to these articles. One thing that bothered me (as Yuna brought up) was that the Pepperberg article was claiming that training the parrot was a measure of its intelligence. I just don't think that's the case. I think there is a difference between stimuli-response actions and intelligence. However, one could argue that most of what humans do that we consider to be signs of intelligence, is just a stimulus-response situation. Which might be what Rick is saying.
ReplyDeleteI also started thinking about what it is that separates our abilities from those of animals. And I thought, is it possible that there is no real difference, just that our brain is more developed and we thus have hightened capabilities? Are we all doing the same thing, just to a different extent? Do we now want to admit that because we want to think we're special and to compare our intelligence to that of a dog seems degrading to us?
And yes, I realize that I've expressed kind of opposite opinions from stat to finish here.
Once I learned one of the obvious differences between human beings and animals is the language: human uses the language, but animals don’t. These two weeks’ reading seemed to challenge this concept. I agree animals have their own unique ways to communicate and to learning, or even to acquire the meaning representation. However, the characteristics of the language like displacement, arbitrariness, productivity, cultural transmission, and so on only are only shown in human beings. The different animals perhaps embody some of these characteristics during their evolution for the survival, but it is appropriate to compare animals with human infants. Does some research study whether infants learn when they are still in mother’s wombs? Why is the antenatal training so popular in the world? Before infants start to speak, have they already acquired some rules for the meaning representation? Does this prior acquisition lead to their “language explosion” when they speak at the age of one and the “vocabulary explosion” at the age of about five or six? I am not sure whether this deduction is correct when we see some peripheral phenomena occur?
ReplyDeleteI am also not very fond of case studies. However, if you are to make certain cases that animals CAN have certain human abilities, they are valuable research. So it depends on what your research question is. If the research question is: "DO animals have the same kind of abilities that we human have?" In that case, since you are generalizing, case studies would not be very useful. But if your question is: "CAN animals have some human abilities?" Then case studies would be useful
ReplyDeleteHowever, I want to emphasize this point that case studies cannot be used for generalization. It is a known FACT that you can have certain genetic mutations that can result in extraordinary abilities. We have case studies of savants, geniuses, prodigies in human beings, so why can't we have savants, geniuses, prodigies in non-human animals as well? In previous discussions people have pointed out this erroneous idea of point-referencing humans. So if an alien came down to earth, and someone like Mozart and Einstein would be their level of intelligence at their equivalence of age 3 to 5 let's say, would it be correct for them to ask the question: "CAN human beings have XXX(Alien) abilities?" And would they be correct in making case studies on Stephen Hawking and then conclude that human beings CAN. I guess from all the research I am convinced that Povenalli is wrong in trying to emphasize dichotomy and not continuity. But then again, what exactly is he and others trying to "disprove" Darwin (by using so many strawmen arguments) is still unclear to me.
Yay dogs! I wish that my dog was this smart. Unfortunately, he is quite the opposite.
ReplyDeleteSoooo my question is going off of Blair and Whitney's comments. I think it makes sense that different species have developed "functional" cognitive abilities. My bichon has a much different purpose than a border collie might have. I wonder if people have tried to train dogs like poodles or Pomeranians to do the things that say a border collie might do. Can poodles hunt?
Next thing. Along these same lines, can we look at humans in any way that is analogous to this? It seems like it would be stereotypical and offensive to do so, but I'm just curious.
Back to the article: I'm fine with case studies. I think they are fun to read. I am curious though if more research is warranted here that will use different methods. For example to make some of the comparisons that have been discussed?
These were interesting readings. I was contaminated last week by the cuteness of voles but nonetheless I still am attracted to studies on animal learning abilities in the context of human language, even when those animals are not especially small and fluffy . I don’t see a clear path between the title “Animal Communication” and the articles for this week, maybe something like “Abilities of Animals related to Human Language”, or a title about animal learning would have been more appropriate.
ReplyDeleteI believe that Blair’s mention of differences between breeds is very smart. It would be one of the hottest topics (for me, at least) inside the field of animal/human cognition. How do you obtain differences between populations (breeds) in terms of intelligence? We already know how to do it for size, speed, aggressiveness, etc. I agree that intelligence is more abstract and complicated but isn’t it worth experimenting with, using the scientific method, and not only the subjective measurements of dog breeders along the centuries? I also like Whitney’s interspecies approach and her touch on the "degree vs kind" argument.
I agree with Shi’s approach on the value of a case study. You can disprove an “ALL”-type hypothesis with only one counter-example and that makes it valuable research, unless you suspect that it is completely unreasonable to formulate that hypothesis with a straight face, but I don’t think it is the case here.
sorry this is late... i forgot... :(
ReplyDeleteAs human beings, we all too often view non-human communication from a language-based standpoint instead of from a functional standpoint. Following Occam's razor, the authors should first aim to determine if the results can first be explained from a functional standpoint whereby these animals are simply signaling to elicit some response on the part of the listener and not actually using referential signals.
A greater concern is that these specific animals had so much exposure and training with human communication. Furthermore, they were often rewarded for communicating and thus we would expect that they would initiate and respond to communication more often. Why this is more than simply a learned behavior?
These authors claim that animals can understand and recombine individual label units to create new signals. I basically am concerned that the authors are assuming referential signal use when a functional explanation might better (and more simply) explain the findings. I would suggest the author explain why this is not the case.
Sterling's comment convinced me to address the following question: is there anyone of you guys actually believing that these animals are using referential signals or that they are showing something more than just learned behavior?
ReplyDelete