VIEWS: 1 PAGES: 56 POSTED ON: 7/19/2013
The Cognitive Dog Class 2: The Great Debate Bruce Blumberg & Carolyn Barney Harvard Extension School Agenda for class • Big apology about the reading!!!!!! • Questions from last week • Plan for next week • The Brain Trust, Part 1. • Great Debate Plan for next week • The behavior of wolves and wild canids • Readings from course pack • Packard (2005) Wolf Behavior: Reproductive, Social & Intelligent • MacDonald et al (2004) Society Introducing the Brain Trust, part 1. The debate: the evolution of human-like social skills in dogs? A word to the wise “we agree with one of our reviewers saying that the present state of this field does not allow for scientific arguments on social cognition evolution” - Adam Miklosi & Krisztina Soproni Miklosi, A. and K. Soproni (2006). "A comparative understanding of the human pointing gesture." Animal Cognition 9: 81-93. Another word to the wise Scientific work almost always needs to be seen through the lens of the intellectual, social and funding context in which this work was done Miklosi, A., J. Topal, et al. (2004). "Comparative social cognition: what can dogs teach us?" Animal Behaviour 67(6): 995. A bit of a cottage Really about the evolution of social cognition industry Hungry UK Why? • Evolution of social cognition in humans is a “hot” topic • The argument is that dogs represent a really interesting animal model • Highly social • Presumably their behavior has been selected for success in the ecological niche of human-dog social groups. • Dogs “are enculturated” vs. Chimps who “can be enculturated” • Easy to find experimental subjects • Everybody likes a good dog story Hare, B. and M. Tomasello (2005). "Human-like Social Skills in Dogs." Trends in Cognitive Science 9(9): 439-447. Using dogs as a model to understand the evolution of social Why the the funding agencies care cognition in humans Why I/we care... • What can we learn that will help us understand • why the dogs in our households do what they do • how to better raise behavorally-healthy dogs • how to better train and interact with our dogs • how to better ensure that our dogs stay behaviorally healthy. The big picture • Do pet dogs use human cues (gestures, gaze, body position, motion, voice) to guide their behavior? • Well, do they? • If so, to what feature are they attending? • Why are they using the feature? • What mental representation do they build, i.e., is it a ‘simple’ association or do they understand the meaning of the cue at some ‘deeper’ level. • What larger story does this tell? In fact there are 4 big questions to ask... • Function: how does it enhance survival • Causation: what elicits the behavior, and are there learned components? • Development: when does the behavior appear and what role does development & developmental context play? • Evolution: do you see it in Niko Tinbergen related species how might it have arisen via evolution? Hare, B. and M. Tomasello (2005). "Human-like social skills in dogs?" Trends in Cognitive Science 9(9): 339-444 Miklosi, A. and K. Soproni (2006). "A comparative understanding of the human pointing gesture." Animal Cognition 9: 81-93. Pointing gestures as Experiments do try to control for olfactory cues cues Call, J., J. Brauer, et al. (2003). "Domestic Dogs (Canis Familiaris) Are Sensitive to the Attentional State of Humans." Journal of Comparative Psychology 117(3): 257-263. Attentional state as Is the dog more likely to grab the treat if the person isn’t facing cues them? Viranyi, Z., J. Topal, et al. (2004). "Dogs respond appropriately to cues of humans' attentional focus." Behavioural Processes 66(2): 161. Is the dog sensitive to the Attentional state as cues apparent attention of the person? What is the mechanism behind Words as labels word use? Miklosi, A., P. Pongracz, et al. (2005). "A Comparative Study of the Use of Visual Communicative Signals in Interactions Between Dogs (Canis Familiaris) and Humans and Cats (Felis catus) and Humans." Journal of Comparative Psychology 119(2): 179-186. Experiments generally show These results are more or less that dog’s choice is biased by typical gestural cues Things to note about experiments • Small number of subjects with repeated trials • Criteria is statistically different than chance • Results are typically aggregated • Typically subjects are adult pet dogs recruited from local training clubs, friends, etc. • Class 11 we will revisit methodology because it is often problematic Hare 2002 - 2005 9 out of 11 dogs used cue vs. 2 out of 11 for chimps Hare, B., M. Brown, et al. (2002). "The domestication of social cognition in dogs." Science 298: 1634-1636. The difference between dogs & Hare 2002 chimps was startling, and the question was why? Hare’s Three Hypotheses (2002) • “Canids in general are unusually flexible in the types of social information they can exploit” • “Domestic dogs ... have learned their skills during their individual ontogenies” • “Selection pressure on dogs during process of domestication for specific skills of social cognition and communication with humans” No wolf performed above chance using any cue. 7 dogs used GPT, 5 used GP, 4 used P to find food above chance. 3 dogs used all 3 cues, 3 dogs used 2, and 1 dog used just one. Gaze, Gaze, Point Control point, point touch Hare, B., M. Brown, et al. (2002). "The domestication of social cognition in dogs." Science 298: 1634-1636. Dogs performed differently This suggested to Hare that it wasn’t an ability common to than ‘socialized’ wolves canids Differences aren’t significant, but interesting that litter-reared did better than family raised. Hare, B., M. Brown, et al. (2002). "The domestication of social cognition in dogs." Science 298: 1634-1636. Litter raised pups did the This suggested to Hare that it same as family raised wasn’t developmental pups 9-12 week pups did as This suggested to Hare that there well as 17-24 week wasn’t a learned component pups Hare’s big conclusion • “These findings suggest that during the process of domestication, dogs have been selected for a set of social-cognitive abilities that enable them to communicate with humans in unique ways.” Things to think about... • Are there flaws with the experimental design and analysis? • There is an assumption that socialized wolves are the same as socialized dogs. Is this valid? • There is an assumption that extensive contact with humans prior to 8-12 weeks is required for pups to preferentially attend to humans. Is this valid? • There is an assumption that social learning doesn’t occur prior to 8-12 weeks. Is this valid? • There is an assumption that you can describe a generic pet dog, and that one can generalize across breeds. Is this valid? Hare 2005 • Hare’s statement: “dogs have an unusual ability for reading human communicative gestures... seems to have evolved during domestication” • Hare’s question: “unclear whether this evolution occurred as a result of direct selection for this ability... or as a correlated by-product of selection against fear and aggression toward humans” • Decided to test hypothesis using domesticated foxes that were explicitly bred to have reduced fear and aggression toward humans Hare, B., I. Plyusnina, et al. (2005). "Social Cognitive Evolution in Captive Foxes Is a Correlated By-Product of Experimental Domestication." Current Biology 15: 226-230. Trut, L. (1999). Early Canid Domestication: The Farm Fox Experiment. American Scientist. 87: 160-169 we will keep coming back to Dr. Belyaev’s foxes... Belyaev & his foxes Belyaev’s Fox experiment... • Initial goal was to produce tamer foxes • Started with a population of 465 foxes • 30% extremely aggressive (threatening?) • 40% moderately aggressive (threatening?) • 20% fearful • 10% quiet & exploratory Belyaev’s Fox experiment... • Criteria for breeding • Flee threshold • Flee distance • later generations, willing to approach • After 18 generations they had produced foxes that had some ‘dog-like’ behavioral and morphological characteristics... Trut, L. (1999). Early Canid Domestication: The Farm Fox Experiment. American Scientist. 87: 160-169 Significant change in timing of We definitely will be coming back to this one developmental milestones Pups & fox kits between 8-16 weeks Hare, B., I. Plyusnina, et al. (2005). "Social Cognitive Evolution in Captive Foxes Is a Correlated By-Product of Experimental Domestication." Current Biology 15: 226-230. Pups & domesticated fox kits This suggested to Hare that this skill was a by product of selection performed similarly for tameness Hare, B., I. Plyusnina, et al. (2005). "Social Cognitive Evolution in Captive Foxes Is a Correlated By-Product of Experimental Domestication." Current Biology 15: 226-230. Temperament of farm foxes This experiment was consistent with Hare’s view that this skill may interfere with was a side-effect of breeding for performance temperament Hare’s conclusions... • 2 alternative explanations for dog’s ability to read human signals • Communication hypothesis: this ability was directly selected for during domestication • Correlated by product hypothesis: this ability is simply a by-product of selection for tameness • He believes his results support correlated by-product hypothesis... • Nothing was being selected for other than tameness (e.g., ability to read human cues) and yet foxes did as well as pet dog pups Things to think about... • At one level, all Hare is saying is that cognition takes place in an emotional context. This highlights the central role that emotions and temperament play in a dog’s choice of what to attend to, and what to do. • This is why we devote so much time to emotion and temperament • What are the specific mechanisms that make a pet dog emotionally prepared to interact & attend to humans? • Can we tease apart the complex interplay of genes, development, developmental context, & learning? • This is why we devote so much time to development Is this the whole story? • Hare’s argument: lowered emotional reactivity was selected for, and at a minimum, this set the stage. • How much more is needed? • Miklosi’s argument: that is not the whole story, social skills were selected for as well... Kubinyi et al, 2007 Social evolution: from wolf and dogs to humans • Research question: can we use the presumed evolution of social cognition in dogs to say anything about the evolution of social cognition in humans? • They claim dogs are an interesting model because... • “Behavior changed in a way that made them successful in the human social environment” • “Behavior of dogs’ ancestor species can be reconstructed from the behavior of the wolf” • “The natural socialization of dogs in the human environment offers a parallel between them and children. Kubinyi, E., Z. Viranyi, et al. (2007). "Comparative Social Cognition: From wolf and dog to humans." Comparative Cognition & Behavior Reviews 2: 26-46. Unique evolutionary history... • Dogs were first domesticated animals (when this occurred will be a topic for another lecture...) • Their model • subset of wolves adapted to new ecological niche provided by humans • traits implicitly/explicitly selected for by humans • Explicit selection for behavioral and morphological traits (19th century) produced dog breeds. The family wolf project: if wolves are raised like pet dogs, do they act like pet dogs? • They take as a given that dogs and wolves have a very different developmental time course... • Critical period for socialization seems to begin prior to 10 days in wolves vs. 3 to 5 weeks in dogs, and in dogs it extends up to 12 weeks. • In wolves, 24/7 contact with humans seems to be required • In pet dogs, minimal contact/presence seems to be enough to scaffold social attraction to another species (such as humans...) • Presence/absence of litter-mates seems to have a different effect in wolves vs. dogs. • Why this should be is the most fascinating question of all to me!!!!!!!! The family wolf project • Its a tough job but somebody has to do it :-) • 13 wolf cubs & 11 pups • Raised similarly • Tracked comparative behavior over approximately 2 years • “... intensive early handling proved to be an effective means of socializing wolves to a level comparable to dogs...” Kubinyi, E., Z. Viranyi, et al. (2007). "Comparative Social Cognition: From wolf and dog to humans." Comparative Cognition & Behavior Reviews 2: 26-46. Kubinyi, E., Z. Viranyi, et al. (2007). "Comparative Social Cognition: From wolf and dog to humans." Comparative Cognition & Behavior Reviews 2: 26-46. Statistically different Must depend on more than social performance on attachment experience... test... Pet dogs seem to be able to use pointing gestures more easily... • “dog puppies as young as 4 months old are able to perform well... without any special, intensive, and early socialization to humans” • Significant difference in performance between 4 month old wolf cubs and pet dog pups, but... • “After extensive training, wolves significantly improved in parallel with increased readiness to look at the pointing human.” Kubinyi, E., Z. Viranyi, et al. (2007). "Comparative Social Cognition: From wolf and dog to humans." Comparative Cognition & Behavior Reviews 2: 26-46. Wolves may be less likely to look at people • Tested wolves are significantly less likely to look to their handler when working on a blocked task, and when they do look, spend less time looking. • Attention is a prerequisite for learning. If wolves are biased against, or dogs biased toward attending to humans one would expect a difference in... • Learning to use human generated cues • Use of those cues to guide Kubinyi, E., Z. Viranyi, et al. (2007). "Comparative Social behavior Cognition: From wolf and dog to humans." Comparative Cognition & Behavior Reviews 2: 26-46. Miklosi’s interpretation • Wolves aren’t as good at using human cues as dogs because of “their decreased willingness to look at the human” • Conversely, “preferential looking at the human seems to be a genetic predisposition of dogs”... • at this is the “foundation on which developmentally canalized complex communicative interactions can emerge between man and dog” • In other words, a genetic bias to look at people was a precursor to the coevolution of dog-human communicative skills. Miklosi, A., E. Kubinyi, et al. (2003). "A Simple Reason for a Big Difference: Wolves Do Not Look Back at Humans, but Dogs Do." Current Biology 13(9): 763. Miklosi et al call this apparent difference in attention: “A simple reason for a big difference” • They suggest 2 processes might have been at work... • A bias to attend to people, and all that that brought along with it, may have been implicitly or explicitly selected for • Natural or artificial selection • Lower emotional reactivity in dogs may allow dogs to “tolerate being gazed at by humans better than wolves” • In either case: a genetic bias to look at people may have been a precursor to the co-evolution of dog-human communicative skills. Some things to keep in mind... • Any bias that exists in dogs... • may be to members of an imprinted species rather than limited to humans (e.g., live stock guarding dogs raised with sheep). • may have a wide variance across individual dogs and across breeds, e.g. a border terrier trying to get a rat in a cage doesn’t spend a lot of time looking back at its handler either... • it may well have arisen as a side-effect of some other difference... • There doesn’t need to be, and mostly likely isn’t, a gaze-at-human gene. • May not have been directly selected for. Social learning: from simple cues to selective imitation Local enhancement: “I think I’ll hang out with Harry. Hmmm, what’s that” Stimulus enhancement: “Hey what’s Harry fooling with. That looks tasty” Observational Conditioning: “Yikes, what is Harry reacting to, I guess I should be scared too” Three simple types of social Just because its simple, doesn’t make it any less useful learning These previous types of social learning can be explained via associative learning, but... • Josep Call postulates 2 alternative mechanisms... • The ‘cue-based’ approach. The animal learns to respond to a given stimulus in a given way, or learns to predict a given future stimulus based on observing another given stimulus. • All about correlation, no mental model of causation, and hence little or no ability to generalize. • The ‘knowledge-based’ approach. Here the animal extracts functional features associated with the stimulus, and builds a functional model of greater or lessor complexity that it then uses to guide its choice of behavior • To the extent that the functional model accurately captures causation, it provides a mechanism to generalize. Call, J. (2001). "Chimpanzee social cognition." Trends in Cognitive Science 5(9): 388-393. Selective imitation • Dogs trained to pull a ring for food via mouth and paw • Control dogs given opportunity to solve this problem on their own. 85% used their mouth to pull on the rod. • Experimental dogs watched 8 trials of a demonstrator using their paw to push down on the rod. • One group, demonstrator has ball in mouth • Other group, demonstrator doesn’t have a ball in mouth Range, F., Z. Viranyi, et al. (2007). "Selective Imitation in Domestic Dogs." Current Biology 17: 1-5. Clear difference in performance between the two But why??? groups Possible explanations... • The dogs really are making the kinds of inferences that I have described on the previous pages... • If so, this is both quite remarkable and quite unexpected • Is there some confounding factor that isn’t apparent to us that makes it appear as if the dog is making an inference when in fact they are responding to something else in the experimental set-up. • In either case, it is a fascinating question to ponder... In the end... • The course will focus on many of the issues that are at the heart of much of the work described above. • origins of the dog • development • emotion & temperament • social learning • In class 11 and class 15 we will re-examine the original work with maybe a few more answers to the questions, and even more questions Next week: wolves & wild canids
"lecture 2 - iSites"