ron2:In my own anecdotal way, I see some cognitive processes in dogs. I also see that humans don't always generalize, even as adults. Which brings me to question the value of using humans as a rule of thumb for the ability to generalize. Which I don't guess we can help, much. We are differentiating, but so do dogs. And that, to me, suggests that it is more likely that dogs do have cognitive ability, including the ability generalize.
I agree that much of human cognitive processes are mechanical and not self-aware. However, when it comes to things like generalizing, I think it depends on how much of a stickler you are for what that means.
In terms of cognition there are two basic definitions: 1) the process of formulating general concepts by abstracting common properties from a set of variables (which clearly doesn't apply to dogs as they don't have the cognitive architecture for conceptual and abstract thinking), and 2a.) the ability to transfer a response learned in relation to one stimulus to a similar stimulus, or 2b.) the ability to learn something in a certain context and apply it to other contexts as well. To avoid confusion I would define both of these 2nd cognitive phenomena as cross-contextualization, not generalization.
You also have to remember that when dealing with levels of cognition in dogs, humans, apes, whatever mammal you choose, that neuroscientists and cognitive scientists are now finding that there are pre-cursors built in to the more primitive parts of the brain, which dogs and humans share. These pre-cursors support, and in fact are necessary to the higher forms of cognition, which are shown to take place in parts of the brain that dogs don't have. For instance:
There seems to be a pre-cursor to language in primitive forms of imitation, mimicry, and non-cognitive forms of expression.
There seems to be a pre-cursor to a sense of self in an area of the mid-brain diencephalon called the periaquaductal gray*.
There is clearly a pre-cursor to logical thought in the ability to recognize patterns.
I would say it's pretty clear that dogs are very capable of the kind of thought generated through these pre-cursors, but not their more evolved constructs.
And I think that's why it's easy to believe that dogs are capable of the higher levels of thought because they involve these pre-cursors, that were a step in the evolution of these higher types of consciousness. We interact with our dogs through language regularly, and they respond as if they understand the meanings of our words. So it's hard not to think that they "understand" symbolic language, because on the simplest level (for below that of conscious thought) they do: they learn to respond to our words. But there's a difference between a learned ability and one that is hard-wired or innate. Also, it's very doubtful that wolves and coyotes give names to one another, but we give names to our dogs, and those names are associated with all sorts of interactions we have with our dogs on a regular basis, and it's hard not to think that dogs have a sense of their own identity when they respond not only to their names, but the names of other dogs they like. Hence we tend to think they must have a sense of self, which is a kind of requirement for first-level cognition, and a subsequent theory of mind. But EEC (embodied embedded cognition) would explain these behaviors quite easily without needing to circumvent all the evidence showing that dogs DON'T have identities, or a sense of self, in anything close to the same way that we do.
Personally, I think it's fascinating to actually study and read how these abilities developed and evolved, what parts of the brain are involved, and how one can know with a pretty high level of certainty where the cut-off points are, neurologically and behaviorally speaking.
But maybe that's just me,
LCK
*From "Empathy and Consioussness" by Evan Thompson:
Jaak Panksepp has proposed that the panoply of emotional states we experience can be analysed into certain core affective comportments, probably common to all mammals, that depend on distinct, basic emotional operating systems in the brain, and are tied to an animal’s social and biological relationships to conspecifics and members of other species. These core affective comportments are seeking/expectancy, rage/anger, fear, nurturance/sexuality, social bonding/separation distress, and play/joy. Panksepp hypothesizes that each of these affective comportments is subserved by its own core neural network in the midbrain-diencephalon. Each network has certain key chemical neuromodulators, and all project to an area called the periaquaductal gray (PAG). Panksepp proposes that the PAG serves as the substrate for a primordial sense of self, again probably common to all mammals, because in this area ‘there is a massive convergence of a diversity of basic emotional systems (fundamental value schema), various simple sensory abilities (perceptual schema), and primitive but coherent response systems (action schema)’ (1998b, p. 568). -- "Affective neuroscience: basic emotional operating systems of the mammalian brain,” Jaak Panksepp (1998a,b)
"Clicker training has not taught me a whole bunch, other than that people can get wrapped up in fads and catch phrases." Bob Bailey
"If a lion could talk we would not be able to understand him." Wittegenstein. "If a lion could talk we would understand him perfectly, but we would learn very little about ordinary lions from him."Daniel C. Dennett
"Dogs don't care who's alpha and who's not. Only emotionally dysfunctional owners and trainers do." Jack Field