Joseph LeDoux has an article at Nautilus on The Tricky Problem with Other Minds. It’s an excerpt from his new book, which I’m currently reading. For an idea of the main thesis:
The fact that animals can only respond nonverbally means there is no contrasting class of response that can be used to distinguish conscious from non-conscious processes. Elegant studies show that findings based on non-verbal responses in research on episodic memory, mental time travel, theory of mind, and subjective self-awareness in animals typically do not qualify as compelling evidence for conscious control of behavior. Such results are better accounted in “leaner” terms; that is, by non-conscious control processes.22 This does not mean that the animals lacked conscious awareness. It simply means that the results of the studies in question do not support the involvement of consciousness in the control of the behavior tested.
LeDoux makes an important point. We have to be very careful when observing the behavior of non-human animals. It’s very easy to see behavior similar to that of humans, and then assume that the same conscious states humans have with that behavior also apply to the animal.
On the other hand, and I’m saying this as someone who hasn’t yet finished his book, I think LeDoux might downplay the results in animal research a bit too much. It does seem possible to identify behavior in humans that requires consciousness, such as dealing with novel situations, making value trade-off decisions, or overriding impulses, and then deduce that the equivalent behavior in animals also requires it.
But LeDoux makes an excellent point. There are wide variances in what we can mean by the word “consciousness”. In particular, he discusses a distinction between noetic and autonoetic consciousness. Noetic consciousness appears to be consciousness of the environment and of one’s body. Autonoetic appears to be consciousness of one’s mental thoughts. He describes the autonoetic variety as providing the capability of mental time travel.
I’m not sure about this distinction, but in many ways it seems similar to the distinction between primary consciousness and metacognitive self awareness. This always brings to mind a hierarchy I use to think about the various capabilities and stages:
- Reflexes: fixed action patterns, automatic responses to stimuli.
- Perceptions: predictive models of the environment built with sensory input, expanding the scope of what the reflexes can react to.
- Attention: prioritization of what the reflexes react to, including bottom up attention: reflexive prioritization, and top down attention: prioritization from the next layer.
- Imagination / sentience: sensory action scenarios to resolve conflicts among reflexes, resulting in some being allowed and others inhibited. It is here where reflexes become feelings, dispositions to act rather than automatic action.
- Metacognition: awareness and assessment of one’s own cognition, enabling metacognitive self awareness, symbolic thought such as language, and human level intelligence.
Noetic consciousness (if I’m understanding the term correctly) would seem to require 1-4. Autonoetic might only come around with 5. Although 4 enables the mental time travel LeDoux discusses, so this match may not be a clean one. If I end up buying into the noetic vs autonoetic distinction, I might have to split 4 up.
But LeDoux’s point is that behavior precedes consciousness, and that’s easy to see using the hierarchy. Unicellular organisms are able to engage in approach and avoidance behavior with only 1, reflexes. The others only come much later in evolution. It’s easy to see behavior driven by the lower levels and project the full hierarchy on them, because it’s what we have.
All of which is to say, that I think LeDoux is right that arguing about whether animals or conscious or not, as though consciousness is something they either have or don’t have, isn’t meaningful. The real question is how conscious are they and what the nature of that consciousness is.
It’s natural to assume it’s the same as ours. It’s part of the built in empathetic machinery we have as a social species. But just because anthropomorphism is natural, doesn’t mean it’s right. Science demands we be more skeptical.
Unless of course I’m missing something?