This question of sentience versus ‘instinct’ (more on that in a minute) is at the core of the problem both from an ethical and behaviorial standpoint, and the more we learn about neurophysiology across the entire kingdom of Animalia the more it is obvious that there is no clean dividing line. Any creature with complex learned behaviors (which includes entire classes of vertebrates and in vertebrates) has some apparent level of ‘self’, at least in the strict survival sense, any many animals clearly have fairly complex internal lives and needs which satisfy at least the first three tiers of Maslow’s “hierarchy of needs”, indicating a degree of sapience.
Whether a creature ‘feels’ pain and has the emotional response of fear or anxiety as a result is a question difficult to answer, but pain due to physiological stress or injury is not a conscious behavior; even in human patients who are have been paralyzed or in a coma, the brain registers the pain impulse and emits neurotransmitters accordingly. Nociceptors, the nerve receptors that detect pain, have been found in all classes of animal with a central nervous system of any kind. A response to pain may or may not be conscious, but the stress response of experiencing pain is built into the basic functions of the neurological system at a fundamental level.
In regard to the question of whether response to a harmful impulse indicates a cognitive sensation of pain as opposed to an instinctual, non-stressing response, this is really begging the question. It isn’t as if ‘pain’ and the associated internal experience of fear suddenly emerged in primates, or mammals, or even vertebrates. It is an evolutionary solution to equipping creatures with central nervous systems and any ability to make decisions over their behavior with a means to avoid potentially harmful circumstances. If it were not stressing it would be of no utility, and we’d all be like jellyfish, happy to float around and absorb nutrients, and then be eaten or broken apart with no awareness as to hazard or volition to avoid death. The ability to sense ‘pain’ as a fundamental environmental stressor and the associated nervious response is essentially fundamental to behavior.
The example of a simple robot which is programmed to avoid light is not really pertinent to the discussion as it is clearly the Mechanial Turk of stimulus and response; it does not ‘feel’ anything because there is no more complex nervous system or analogue to serve as a substrate for cognition of any kind, and is just following a relatively simple and discrete instruction set with little or no heuristic capability. Even the most simple creature with a networked nervous system such as the nudibranch (inaptly but commonly referred to as a “sea slug” even though many are stunningly beautiful creatures with sophisticated adaptations) is vastly more complicated in neurophysiology and behavior than even our most sophisticated robots.
As for inaccurate assumptions or claims in cognitive neuroscience, it should be understood that the field is really a relatively recent one in comparison to many sciences, including advanced sciences such as modern physics, molecular biology, or computer science. The fundamental tools to start working with even individual neurons or small networks at a functional level (e.g. on living organisms) only came about in the early ‘Seventies, and being able to actually assess brain states with any fidelity beyond measuring gross neural oscillations with the advent of PET and functional MRI scanning, and many knowledgeable people within the congitive neuroscience community believe that even this degree of resolution is insufficient to observe individual processes of cognition, and indeed, whether it is possible to identify discrete cognitive activity given that all behaviors and reported thought processes seem to activate most or all areas of the brain to a significant degree.
There are a lot of assumptions and commonly repeated factoids in neurophysiology and cognitive neuroscience that barely rise to the level of urban myth despite the fact that they are commonly held by experts in the field as discussed by Dr. Suzana Herculano-Houzel in this excellent Brain Science Podcast episode, which demonstrates just how complicated and disjointed the discipline of neuroscience is. (It’s still better grounded in demonstratable fact than psychology, though.) Anything you hear, even from a leading expert in the field, should be treated as informed speculation rather than hard fact at any level above basic physiology of neurons, and even that is subject to constant revision as we learn of new ways neurons adapt and function.
Stranger