How it would feel to be a rat under remote control
A granular analysis of an abnormal situation
Over the last decade, a handful of lab experiments have shown that it’s possible to control a rat’s actions through external electrical stimulation. These unsettling demos, whose victims are sometimes referred to as “ratbots”, suggest the possibility of a future use on humans, raising troubling questions of free agency and self-control. They also lead one to wonder what it would feel like to be the rat while it’s being remote-controlled. Does the rat realise something is wrong? Or does it still feel like its actions are proceeding as it intended?
Although most people don’t believe that rats have anything resembling a full human consciousness — which is good for the experimenters’ ethical reviews — let’s imagine for a moment that the rat in question did possess the equivalent of a human mind and consciousness. Better still, imagine you were the rat. What, if anything, would make you notice that something was off?
The details of the control mechanism are, of course, relevant to answering the question. In the above experiment, technicians controlled the rat’s actions in the most straightforward manner, via one of the rat’s motor pathways (the nigrostriatal pathway). This is roughly similar to triggering a muscle reflex which automatically moves your body without you intending it. The effect on the rat’s thoughts would be subtle and indirect. Other experiments used direct cerebral stimulation to control the rat, by immediately eliciting sensations and rewards in a manner similar to classical operant conditioning.
In this post we’ll focus on the first experiment where the rat’s actions are being directly controlled, because it turns out to be the more interesting one. In this setup, you (i.e. the rat) would sometimes intend one action, but then see a different one occur. The key question is: how would your thoughts be affected by your actions when your actions are no longer caused by your thoughts?
The question of how thoughts interact with actions has been a problem of philosophy since the time of the 17th century dualists, who puzzled over the invisible connection between the “immaterial soul” and the physical world. Modern materialists (and most neuroscientists) don’t make as big a deal of that particular question. They assume that both thoughts and actions are a type of general brain activity. Nevertheless, they usually leave the details of that mechanism to one side. In this post, however, the question can’t be sidestepped since that is the very connection being interfered with.
To explain why, let’s start by looking at the simplest case, in which external senses directly cause your actions, with no thinking involved. Such actions would play out like automatic, well-worn motor habits. Were an experimenter to interfere in such processes, and cause you to perform a different action, you would have no immediate reason to notice any alteration; that is, until you realised that you were not in the situation in which you expected to be. Since the cause of the action and the action itself are both “invisible” to the conscious mind, it is not the incorrect action but rather the unexpected feedback from that action that tells you something is wrong — e.g. you see your arm move in a direction you hadn’t intended.
However, it would be unreasonable to expect that your predicted outcome and reality were ever going to match up perfectly. As you walk towards a door or piece of cheese, any number of people might pass by your peripheral vision, and you may hear incidental noises. Yet these don’t usually cause you any alarm. So the fact that things didn’t turn out exactly as expected is not by itself enough to cause you to take note.
In addition, you can imagine a hypothetical scenario where some alien force, invisible to you, was slightly adjusting all your actions so they land more accurately on their mark. Whereas previously you had to make small readjustments, now you always grasp the doorknob or cup you reach for on the first try. Would you notice that something was wrong? Not likely. Even if you noticed an overall change in your skills, it’s more likely you would attribute it to a natural, mature level of physical talent. (This scenario isn’t even hypothetical: it describes a real technology intended to help people with Parkinsons control their tremors.)
So what would have to change for you to notice? Let’s go back to the example of the ratbot above. Say that you intended to go towards a door on your left, but were forced via external control to go to the right. It’s fair to say you would quickly notice that your expectations weren’t being fulfilled. But why would you notice this, when you didn’t notice in the above two examples? To answer this question, we must first ask why it was that you wanted to go to the door at all.
Walking towards a door is a learned action. So we can assume that you were trying to accomplish something by carrying it out. Perhaps you were hungry and it was a pantry door. Or you were walking briskly to class because you didn’t want to be scolded for being late. Missing the door causes anxiety because the action had an intention behind it: to achieve a particular goal. Not accomplishing it means you would fail to satisfy the reason behind the intention.
The mismatch between expectation and reality is only a problem insofar as you cannot proceed with your plans. This also means that your expectations only cover as much as you need to succeed; they are not meant to be a complete model of reality — otherwise slight variations in circumstances would through the system off-kilter. This is why the two cases above — adding noise, and generating perfect actions — don’t register as problems. Additional noise doesn’t detract from the fact that the subset of desired sights and sounds are still present. And perfectly carrying out the action is, if anything, a boon. Failure to meet an expectation is only a disappointment if it undermines some need. The internal expectation of success, much like the feeling of hope, had temporarily suppressed the underlying tension. So when you fail to meet those expectations, the tension flares up again.
This is where conscious awareness first enters the picture. Awareness is not continual; it is a moment-to-moment notice. Most activities can be carried out without explicit awareness of your actions. To even notice that something is wrong you must first have a reason to notice. On failing to meet expectations your mind now triggers the underlying tension, and pays attention to the things around you. But this is not a “general” alertness; it is searching for a resolution to replace the one that was cut short. In the same way that the desire to find your keys makes you notice shiny objects around you, the tension that motivated you to go to the door will now look for something that fits its “success criteria”, to close the gap between where you are and what you’re trying to accomplish.
The first thing you’ll notice, then, is where you want to be, but aren’t — namely the door. This happens so quickly you won’t even realise it’s happening. It would feel more like suddenly being alert after a stretch of complacency. You might interpret it as re-orientating towards your goal. This initial, sub-second reaction scans your environment and re-registers your intention, i.e. “I’m paying attention now, and that is where I want to be”.
What you ultimately become aware of is not random, nor is it an impassive recording of all experiences. You are paying attention because of a specific tension; so what you are looking for is also specific. If you felt hungry, you would look for and notice food, or something you know leads to food, such as the pantry door. The outcome of such awareness is the formation of a new memory, a mental image of what you projected to be useful, which can later be recollected as a thought later, when it may become useful again.
Of course, simply noticing where you want to be is not enough to ameliorate your situation. It gives only a preliminary awareness of your plight. As your mind notices that your body is moving inexorably away from you goal, you register another, more urgent problem. For the first time you are now engaging with the actual problem of being remote-controlled. From your perspective, it seems to you like you are walking towards a door which keeps moving away from you — even though it is actually you that is moving away from the door. This is a different problem than the previous one, when you were merely trying to get to your goal.
The fact that this new situation represents an emergency is something you must have learned beforehand from a similar experience. Your mind wouldn’t naturally know to be concerned about it. For example, you may have learned it as a child who, on being dragged away from a toy store, discovered they could no longer access their desired toys. Now, like the same child trying to escape a parent’s grasp, you will try to escape the grip of the unseen, unknown pull on your actions.
Perhaps some external force, like a rope or foreign hands, is pulling at your legs. Such a thought might arise, again, as a memory from past experiences when that might actually have been the case. At that time, you may have looked at your legs and spotted the entanglement, which helped you extricate yourself. You became aware of it — i.e. the memory of that sight was recorded — since, like the pantry door above, seeing the entangling rope was a useful step towards a resolution. Now the same memory could address your current predicament, so you recall it and register it as a useful “explanation”. “Explanation” here means a thought that can help you resolve the underlying issue. Recollecting the memory of entangled feet can be useful as it could trigger the right actions that resolve the issue.
This, however, raises an obvious question — how can a mere thought or recollection of entangled feet cause a useful action now?
The same question that preoccupied the early dualists is before us again, namely: what is the connection between thoughts and actions?¹ We, perhaps, can take a hint about this matter from 19th century philosopher Henri Bergson:
If there be memory, that is, the survival of past images, these images must constantly mingle with our perception of the present, and may even take its place. For if they have survived, it is with a view to utility; at every moment they complete our present experience, enriching it with experience already acquired …
…any memory-image that is capable of interpreting our actual perception inserts itself so thoroughly into it that we are no longer able to discern what is perception and what is memory.
— Matter and Memory
What Bergson was foreshadowing is a model of thoughts as self-generated stimuli. Recall from the discussion above that the mind only expects an sensory experience to happen if it is a useful step towards its present aims; and it excludes the rest as noise. We also showed that the things you notice and later recollect as thoughts are experiences that were useful precursors to beneficial actions. The parallel between these two should be obvious; they both capture those stimuli that presage the accomplishment of your aims. The only difference is that the first includes actions and the latter is a precursor to actions. Thus thoughts can be viewed as expectations which, in the absence of intervening actions, are experienced as self-generated stimuli. In other words: you imagine what it is useful for you to expect².
The benefit of this hypothesis, that thoughts are self-generated stimuli, is that it helps us explain how thoughts can cause actions. The same external experiences that would cause you to disentangle your legs when experienced from the outside, would also cause those same actions when elicited as a thought. This is what helps you transfer what was learned from the first situation to the second. You merely recreate the prior experience of seeing your entangled legs as a thought, which then causes you to act as though those experiences were directly visible, even though they aren’t. This is similar to reaching for a light switch in the dark; once you’ve learned to expect the light switch to be in a certain place (the useful expectation), you can later visualize where it is in the dark and use that image to guide your actions as effectively as if you could actually see it.
Though this sounds promising, you, the rat, are still smack in the middle of a complete breakdown of alignment between thoughts and actions. So you would find that imagining historical solutions also fails to have an effect. As you run out of ideas for how to achieve your goal, you may temporarily stop trying to move. Then you notice your legs are still moving. The world suddenly seems dreamlike. A new fear of your own uncontrolled body, acting in ways you didn’t intend, triggers a new tension.
Now, rather than struggling to reach the door as before, your mind will try to find a way to prevent this uncanny movement. You may resort to explanatory thoughts such as a seizure, or in the extreme, madness. Again, each of these thoughts is a useful set of images and words you bring to the situation from prior learning — none of this is innate. And each resolves the underlying problem by giving you a possible path for action. So if you imagine that you are having a muscular seizure, you might try to make slower, more methodical movements.
Ultimately, where your mind and sanity end up depends on what you’ve learned before the ordeal, including what should or shouldn’t signal an emergency. The knowledge you bring to the situation will determine how you frame it. In many ways, it is your own past, your own life’s learning that is haunting you and causing you to feel alarmed, not the situation itself.
Over time, your mind might learn to adapt to the new order of things, like a prisoner adapting to his daily routines. In fact, it is not guaranteed that a person under remote-control would necessarily be horrified by their predicament. A child who was continuously controlled from birth would have no reason to be alarmed. He would watch his own life pass by like a movie or a dream; or like a roller coaster which can hurt him, but whose course he cannot change. What he notices and remembers, his whole conscious awareness of his life, might appear surreal to you or me. But insofar as he is still human his thoughts would always be based on those needs that he developed within his own, admittedly alien situation.
(I promise the next post will end on a lighter note; but given the initial starting point this one was hardly going to conclude cheerfully.)
¹ I find it curious that the dualists were so concerned with the mystery of how thoughts cause actions, that they apparently felt confident they knew how thoughts caused other thoughts. Why? Both connections are equally invisible and opaque.
² It is no accident that the same cloud shape is used in the diagrams above to represent (1) sensory experiences, (2) expectations, and (3) thoughts. All three are facets of the same phenomenon.
