Determined, page 26
Thus, we have a hierarchy. For a single shock, you add more copies of some molecule that already exists; for four shocks, you generate something novel to interact with that molecule that already exists; for a massive cluster of shocks, you start a whole construction project. All very logical. And this is precisely what Kandel showed as well (taken from that same Nobel Prize talk):
What exactly did he show goes on in an Aplysia SN when this is happening? Just skim this paragraph, and don’t memorize a word of it. In fact, probably don’t even read it—just know how to find it again later. The details: (A) What’s actually happening in step #1? The neurotransmitter 5HT triggers the release of cAMP, which activates previously inactivated PKA, which works on the K+ channel to trigger an influx of Ca2+ through Ca2+ channels, which results in the release of a greater amount of neurotransmitter. (B) Step #2: Enough cAMP has poured in not just to activate step #1 but also to spill over and cause MAPK to cleave CREB-2 from CREB-1, freeing the latter to dimerize into pairs of CREB-1, which interacts with the CRE promoter, which turns on an early-phase gene that leads to the synthesis of the enzyme ubiquitin hydrolase, which stabilizes PKA, allowing it to have its effects longer. (C) Step #3: The influx of cAMP is large enough that not only are steps #1 and #2 activated, but #3 is as well; this leads to liberation and dimerization of enough CREB-1 to not only activate the ubiquitin hydrolase gene, but the C/EBP gene as well; C/EBP proteins then activate an array of late-response genes whose protein products collectively construct a second synaptic branch.[*]
Almost half a century of work by Kandel, his students and collaborators, eventually a whole field of neuroscientists building on those findings, all to answer a just-so question: Why did the traumatized Aplysia retract its gill for so long? We have built a machine on both the level of neurons communicating with each other in a circuit and the level of chemical changes inside a single key neuron. This is a machine that is entirely mechanistic in biological terms and that changes adaptively in response to a changing environment; it has even been used as a model by roboticists. I dare anyone to invoke the concept of free will in making sense of this Aplysia’s behavior. No Aplysia, encountering another one, would say, “It’s been a tough season, thanks for asking, lots and lots of shocks, no idea why. I had to build new synapses on every neuron in my siphon. I guess my gill is safe now, but I sure don’t feel safe. This has been hell on my partner.” We’re watching a machine that did not choose to change its behavior; its behavior was changed by circumstances via logical, highly evolved pathways.[2]
And why is this the most gorgeous piece of neurobiological insight ever? Because pretty much the same thing goes on in us when we have become the sort of person who would pull a trigger, or run into a burning building to save a child, or steal an extra cookie, or advocate hard incompatibilism in a book destined to be read only by two people who will hate it. The circuits and molecules of the Aplysia are all the building blocks we need to make sense of behavioral change in us.
Which, no doubt, seems absurd, totally implausible, leaping from Aplysia to us. Thus, we’re going to get there with a few in-between examples (but in less agonizing detail than has gone into understanding Aplysia behavioral machinery). When we’re done, the hard reality is that we are unimaginably more complex than an Aplysia but are biological machines with the same building blocks and the same mechanisms of change.
Aplysia californica. As should be obvious, the one on the left is happy, in an unreflective kind of way. The one on the right is a wonderful Aplysia stuffie that could be your child’s comfort object all the way until their freshman year of college.
Detecting a Coincidence
Our next neuronal machine blinks its eye. Go up to it, spritz a little puff of air at its eyelid, and the eyelid blinks automatically, as a protective reflex. We already know the simple circuitry needed to pull this off. There’s a sensory neuron that has an action potential in response to an air puff. This then triggers an action potential in a motor neuron, causing the eyelid to blink (see the following page).
Now let’s add a totally useless additional piece of circuitry. We have a second sensory neuron. This one doesn’t respond to the tactile stimulation of an air puff. Instead, it responds to an auditory stimulus, a tone. Neuron 3 projects to the blink motor neuron, where it isn’t excitatory enough to cause an action potential in neuron 2. Play the tone, and nothing happens in neuron 2:
Let’s make that side path even more ornately useless. Now the tone is played, activating neuron 3. As before, neuron 3 isn’t able to cause an action potential in neuron 2; however, it does so in neuron 4. But as it turns out, the neuron 4 action potential has only about half the excitatory power needed to evoke an action potential in neuron 5. So stimulate neuron 3 with a tone, and the net result is that nothing happens in either neuron 2 or neuron 5; a tone still does nothing to blinking:
Let’s add another useless projection to this circuit. Now neuron 1 sends a projection to neuron 5 (along with its usual projection to neuron 2). But when an air puff triggers an action potential in neuron 1, it gets only halfway to the excitation needed for neuron 5 to have an action potential. So: air puff, neuron 2 activates, nothing happens in neuron 5:
But now let’s activate neuron 1 and neuron 3. Play a tone and release an air puff. Crucially, the tone comes one second before the air puff, and it takes a second for any action potential to reach the axon terminals. So:
At time zero: play the tone, neuron 3 has an action potential.
After one second: neuron 4 has an action potential (thanks to neuron 3), while the air puff is now causing an action potential in neuron 1.
After two seconds: neuron 2 has an action potential (thanks to neuron 1), triggering an eye blink. Meanwhile, the action potentials from neurons 4 and 1 arrive at neuron 5. Again, neither of those two inputs is enough to trigger an action potential alone, but when they are combined, neuron 5 has an action potential. In other words, neuron 5 has an action potential if and only if the tone is played and is followed by an air puff one second later. The circuit allows neuron 5 to detect that the two stimuli coincided. Or to use the jargon of the field, neuron 5 is a coincidence detector.
After three seconds: neuron 5 has its action potential, causing it to stimulate the axon terminals of neuron 3. Which, as it turns out, accomplishes nothing—it isn’t strong enough to, say, cause those axon terminals to dump much neurotransmitter.
But play the tone followed by the air puff a second time. A tenth time, a hundredth time. Each time neuron 5 stimulates the axon terminals of neuron 3, it slowly causes neuron 3 to build up more neurotransmitter there, release more of it each time, until . . . finally . . . when neuron 3 is stimulated by the tone, it triggers an action potential in neuron 2. And the machine blinks before the air puff happens, blinks in anticipation of it (see figure on previous page).
It’s called eyeblink conditioning, and it works this way in mammals—in lab rats, rabbits, in humans. It’s useful, adaptive—it’s great to be conditioned to close your eyelids protectively before, rather than after, a noxious stimulus occurs. We know the underlying circuit in a different, famous setting, one that gives the phenomenon its name: Pavlovian conditioning. Ol’ Doc Pavlov lets his dog smell dinner; dog salivates. This is the circuit of neurons 1 and 2. Neuron 1 smells the food, neuron 2 stimulates salivary gland, dog drools. Thus we have an unconditioned stimulus (the smell), which automatically evokes the unconditioned response of salivating. Now ring the bell just before the food arrives; pair the two over and over and, thanks to neurons 1, 3, 4, and 5, you establish a conditioned stimulus and a conditioned response—ring the bell and the dog salivates in anticipation of the smell of the food.
The key spot where change happens is in the place where neuron 5 terminates on neuron 3. How does the former repeatedly stimulating the axon terminals of latter result in the latter increasing the amount of neurotransmitter released, eventually gaining the power to elicit an eyeblink on its own? Go back to page 277’s description of the inner workings of the SN in an Aplysia. How does eyeblink conditioning work? By neurotransmitter from neuron 5 releasing cAMP inside neuron 3, which frees PKA from its brake, which activates MAPK and CREB, which activates certain genes, culminating in, among other changes, the formation of new synapses.[*] This is not “Neuron 5 causes the intracellular release of chemicals that kind of function like stuff in Aplysia.” It’s the same chemical messengers.[3]
Think about this. Humans, being conditioned to blink their eyes, and marine sea slugs, conditioned to withdraw their gills, haven’t shared a common ancestor for more than half a billion years. And here we are, with their neurons and ours using the same intracellular machinery for changing in response to experience. You and an Aplysia could trade your cAMPs, PKAs, MAPKs, and so on, and things would work just fine in both of you.[*] And you’d both be using serotonin to kick-start this. These Aplysia/human similarities should demolish anyone’s skepticism about evolution.[4]
More important for our purposes, these findings show that (just as with the Aplysia gill-withdrawal reflex) we can build ourselves deterministic circuits with deterministic neurons that explain an adaptive change in human behavior in response to experience.[*] All without having to invoke the notion of our “choosing” to start blinking our eyelids when we hear a tone.[5]
Booyah, we’ve crushed any philosopher whose lifework is premised on the notion that we have free will because we can be eyeblink conditioned. Yeah, yeah, this is not a very fancy outpost of human behavior. Nevertheless, it’s fancier than you might think.
To appreciate that, what happens to lab rats if, when they were pups, they were intermittently separated from their mothers for a while? Rats that experienced such “maternal separation” early in life are, as adults, a mess. They are more anxious, show more of a glucocorticoid response to mild stress, don’t learn as well, are easier to addict to alcohol or cocaine. It is a model for how one type of early-life adversity in humans produces dysfunctional adults, and people know tons about how each of those changes comes about in the brain.[6]
So get this—take a rat pup and maternally separate him, and as an adult, it will be harder to do eyeblink conditioning on him. In other words, along with all the other deleterious consequences of maternal separation, you have animals that don’t acquire this adaptive response as readily. It is caused by an epigenetic change in the brain, such that forever after, there are elevated levels of receptors for glucocorticoid stress hormones in the equivalents of neuron 2. Block the effects of glucocorticoids in that adult rat, and eyeblink conditioning becomes normal.[*] Conclusion: early-life adversity impairs this circuit by making a key neuron in the circuit more sensitive to stress.[*],[7]
Take one lone heroic rat that, for some reason, can save the world from disaster by developing a conditioned eyeblink response. And he screws up, doesn’t do it, lets the world down. Afterward, everyone is pissed at the rat, blaming him for not conditioning. To which he can say, “It’s not my fault—I didn’t get conditioned because, one second before, my interpositus nucleus wasn’t as responsive to the conditioned stimulus; because a few hours before, my stress hormone levels were elevated, which guaranteed that the interpositus would be particularly resistant to conditioning; because back in my childhood, my mother was taken from me, and this changed gene regulation in the interpositus, permanently increasing levels of a hormone receptor there; because back millions of years ago, my species evolved to be highly dependent on maternal care after birth, and the genes needed to make lifelong changes in circuitry if the mother is absent.” A change in behavior due to specific changes that can be identified in a circuit, arising from circumstances a second, an hour, a lifetime, an evolutionary epoch earlier over which the organism had no control. No rodential moral responsibility involved, no grounds for everyone blaming the rat.
But still, this is just about blinking your eyes. On to the sorts of scenarios that this whole book is about.
When They Become Thems
Not many of the world’s problems arise from the fact that a neutral stimulus can be conditioned to evoke an eyeblink reflex. But a lot of them sure arise from the same going on in the amygdala.
Take a lab rat or a human volunteer and give them a shock. The amygdala activates; you can show this in the rat by recording the activity of neurons in the amygdala with electrodes, while in humans you show the same with brain imaging. To prepare us for the subtleties to come, right off the bat, the link between shock and amygdaloid activation is modulated in all sorts of interesting ways. For example, in both the rat and the human, the amygdala activates more if the shock occurs unpredictably, rather than if you know when the shock is coming.
Once the amygdala activates, it triggers a variety of responses. The sympathetic nervous system is activated, the heart beats faster, blood pressure rises. Glucocorticoids are secreted. Your typical rat or human freezes in place. Nontrivially, if that rat has a smaller and weaker rat next to them, the rat that has been shocked becomes more likely to bite the other—which lessens their own stress response.
So this is a version of an SN-MN circuit, by now familiar. Now, before each shock, play a tone as a conditioned stimulus. Do it a bunch of times and you know what happens—the tone itself will eventually have gained the power to activate the amygdala, and we have a conditioned fear response. Beautiful work by Joseph LeDoux of New York University has revealed the circuitry to explain this. Look at it closely and, what do you know, it’s the same basic wiring as for conditioning an eyeblink or a gill withdrawal. If timed right, information about the unconditioned stimulus (the shock), mediated by the somatosensory thalamus and cortex, and about the conditioned one (the tone), mediated by the auditory branch, just as with conditioned eye-blinking, simultaneously converge on the amygdala. Local neurons there act as coincidence detectors, repeated stimulation of the auditory branch induces all sorts of changes in the amygdala involving cAMP, PKA, CREB, all the usual, and a tone now elicits the same terror that a shock does.[8]
We saw that something as simple as eyeblink conditioning reflects a nervous system that has been sculpted by all that came before it (e.g., early maternal experience). The acquisition, consolidation, and extinction[*] of the conditioned fear of something neutral like a tone reflects the organism’s history even more. Extinction will occur faster if, in the seconds before, there are high levels in the amygdala of endocannabinoids (whose receptor also binds THC, the most active component of cannabis)—this makes it easier to stop being afraid of something. The amygdala becomes less likely to store away a conditioned fear response as a stable memory if, in the previous hours, the individual has taken an SSRI antidepressant like Prozac (which makes people ruminate less about negative thoughts). The amygdala will be less active and harder to condition if, in the days before, it was exposed to high circulating levels of oxytocin, which helps explain how oxytocin can promote trust. In contrast, if the organism has been exposed to high levels of stress hormones in the previous month, it becomes easier to generate a conditioned fear response (thanks to the hormones increasing activity of the gene that produces the mammalian version of C/EBP, which appears in the figure on page 277). And pushing way back in our “one second before, one minute before” arc, if an organism was exposed to lots of Mom’s alcohol back during fetal life, it has a harder time remembering a conditioned fear. And of course, what versions of the genes related to those in that figure are present, and whether the individual’s species evolved those genes in the first place, will influence how readily conditioning occurs. How easily an organism learns to be afraid of something as simple as a tone is the end product of all these influences on the workings of this circuit, all factors over which the individual had no control.[9]
All this for a tone.
Consider something else that activates your amygdala. In this case, hearing the word rapist. You’re not genetically programmed to activate your amygdala in response to it, not the way it would automatically activate if, say, you were dangled upside down by a thread up in the air while covered with spiders and snakes. Instead, the amygdala came to respond to the word through learning—you learn what the two syllables mean, what the act is; you’ve learned about its impact in general, how being raped, as it’s been said, is like living through your own murder; you know someone who was or, unbearably, you yourself were. In any case, you now have an amygdala that activates automatically in response to the word, as surely as if you were given a shock.
Now let’s take a neutral stimulus and rely upon the coincidence detectors in our amygdalas to generate a conditioned fear response. Something more complex than a bell that would make Pavlov’s dogs salivate or a tone that would cause a lab rat to freeze:
When Mexico sends its people, they’re not sending their best. They’re not sending you. They’re not sending you. They’re sending people that have lots of problems, and they’re bringing those problems with us. They’re bringing drugs. They’re bringing crime. They’re rapists.
—Donald Trump, in the speech that famously opened his presidential campaign, June 16, 2015
Students of history and current events: Let’s play a game called “Match the Conditioned and Unconditioned Stimulus.” Get them all right and you win a prize, so have fun!
Conditioned stimulus and the people who labored to generate that association
