I originally posted this on Reddit AskScience, but it didn’t garner a single comment, much less an upvote. I thought it was a very interesting question, so maybe the SDMB will be more responsive.
I’ve been battling a nasty respiratory virus for the past few days, and I noticed something. My throat is so swollen and sore that I can’t cough up any dead viral phlegm, but the existence of the phlegm in my airway is causing me post-nasal drip (I think), resulting in my throat remaining sore and tight.
This loop has since been broken, but how? With billions of stimulus-response interactions going on in the body at any given time, do any ever get caught in an infinite loop, and what breaks the loop?
Animals have a fundamental limitation that each action costs energy so even if they get into a “loop” they aren’t really in the same state at the end as they were at the start, because now they have less energy. If they get energy from the environment by eating, now they are in the same state but their environment has changed because there’s less food in it.
Computer programs face the same fundamental limitation, but for the kind of time intervals in question the amount of energy available is effectively infinite, so they can be thought of as having “infinite loops”.
It is hypothesized that the Spanish flu killed via an infinite loop of cytokine creation where creation of cytokine triggered the creation of more cytokine, which of course triggered the creation of more cytokine… In computer science parlance, it was a type of fork bomb that performed a Denial of Service (DoS) attack against the body.
And an organism has a lot more inputs into the system than a computer program. A program in an infinite loop looks at one variable, if the value of the variable is within a certain limit it goes through the loop and looks at the exact same variable again. If nothing can happen in that loop the change the value of the variable, the program will loop “forever”.
But of course it won’t really loop forever. Eventually somebody is going to hit Ctrl-Alt-Del, or there will be a power outage, or the sun will expand to a red giant and boil the oceans.
Animals that get stuck in a behavioral loop repeating the same behavior over and over again will eventually die. There’s your halting state. This really does happen, lots of animals have very stereotyped behaviors and if the stimulus remains the same they will repeat the same behavior over and over. A famous example is when a cuckoo lays an egg in another bird’s nest. The cuckoo chick hatches and murders the other babies in the nest, while the parents stuff bugs and worms into the murderous chick’s mouth, unaware that anything is wrong.
But eventually that cuckoo chick will either die, or grow up and fly out of the nest, and the repeated behavioral loop stops.
Or take your coughing example. Some diseases cause coughing so bad that the victim can break ribs. The feedback cycle gets worse and worse, with the coughing intensifying the symptoms and leading to more coughing, and eventually you cough yourself to death. End of cycle.
Right. But the important question is whether the loop ends with the death of or the recovery of the organism. If each “cycle” of the loop drains your nutrient supply, then you may starve before some other loop-ending condition materializes.
A little less dramatically, diseases of chronic inflammation can also involve positive feedback: some damaged tissue releases inflammatory cytokines, which recruits immune cells which continue the process. In a normal, healthy inflammation process, the tissue and immune system eventually switches to a healing response.
Very generally speaking, cell signaling pathways all have many possible layers of negative feedback. Signaling proteins are degraded after they bind to their receptors, activated receptors are pulled from the cell surface into the cell, signal transduction in the cell also activates pathways which deactivate the receptor, etc. Some receptors are essentially single-use, and the process of activation removes the intracellular “business end” from the extracellular portions that bind to signaling proteins.
I could go on about any of those examples and more if you have specific questions…
Another way of looking at “what prevents …” is more evolutionary.
Critters that had feedback loops that did get stuck tended to die early. e.g. The ones whose digestive enzymes ate through their proto-stomachs after mis-recognizing the damaged lining as just more food to digest have long since failed to reproduce.
So we’re left today with critters that mostly are designed to never get stuck in loops. At least for most of the situations most of the time, even when battling infectious invaders.
And on those rare occasions they do get stuck in loops, the situation slowly changes enough they wander back out of the loop. Usually. And for those who do get really, really stuck in loops, they die.