On the world of Ruquelis, the horrific practise of cannibalism stands on three legs, however, this question is concerned only with Leg #1: Genetics.

In How to have 'easy' sexual morphs in individuals with two identical sex chromosomes in species with XY or ZW sex determination?, I describe the genetics of lilim, and in my answer, I describe how the matter of the lilim not breeding true can be solved. The TL;DR is that in the lilim mother, nondisjunction of the sex chromosomes results in the offspring inheriting three sex chromosomes, and in the diploid offspring, the unnecessary maternal sex chromosome is then selectively destroyed before the first cell division, resulting in the desired inheritance pattern.

The problem I have is that by the time the person in my story who sees and can deal with this problem comes onto the scene, the lilim number in the billions across multiple worlds... however, time travel is available. The problem with time travel is that if a person was to go back in time (around 20,000 years, or approximately 1100 generations) and genetically engineer the problem away in the past, when lilim were few in number and restricted to a single world, it would invalidate the reason for the time travel in the first place, thus causing a temporal paradox... but that isn't the only way.

The proposed solution is therefore a little more complex:

  1. The genetic engineer goes back in time (20,000 years/1100 generations) and engineers in the solution to the small initial population of lilim... literally a few thousand people on a single world. However, the modification is present but not activated, resulting in no visible change.

  2. Back in the present, a relatively easy action is used to spread the 'activator', which causes the genetic modification (which is present in the majority of the population by this time) to be activated, thus avoiding the paradox.

The idea that I came up with involves the activation protein that causes the gene causing nondisjunction of L and X chromosomes to be expressed. During cell division, a nondisjunction activator protein (NAP) is produced. The new activator proteins fold themselves up... but fail to fold correctly, becoming NAP-a and therefore fail to activate the LX-nondisjunction gene, therefore not causing the required LX-nondisjunction.

Then, in the future, an alternate differently folded version of NAP is spread, NAP-b. The small NAP protein is heat-resistant, tiny and persistent, and can easily spread through the human body and get inside cells. The only effect of NAP-b is to cause newly-produced NAP proteins to fold into the alternate NAP-b form... and to be able to activate the LX-nondisjunction gene.

So, when the time comes to switch the lilim from producing 1:1:1 male, woman and lilim children to 1:1 male and lilim children, all that need be done is for NAP-b to be introduced into the environment, by adding it to food and water. It will get into people's bodies, will be passed down to children, and will cause any NAP they produce to fold into the useful NAP-b rather than the ineffective NAP-a. Once it is out there, there will be no getting rid of it.

NAP will also have no effect on men or women, because the NAP gene occurs only on the L chromosome which is unique to lilim... but with the NAP-b protein in their bodies, eating a person with NAP-b in their system would be enough to cause the switch to activate... not that Ruquelian cannibals eat only other people.

This is similar to what happens with prions, in that a differently folded protein causes correctly folded proteins to become incorrectly folded... except that in my proposed scenario, neither NAP-a nor NAP-b causes any symptoms other than the expression or non-expression of the nondisjunction gene in lilim and alters the folding pattern of the NAP gene product, which also occurrs only in lilim.

I anticipate that mutation may have occurred between the time at which the genetic engineering is done and the time at which NAP-b is spread and not all extant lilim will be affected by the introduction of NAP-b into the environment, however, enough will be.

Is this a feasible way to genetically engineer a population when it is small, but only activate the change when the population is large, without causing a temporal paradox by time-travelling to the necessary early point in time?


In this world, temporal paradoxes may occur, but result in temporal loops in which the paradox and non-paradox states play out alternately and repeatedly but not necessarily identically, until random chance invalidates the paradox-causing event, causing the loop to break. Each iteration of the loop is separate, may play out differently, and information is not (usually) retained between iterations.

Because those who can travel in time exist partly outside of time, the business of having to replay their parts if a time loop occurs is especially egregious, especially over great lengths of time.

So, those who can time-travel are very careful to try to avoid paradoxes entirely. Since they know that simply altering the lilim's genetics in the past to make them breed true in the past will invalidate the reason for wanting to travel to the past to do so, but by introducing the necessity for a future action to 'activate' the past change, the problem remains to be solved in the present until the now-easier future activation of the change takes place.

This question is concerned with the deliberate avoidance of paradox states as much as it is with the feasibility of the mechanism of activating the past change.

Edit 2:

Time travellers are concerned with creating paradoxes that require unlikely events to break out of them. They are not particularly concerned with paradoxes that required an unlikely event to break into them.

Consider: The genetic change is made and NAP-a spontaneously mutates into NAP-b in an unlikely event. NAP-b spreads through observed populations and invalidates the time travel event, and in the next iteration of the loop, the spontaneous change may not occur, thus breaking the loop. All is well as long as there is no NAP-b at the end of the loop. This is a self-correcting 'fail-safe' loop.

However, if a time traveller was to cause a paradox by making an immediately effective genetic change, it would invalidate the reason for the time-travel, causing the loop to recur until something happened to break the loop. That might be the time traveller deciding to put in an activation mechanism... or it might be anything else that would prevent the time travel. This is a 'catastrophe' loop that relies on an unlikely divergence to break it. Time travellers like to avoid these catastrophe loops, since they are personally involved and literally, often have random catastrophic consequences.

  • $\begingroup$ There are some unneccessary details in your already long question. Maybe you could reduce it to only the important parts. For example everything about reincarnation and that hialutabu drug seems to be completly irrelevant for the question. Also as this is about time travel you should mention how time travel in your world works regarding changes in the past and specifically paradoxes. $\endgroup$
    – datacube
    Sep 27, 2023 at 6:31
  • $\begingroup$ @datacube Done. $\endgroup$
    – Monty Wild
    Sep 27, 2023 at 7:00

1 Answer 1


With decent genetic modification, this is twice as complex as it needs to be.

I'd pick a genetically modifying virus over modifying an entire population, and then providing them with specific supplements to kick a new folding mechanism off. In fact, there's lots of tricks I'd pick first, because there's a lot that can go wrong.

For one - prion infection is incredibly inefficient. Lots of people got BSE, but only a small percentage of those who ate contaminated meat.

Two, an unused and non functional gene is a prime candidate for mutation. Nothing happens if you change it, so there's no selection pressure killing off those with altered copies

Three, the inverse of point 1 might happen - if a possible prion exists, your process might get kicked off early by the spontaneous formation of these prions

What would I do instead?

  1. A virus - this one isn't too tough. Just pick something nice and contagious.

  2. a molecular clock - ok, this is a bit more out there, but less so than the prions. You set up a sequence in such a way that it is inactive when it has a tail on the protein, which is part of the transcription sequence. You then set it up so it acts as a special type of transposon - each meiosis event it replicates exactly once, and loses one base pair while it does so. So, once a generation, the protein will get shorter. Set up the maths properly on generation times, and you have a countdown clock to your protein working. We might be worried about mutations, which is why we set it up as a transposon - hopefully having a number of copies will keep it safe.


  1. Looking again at the new requirements, this is a job for Crispr - you add both the gene to make the cas protein, and the sequence to make the guide RNA to the genome of the cell, under a region whose transcription is only triggered by a synthetic hormone. When you supply the hormone, the crispr mechanism activates and neatly inserts the gene to make the protein you'd like. It's then permanently in the genome of the organism. This is efficient, and, barring an ethics board, almost achievable with today's technology. No messing around with alternate protein folding, and triggerable with a substance that you could dump into the water supply. It's quick, compared to prions, which can take years to show up, and requires dosing the population with far less of the proposed substance
  • $\begingroup$ The trouble is, with a molecular clock, we're talking about maybe 2800 generations between the change being made and it activating. Over that number of generations, the generation time could vary, and you'd get long tails that you'd have to allow for. A virus is something that you can't guarantee that everyone will get, nor guarantee that it'll affect the germ line. As for protein infection, the fact that it is transmissable by means other than the initial infection is good, not bad. I have also allowed for a certain amount of mutation to be acceptable. $\endgroup$
    – Monty Wild
    Sep 27, 2023 at 7:12
  • $\begingroup$ Besides, a spontaneous mutation of NAP-a to NAP-b would itself cause temporal loop that would prefer a lack of mutation in order to allow the time-travel event to take place. $\endgroup$
    – Monty Wild
    Sep 27, 2023 at 7:15
  • $\begingroup$ However, despite these problems, this does provide food for thought, so +1 from me. $\endgroup$
    – Monty Wild
    Sep 27, 2023 at 7:17
  • $\begingroup$ @MontyWild - what I meant is "eating the protein is not guaranteed to give you the prion - your gut does a great job of breaking down proteins, it's a relatively unlikely event that a successful prion infection happens from a food source $\endgroup$
    – lupe
    Sep 27, 2023 at 8:28
  • $\begingroup$ NAP is supposed to be small enough to get directly into cells. It could get into your system directly through mucous membranes. Gut absorption not required. $\endgroup$
    – Monty Wild
    Sep 27, 2023 at 8:33

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .