Did Roth miss the Epigenetic Boat?

This post is a follow-up to Chana McCarthy’s keen insights in her guest post.  Allegiant was not the book I expected; and I must admit it is not love at first sight. But, having read the bombshell ending yesterday, and with a little recovery time I am liking it more and more. I am eager to go back for a re-read, so I can’t hate it too much.  I’ll come back with a post about what I did like later.


But, like Chana, I was most disappointed in the science of it, particularly when the psychology and neuroscience were so well handled in Divergent and Insurgent. The “genetic engineering study gone bad” is simply not believable on any level, even from the basic test they do in the lab. Helllooooo? If you need to do a genetic test on someone, it is far easier to take a cotton swab and scrape out some cheek cells than it is to inject your subject with some sort of micro-computer-packed serum.


I’m quite sure that the technology of basic cheek cell harvesting, of the type practiced in 9th grade biology classes all over, is not going to be lost over the next 300 years, no matter how many Purity Wars we have. The compound scientists may be reasonably well off financially, but there is no need for them to waste resources on a complicated serum when a Q-tip will do.

Second, if you have the technology to go in and “knock out” certain genes (for cowardice, low intelligence or whatever) it stands to reason that you would have the technology to reinsert the original sequences. It’s done with mice all the time. A basic understanding of DNA replication should make it clear that there is no reasonable mechanism for “healing” genes over time… the only way to “heal” a mutation is to rewrite the DNA sequence to get the original gene back: not something to be done one step at a time, over generations.  In fact, assuming the “experiments” started out with a population of diverse broken genes (some intelligence, some courage, etc) the last thing you want to do is isolate them and let them interbreed with each other for generations.  You are just as likely to wind up with people carrying multiple mutations that they inherited from different parents as you are to see people “healed.”  There is only one good reason for isolating genetic undesirables together: to make it easier to exterminate them all.

Interestingly, there is one hot new area of science that could have been used to make a bit more sense out of this storyline: epigenetics. Which is, in a nutshell, modifying not the DNA itself, but how it is packaged, to make certain genes more or less expressed or silenced entirely.

Let’s say, hypothetically that the earlier generation had identified certain “socially undesirable” genes such as for low intelligence– the “murder gene” David describes as the first discovered is based on fact, for instance. (It’s called MAO-a, and there are some fascinating science and ethics issues there; again, I am mystified that Ms. Roth got this right and so much else wrong). Let’s further assume the ancestors figured out not how to erase or mutate this gene, but how to  selectively methylate the DNA sequence or wind it around a histone so tight it could not be expressed, and whoops, turns out that was a really bad idea. It is far more feasible that some method for undoing that epigenetic effect might be developed that would require multiple generations to reverse the damage, because this would be more of an “uncovering” of the original good gene than the reversal of a mutation already permanently encoded into the DNA.

Furthermore, epigenetic processes are known to be influenced by the environment. (If you have 13 minutes, click on the link to see a wonderful NOVA clip explaining this). It is in fact possible that the strict environmental controls–different diets, education, training–set up in the Faction system might speed the reversal process along, as will as keep the population sufficiently under control to deter violence.

Finally, the Choosing/Initiation process would serve as a kind of artificial selection process to pull a higher number of Divergent out into the Factionless population, where they would breed more Divergent. Hence, the higher number of Divergent amongst the Factionless.

Epigenetics works so well, in fact, that I wonder if that wasn’t Ms. Roth’s original plan, and if she was possibly talked out of it by an editor who thought explaining it would be too complicated. If so, it was the wrong move. Like mirror neurons, epigenetics is a topic that is working its way into psychology textbooks. I checked and it is in the newest edition of the texts I use for both the introductory and the biopsychology courses I teach, though not in the previous editions, published in 2009-10.  A skilled writer like Ms. Roth could have pulled it off.


  1. Thank you for articulating some of the logical problems with this book. The genetically pure/genetically damaged (GP/GD) conflict didn’t work for me. I didn’t understand why they used a behavioral experiment at the choosing ceremony to select factions which are based on personality if the differences between factions were purely genetic (which doesn’t seem possible). Why do the Amity need peace serum if they were genetically selected for Amity?

    You are correct that the experiment ended up forcing divergent out into the factionless population who bred more factionless…but I wonder how successful that breeding experiment would be…for example, I would think the Amity would have the highest parity rate of all the factions. The founder effect argues against doing a human genetic/breeding experiment this way (small closed population with low genetic variability).

  2. Louise — this is terrific! I have not kept up with the research in epigenetics, so thanks for the link to on the Nova piece. I can readily imagine ways the series might have developed had either epigenetics or mirror neuron theory been the world-building premise.

    Michael Crichton, for one, demonstrated time and again that complicated, well-explained science can be the foundation for best selling science fiction — there really is no need to ‘dumb it down’. Given her promising start in Divergent, I can only assume that Roth was rushed to get the sequels out and conclude that her editors and publisher did her a massive disservice by failing to encourage more detailed planning and research.

  3. Ugh. this is by far my biggest problem with the book. I actually stopped reading and complained to my husband for a good 10 minutes after I got to the reason for the experiments. My husband asked me if I could trust the Bureau, and I think that’s beside the point. I don’t think David was lying about that particular aspect. So my problem is this:

    We have a bunch of mutated genes. We tried to make people brave, selfless, etc., by changing the genes. This ended up with a bunch of unintended consequences. To fix it, we’re going to isolate people by their mutated genes, because that TOTALLY won’t make the consequences worse. Nope. The Dauntless don’t get more ruthless or anything, obviously.

    I mean, this would be okay if they said Chicago was a failing experiment. But how the heck was separating people by their genes successful? I honestly do not get it. At all.

    (On a different note, I’ve decided I like the ending.)

  4. Despite the overly plentiful use of four letter words, I cannot resist sharing a link to Andi’s review of Allegiant at Goodreads — it’s scathing and hilarious. The genetic experiment comments are illustrated with a priceless cobweb-covered skeleton at a computer. I really did laugh out loud.


  5. Louise Freeman says

    It really could be made much better with a very simple edit:

    David: Eight generations ago, our ancestors discovered they could supress genes for cowardice, stupidity, dishonesty, selfishness and aggression by epigenetic mechanisms.

    Tris: What is epigenetics? Do you mean we are some sort of mutants?

    David: It is different from mutating the gene; it means to supress its expression by attaching many methyl groups to the DNA. Unfortunately, repressing the genes just made everything worse, and we found it is a lot harder to remove the suppression than to impose it, so the problem would take many generations to fix. So, we set up the people with repressed genes in experimental cities, where we would pipe different demethylating agents into the water, to try to determine which worked best. In cities such as yours, we also tried to stabilize the population by grouping people with similar repressed genes into factions, where we assumed their common personality factors would make them get along better. As an added bonus, epigenetic mechanisms could also be heavily influenced by environment, so we hypothesized that different diets, and different training of adolescents might speed the recovery process.

    Tris: Oh, I can see how that makes sense.

    …And we’re done….

  6. Thank you for writing this post. Epigenetics was such a cool topic in grad school to learn about and we are going to hear more about in the general press as more things about it are discovered.
    My wife is reading these books for the first time and I am rereading them so we can discuss anything she wants and with my background in molecular biology/genetics, the genetics info dumps in Allegiant is giving me a headache right now. I needed a palate cleanser before going back to the book.

  7. Louise M. Freeman says

    Thank you for responding! I’ve done some more work on this in preparation for some classes I’ve taught and it keeps getting worse. For one thing, assume Tris is descended from some ancestor who agreed to have her “stupidity” gene knocked out, and thus became Erudite (with its associated pathological lack of humility). If Tris is “healed”, she should have gone back to being stupid. But she’s not, since she has aptitude for Erudite. By this mechanism, a Divergent is not someone with healed genes, but someone who has accumulated multiple broken ones! More damage, not less.

    Epigenetics helps the multigenerational healing process, and would allow very interesting nature/nurture interactions in the different experimental cities. But it doesn’t help the underlying genetic illogic.

    Unfortunately for my epigenetic hope, as far as I know, no epigenetic effects have ever been shown to last more than three generations in humans, either.

    Roth, unfortunately, should have stuck with the personality psychology and the mirror neurons.

Speak Your Mind