So much bullshit has been written about epigenetics that if you’ve heard the word at all you’ve probably heard bullshit about what it is and does. Epigenetics is a real thing, observed and explained by real molecular biologists, and it is not the same thing described by Jean-Baptiste Larmarck nor does it have the same effect. Epigenetics was discovered in the context of tissue differentiation: how does one cell know that it’s supposed to become part of your liver, and another cell know that it’s supposed to become part of your lungs? This puzzle, tissue differentiation, is at the very heart of the development of higher lifeforms. It is why we can have specialized organs and why different parts of the body can have different functions. If our distant one-celled ancestors had not somehow gotten this right, we would all still be one-celled organisms. Or, at best, colonies of mold or bacteria.

The cells know these things because during the process of gestation the DNA of the cells gets a different “program” – that is, histones and methyl groups attach to the DNA allowing different parts of the DNA to be expressed or suppressed. The ‘germ line’ or pluripotent cell lacks these structures, and then as we develop, the cells in different places are subject to different stimuli that cause these things to bind to their DNA and, as a result, their epigenetic “program” is changed and different kinds of tissues grow.

In the context of cell differentiation, it was reasonable to conclude that the germ line lacks these structures, and indeed a cleansing mechanism was found that strips these structures from the genome.

But there was another phenomenon that happens, and eventually the explanation for that phenomenon was also found in epigenetics.

Here is an example of the other phenomenon that epigenetics explained. Some species of shellfish have shallow-water populations with smooth shells, and deep-water populations with rough-textured shells that have lots of corners and angles. And if you take some of the rough-textured individuals and put them in a shallow-water environment, they have rough-textured offspring for some number of generations and then magically their grandchildren or greatgrandchildren or something will be hatched with smooth shells. Likewise if you take some smooth-shelled individuals and put them in a deep-water environment. After a generation, or two, or three, rough-shelled offspring show up. And these populations have the same genome. There is not so much as a point mutation that separates them. So how’d that happen?

Well, that happened because that cleansing mechanism that strips the histones and methyl groups from the DNA of the germ line has a few that it only strips a small fraction of the time, and a few that it only strips about half the time, and so on. These points of configuration, in contrast to the others, are (usually) inherited by the offspring. There is a configuration for smooth shells, and a configuration for rough shells. The histone that chooses between them only gets stripped from the genome of the reproductive cell a small fraction of the time. The higher pressure conditions of the deep water environment are a stimulus that promotes the acquisition of that particular histone at a crucial point in development, making the deep-water population more statistically likely to acquire that histone than the DNA-cleansing mechanism is to remove it (some handwaving here – this is the simplified version).

The Saltation problem was also eventually explained by epigenetics.

Saltation addresses a phenomenon observed in the fossil record, and sometimes even over just a few lifetimes, which Darwin’s theory of gradual one-adaptation-at-a-time changes, as understood at the time, had no way of accounting for. Large changes in the body plan or adaptive range of a species are often observed to occur with shocking rapidity. Saltationism was advanced as a theory of evolution, and proposed that inheritance was somehow described fractally by the DNA, with mutations at a higher level allowing entire structures to be changed instantly, explaining how numerous changes each of which seems to require the others for utility can all happen together so rapidly when a species needs to adapt.

If you’ve been paying attention, you’ve already figured this out. Think back to the shellfish. Two populations with a different morphology, leaving behind markedly different fossils. If one abruptly changed to the other in the fossil record, it would be accounted as a saltation event. So all we need to imagine is a change in the environment that makes it very likely, or very unlikely, to acquire that histone, and suddenly one “version” of the species is gone. In fact species may retain the potential for a range of forms and configurations that have not actually been seen for thousands of years, ready to re-emerge when the environment changes. And a mutation that is highly favorable to one version or the other may render every mutation that leaves that version more likely to occur highly advantageous, until eventually traits are acquired which are not compatible with some of the other configurations.

So these Saltation events, as it turns out, are triggered by environmental changes that favor one configuration of the genome over another – without the need to make wholesale changes to the genome itself. Neat, huh?

And now we have to deal with the thing that most of the bullshit is about. It turns out that this epigenetic process does not stop when you’re born.

Here’s an example of the kind of phenomenon we’re talking about now. Some mice are taught to fear the scent of some chemical by repeatedly exposing them to the scent and then zapping the daylights out of them. I swear some of these experiments are done just to be mean to the mice. But that’s what they did, and then they checked to see if they’d scared the parents so hard that offspring they had later in life, who’d never been exposed to that scent themselves, had also learned to fear it. And it turns out that they had.

So think about that. What the hell is going on when an adult mouse learns to fear something, and then their pups who’ve never been exposed to it, are born knowing that they also fear it?

As it turns out, epigenetics is going on again. The adult mouse, when it learned how horrible that smell was because it always signalled an unavoidable, painful zap, was under a stimulus. And this stimulus, just as other stimuli had done while its body was developing in utero, promoted the acquisition of some histones or methyl groups attached to the DNA, specifically in the germ line. These were not cleared by the cleansing mechanism. And the mouse pups were born with an epigenetic configuration that somehow resulted in developed adults who “know” that smell is bad. So-called “genetic memory” like this can carry forward three, four, even five generations before the cleansing mechanism that runs when new gametes are produced finally happens to get rid of it.

So here is the nub of truth at the center of the bullshit. Some characteristics acquired during a mouse’s lifetime (fear of a smell) are passed on to their offspring. Some learning that has taken place in a mouse’s lifetime, benefits the offspring as well as the parent. The ‘benefit’ in this case is debatable, but if we imagine they had learned instead to climb up on a little rock to avoid the imminent zapping, that would have been beneficial.

This is a fairly astonishing, very powerful adaptive mechanism. Once you get something like this working, your pups are born already knowing to be scared of whatever serious threats in their environment mama and daddy absolutely feared and hated the most. And it can change, within just a few generations, whenever the environment changes. And of course fear isn’t the only thing it’s applicable to. Your genome can be reconfigured in thousands of ways, possibly millions, inheritable from your parents and acquired during your parents’ lifetime, to adapt you to the environment your parents faced.

So, you can guess where the first river of bullshit ran, can’t you? The same kind of people who got snowed by Jean-Baptiste Lamarck ‘Want To Believe’ and imagine that every kind of self-improvement you do is a benefit to your kids (which to be fair it can easily be) and forms a new ‘baseline’ from which your children’s self-improvement efforts can produce grandchildren who benefit from both, et cetera ad infinitum. And they’re eager to sell you all kinds of self-improvement programs, for the benefit of your family line down through the ages starting with you.

And, just as obviously once the mechanism is understood, that isn’t the way it works. Each site can be attached or unattached. It can be wound around a histone or in an accessible portion of DNA. It can be downstream of a transcription inhibitor that’s suppressed, or unsuppressed. There is only so much configurability, so many states.

And generally, given that we assume all these choices are part of nature’s effort to produce optimized organisms for various different environments, they’re expected to be on a Pareto-optimal front. Making something better means adapting to some condition, and that will probably make the adaptation to some other condition worse.

And now we get to the second river of bullshit. It turns out that when real research was done into what gets expressed in human epigenetics, a lot of it turned out to be adaptation to some highly adverse circumstances. And that Pareto-optimal front I mentioned turns out to be a real thing because mostly they’re maladaptive to a happy fulfilling life with healthy relationships and a stable job.

Early life stress, child abuse, addiction, starvation, anxiety, fear conditioning (like the mice), and so on, all cause their own set of epigenetic changes – and most of them, as described in the literature, are bad. An abused child often acquires epigenetic changes that will still be causing lack of ability to trust, lack of curiosity, and blunted emotions in their grandchildren. It’s a good way to get through an abusive childhood. It’s a bad way to live for the rest of your life or provide a decent life for your family.

So the second river of bullshit flows from charlatans who want to sell you on the idea that they can somehow magically clear your epigenetic profile, to help you be as you might have been if that cleansing mechanism had gotten everything, as if you (or your dad, or your grandmother, etc) had never been abused as a child and didn’t have those issues to deal with. And naturally, their snake-oil isn’t free.

Again, obviously once the mechanism is understood, that isn’t how it works. The only place and time that cleansing mechanism even runs is when you’re producing new gametes. While it’s always good for the pregnant woman to be in the best possible conditions as the child grows, during that time it is already too late for any epigenetic cleansing to benefit the child. And needless to say it’s far too late for any kind of cleansing that will help her. She’s inherited what she’s inherited. Once she is pregnant the child has already inherited what the child has inherited. There is no way to trigger that cleansing in any kind of differentiated tissue. While she can add some epigenetic markers during her life, she probably doesn’t want to. Mostly that would involve having horrible experiences and would cause configurations that won’t be beneficial to her children in any but horrible circumstances.

So, anyway, epigenetics is legitimately remarkable. It is nothing short of miraculous as far as adaptation and change goes, from a level of individuals all the way up the chain to the level of species. It’s amazing, even awe-inspiring.

But … The truth is awesome enough by itself. You don’t need to buy the bullshit too.