"The Ghost in the Cell," by Scott C. Johnson. Published by Matter. Available via Web, ePub, Kindle. $.99.
Reviewed by Annalee Newitz
It's a prize that scientists have sought since the early nineteenth century: a biological marker that predicts violent behavior in humans. In the 1830s, phrenologists believed head bumps could reveal a criminal personality — often, prostitutes and the poor were said to have bumps that marked them as deviants from birth. But today, it seems this pursuit may have moved beyond the realm of pseudoscience.
Thanks to recent discoveries, we have evidence that the genes of abused children are marked by the experience. Over time, these effects leave them prone to depression and make it harder for them to control their violent impulses. Could we be on the cusp of discovering a scientific approach to a social problem? In an essay for Matter magazine, former war correspondent Scott C. Johnson suggests that we are. Unfortunately, Johnson fails spectacularly to explain the complexity of this problem, and winds up telling a story that distorts both the science and the reality of abuse in many people's lives.
Epigenetics is the study of how environment and non-genetic molecular activity affect development. Johnson shows us how seemingly-unrelated studies in the field converge to offer a picture of the violence-haunted brain. One study that Johnson describes suggests there may be an epigenetic pathway to suicide, whereby an abusive childhood leaves the brain starved for the neurotransmitters that normally keep our moods on an even keel. Johnson also delves into public health studies which show that violent behavior has its roots in childhood.
Johnson argues that we need to look at these studies in light of another epigenetic study that shows an abusive childhood can leave its genetic imprint even on the grandchildren of the abused person. The suggestion is that violence isn't just a social cycle, but a hereditary one. These studies and others like them have renewed many researchers' interest in childhood interventions to prevent violent adults. They're especially significant if it turns out to be true that the epigenome, as Johnson puts it, is "more malleable" in early life.
Johnson's science reporting here leads to a few interesting conclusions, but he glosses over many of the questions and controversies in epigenetics. He doesn't mention that other epigenetic studies have shown that humans retain neural plasticity throughout most of their lives. Nor does he admit that "criminality" and "propensity to violence" are traits that are difficult to define, especially at the level of neuroscience and genetics.
But most of The Ghost in the Cell doesn't deal with science. Instead, it deals with the violent life of one woman.
Throughout the essay, Johnson foregrounds the story of Yokia Mason, who lives in a high-crime, African-American neighborhood in Oakland, CA. We assume she's there to provide a human face for his descriptions of how epigenetics researchers and others are tackling at the influence of childhood experiences on gene expression. But we never learn how Johnson met Mason, nor why she's telling her story to a journalist. Has she been working with some of the scientists in the article? No. Is she the subject of a scientific investigation into genetics and violence? No.
So why is Mason in this essay at all? As we learn more about her background, one begins to get the extremely creepy feeling that Mason went to the most impoverished neighborhood in Oakland, the city where he lives, and asked people there to tell him about their violent backgrounds. Though Johnson and the scientists he interviews are careful to explain that they are not trying to bring race into this story of genes and violence, the entire shape of this article belies that assertion. It seems as if Johnson chose to profile Mason entirely because she's a black woman who lives in a low-income area with a lot of gang violence.
Mason feels "haunted" by violence, she tells Johnson, and he takes this to mean that she may be one of these people whose brains are primed for violence by a history of abuse. By extension, he hints, her entire neighborhood may be a hotbed of people whose brains are hardwired for crime. And this is where Johnson's essay moves from merely problematic to simply bad science writing.
As any of the scientists he interviewed could have told him — if he'd bothered to ask — Mason is a classic example of somebody whose situation is too overdetermined for us to make any solid claims about her propensity for violence. Is she influenced by a genetic predisposition she inherited, by her own childhood experiences affecting her epigenetically, or by social issues we can't quantify in the lab like economic difficulties, stress from having children when she was a teenager, or living in a neighborhood with gangs? We can't possibly know without studying her brain intensively, and even then the answers would be extremely murky.
Holding Mason up as the human face of these epigenetic studies is worse than misleading. She's a narrative non sequitur whose presence brings up the nineteenth-century notion that black people are somehow biologically destined for poverty and violence. One wonders why Johnson didn't choose to tell the story of somebody who is actually the subject of studies on violence. Instead he simply says that Mason is "the sort of person" who has been studied before, mostly by policy makers and social scientists:
Thanks to the work of Meaney and others, many researchers now believe that the neglect and fear that pervaded the Mason household will have left a powerful chemical imprint in the cells of Yokia and her siblings. If the scientists are right, we may have to change the way we think about tackling violence and crime.
You know, like maybe we'll tackle it by going to the poor areas of Oakland and giving everybody there forced gene therapy?
Given that child abuse is a social problem that's also common in middle-class households of all phenotypes across the United States, one would have hoped that Johnson might have explored this side of the story. What do these epigenetic studies tell us about rage-addicted CEOs whose brains were shaped by abusive parenting and sadistic prep schools?
Instead, Johnson devotes an incredible amount of time to a voyeuristic look inside Mason's life, which he reduces to a series of bullet holes, homicidal men, and children who are already in trouble with the police. This story does not humanize Mason; it objectifies her, and reifies the notion that there are some communities where crime is being bred into the next generation because of child abuse.
Epigenetics is a fantastically promising new field of study, which may ultimately shed light on how childhood shapes our neuroanatomy as well as our neuroses. But Johnson does the field a disservice by situating its story in a narrative like this one. By the end of this essay, you'll feel like you've just read a tale of phrenology rather than epigenetics, where science is being twisted to justify the idea that the people in some communities are just born broken.
Annalee Newitz is a science journalist who is the editor of io9.com and the author of the forthcoming book Scatter, Adapt and Remember: How Humans Will Survive a Mass Extinction (Doubleday).
UPDATE: Scott Johnson, the author of "The Ghost in the Cell," has responded to this review over at Medium.
O. Pagan says:
… And this is precisely one of the reasons why it is very important to educate the public about science. One does not need a PhD to write good, even great popular science, but whomever aspires to do that absolutely needs to do her/his homework. I am aware that I am preaching to the choir, but writings like this may be the only thing that someone will ever read about the topic. This is very dangerous…
Bobbie Johnson says:
Hi Annalee — one of the editors of MATTER here.
This disappointing review gets so many things wrong and makes so many insinuations and accusations that I felt we had to respond.
I’m going to keep it short, but: rather than simply cherry picking a horror story to “put a human face” on epigenetics, as you speculate, we specifically chose to work with Scott Johnson because he is probably the best-placed person in the world to write this story. For the last two and a half years he’s been reporting on violence, trauma and health in Oakland from the inside out as part of a fellowship with the California Endowment. It was through this reporting on systemic, inter-generational trauma that he met Yokia Mason, and from there he began investigating the current state of thinking about the epigenetic imprint of violence. It was not, as you suggest, the other way around.
As far as the science goes, I think your characterisation not only does a great disservice to the story, but also to the people investigating the epigenetics of trauma. As well as the scientists (whose work we are careful to give plenty of context and caveats), the world’s foremost experts on violence prevention and community trauma — people like Gary Slutkin and John Rich — are now looking at epigenetics as an important element, but not the only element, in trying to solve some of the biggest problems we face as societies.
Ultimately, however you paint it, The Ghost In The Cell is not a story about race, it’s a story about place — and it’s not a story about genetics, it’s about how our environment can leave an imprint on our biology. There is no evidence that anyone is hardwired to do anything, and nor do we say so: in fact, The Ghost In The Cell makes it very clear that epigenetics is one part of a hugely complex picture, and gives plenty of context along the way.
Scott’s away from his computer right now, and no doubt he will respond when he gets the chance. But if you think this is a conversation worth continuing, we’d love to follow it up because it’s important: I’ll drop you an email now.
Ronnie Raygun says:
Epigenetics used to be called Lamarckianism and has been disproved to be crackpottery.
Annalee Newitz says:
It would have been very helpful for Johnson to have stated outright in the article how he met Yokia, and what the context was of his involvement with her. As it stands, the reader is at a loss to understand how he knows her and what her stake is in all this. I am still unsure why she is in a story about epigenetics, since there is no evidence that she is one of the people affected by an epigenetic predisposition to violence.
Even if “Ghost” is a story about place, it still just so happens to be a place that is predominantly African-American, poor, and full of gangs. These are not the only places on Earth where childhood abuse happens, but nowhere in this story do we hear anything about white suburban violence, or childhood abuse among any groups other than Yokia’s family. The structure of the story suggests that Oakland is a problem — as if there are generations of people here who are breeding violence into their children.
But this isn’t what the science of epigenetics is about. These studies suggest that childhood abuse — which takes place everywhere — will affect the adults subjected to it. If that is the case, then this story shouldn’t be about place at all. It should be about a variety of people, from a variety of places, who have suffered from childhood abuse and now show symptoms that suggest a perturbation in their epigenome.
I strongly feel that you can’t just take a person with a violent background and say, “Well here is a person who might suffer from this epigenetic disturbance, so here is her story.” There needs to be more evidence that she actually does suffer from it, and isn’t simply suffering from poverty, disenfranchisement, and the resultant stress.
Carl Zimmer says:
Ronnie, you are mistaken. See this site for an introduction to the non-crackpot science of epigenetics: http://www.nature.com/scitable/topicpage/epigenetic-influences-and-disease-895
John Timmer says:
Researchers are indeed looking at epigenetics as a component of the body’s response to violence and deprivation; the review doesn’t question that at all. What it does point out was that, for whatever reason, you’ve held up a figure as emblematic of that potential connection without any sense whatsoever if she is.
To draw an analogy, imagine writing a feature on the genetics of cancer, and using a narrative figure that hasn’t clearly been diagnosed with cancer. Or, to use something with a greater social stigma, how about schizophrenia?
To me, separate from any potential racial issues, it just seems fundamentally misleading.
Jim Giles says:
Hi, Jim, MATTER co-founder here. In reply to John Timmer’s point…
If cancer was a poorly understood condition, would it be wrong to focus on an individual who doctors believed suffered from the condition? Because that’s the equivalent. The scientists we spoke to very much believe that childhood abuse of the type that Yokia suffered leaves an imprint at the epigenetic level.
Anyhow, although I disagree with the thrust of the review, I think that it’s great that Annalee brought up these issues. Our author just responded over at Medium (https://medium.com/oakland-stories/ea2d5b560f51) and I’d like to invite anyone else who wants to continue the discussion to write a follow-up to his post. Medium is invite only, but if you drop me an email at jim@readmatter.com I’ll get you an account.
Dr. MacA says:
From reading the free preview, it is definitely a poor choice to profile someone who has not been confirmed to have an epigenetic link to violence in an article about epigenetic links to violence. From all those who might be profiled, Ms. Mason is a particularly poor choice: as stated in the review, it is impossible to tease one factor in her experience of trauma apart from the others. It is unclear what influences like fetal alcohol syndrome or traumatic brain injury might be, or how effects of threats outside the home may manifest differently than those from within the family. There is not even a way to make a distinction between her experience of physical violence from emotional or sexual abuse, or of witnessing violence inflicted on others.
This conflation of Ms. Mason’s very complex personal trauma with differences in the brains of suicide and accident victims is deceptive: we do not know whether her brain has the trait studied by Mr. Meaney (and since the brains were collected post mortem, it is unlikely that the experimenters know very accurately who among the subjects experienced early trauma, what the reasons were for each suicide or how many of the auto accidents were risk-taking suicides). Even if we did, it would be impossible to know if Ms. Mason’s reports of suicidal feelings were the result of experiences of violence or a predisposition handed down from her father. A better subject, should the story need a human interest component, would be a veteran with no family history of child abuse who engaged in family violence after returning from war. Even so, the complexities of controlling for PTSD, TBI and economic stress would be considerable.
Violence is a very ambiguous thing. Few people have the dramatic home lives of made-for-tv movies, yet a much larger proportion of the population reports experiencing abuse in childhood. According to childhelp.org (citing statistics from http://www.acf.hhs.gov/programs/cb/stats_research/index.htm#can), six million children each year are listed as victims in police reports of child abuse and neglect; fatalities due to abuse are listed at five per day, or 1,825 per year. (We are ignoring problems of underreporting to simplify things, but it is widely acknowledged that there are both structural and social barriers to an accurate count). This means that a lot of the abuse is happening in families that are not the house of horrors that the author ascribes to the Masons. This makes the author’s correlation problematic: like basing a debate on gun violence on mass shootings, it ignores the statistical majority of violent events in favor of the sensational.
We know that abuse is often handed down from perpetrator to victim, and that the victim is more likely to become an abuser in later life. But we also know this about sexual abuse, whether the abuser and victim are related or not. It would be a rash conclusion, therefore, to assume either that there is an epigenetic mechanism or a traditional genetic one. Physical abuse does not travel alone: it is often the occasional reinforcement of steady emotional abuse. There are significant linkages to drug use, especially alcohol, and to social isolation. Intervention programs that teach parents better emotional and stress-management skills than they learned in their families of origin show significant benefit.
Epigenetics at its best can be seen as a very hopeful development, as it admits the possibility that we can influence health through behavior. But it is early days, and extending this to complex social, emotional and economic situations that involve many people over generations is completely unwarranted. This kind of confusion has the potential to taint the contributions that epigenetics may make. It is well established that factors like stress and diet can affect the development of diseases that also have a strong genetic component. But even these mechanisms are poorly understood and emotionally loaded. If epigenetics is to be considered real science and not another reactionary weapon in the culture wars, experiments must be designed to isolate the factor of interest as much as possible – twin studies, data from adopted children and brain imaging of living subjects who can be interviewed at a minimum. It also has to present its findings in ethical and straightforward ways. This presentation was neither ethical nor respectful to the real human being profiled, which makes both its methods and intentions suspect.
Sarah M Mah says:
Wow, what an interesting article, and some of the ethical issues you’ve raised are things I have been concerned about as someone currently doing research in social epigenetics. How will we conduct research in a way that takes into account the intersection of sex, race, and class? What steps will we take to ensure the media covers our science in a way that is accessible, but accurate and precise? And how to we safeguard against the exploitation of individual people, both their lives and their genetic/epigenetic makeup?
Thank you. Hope to read much more from you in the future.
gillt says:
for a visual on the importance of epigenetics in biology consider the many forms of honey bee.
http://www.alexanderwild.com/Art/Spring-Print-Sale-2013/28223041_fPWMMV#!i=2457420773&k=L6Z7KC2
gillt says:
“Nor does he admit that “criminality” and “propensity to violence” are traits that are difficult to define, especially at the level of neuroscience and genetics.”
This would be a good opportunity to mention the genetic research actually being done in neuropsychiatric genetics.
“Genomic architecture of aggression: rare copy number variants in intermittent explosive disorder.” American Journal of Medical Genetics
Also the Williams-Beuren locus is associated with aggressive behavior (duplications) and reciprocally, extremely friendly behavior (deletions). It’s on chr 7q11.22