Anthropology, Genetic Diversity, and Ethics 
A workshop at the Center for Twentieth Century Studies 
University of Wisconsin-Milwaukee  
Eric Juengst*
[Participant Information]

Thanks, Trudy.  My name is Eric Juengst.  I'm a philosopher by training; I teach medical ethics and philosophy of science at the Medical School at Case Western Reserve University.  I also worked for a while at NIH with the Human Genome Project, back when it was the only initiative that could claim the title, "the lightning rod of human genetics."  These days, there are a number of genetic research initiatives that seem to be lightning rods for public attention and controversy  That's in part because the Genome Project itself is at the stage where it's getting ready to reproduce, it is expecting not just one baby, but at least sextuplets!  I want to begin my remarks by showing you the six progeny of the Genome Project that I think are relevant to our discussions.  Right now, there are discussions going on within NIH and the scientific community about how best to use the tools of the Genome Project to pursue each of six new areas, initiatives: 1) pharmacogene, 2) functional genomics, 3) the Environmental Genome Project, 4) The Human Genome Diversity Project, 5) complex trait genetics, 6) Public Health Genetics.  I won't go through them all, except to say that the Human Genome Diversity Project, which most of you know most about, is just one of the siblings, the one that the anthropologists and evolutionary biologists are interested in.  But all these initiatives have something in common with the Diversity Project and each other, and that is all these initiatives are based on the technique of looking at the variations between the genome of individual human beings and looking at the differences between those, the gene frequencies of different groups, different populations.  They all involve, in other words, comparative human genomics, which in practice is comparative population genomics.  They involve comparing genetic differences between human beings. 

If comparative population genomics is going to be the basis for the next century's medicine and pharmacology and epidemiology, it bears spending a little time thinking about the assumptions that go into the science.  And one of the first questions is: what counts as a group for purposes of genomic comparisons?  What are the groups that are appropriate to compare when you are doing genetic studies?  That's the question for the first part of our workshop, the question for the first session here this morning.  It sounds like a pretty theoretical -- even philosophical -- question, but it has some practical consequences for researchers, some of which are listed in the questions in the workshop brochure. 

First, who should be responsible for identifying -- targeting, if you like -- the groups that we study?  The scientific community?  The political entities that govern the groups?  Representatives from different groups?  There was one proposal that essentially we ought to do this research like we do the rest of clinical research in medicine, and simply put out a call for volunteers, and let groups come forward on their own, without any prior identification by us of who would be interesting to study for particular reasons.  Well, most of the time that's not how it is working; most of the time we have an idea ahead of time of who we'd like to study.  But that's at least a logical possibility. 

Secondly, should researchers try to avoid using socially defined group boundaries in designing genetic studies?  Why or why not?  And finally, what about human, nested human groups, in which a research on a subgroup has implications for the larger whole of which it's a part.  How would you distinguish between subgroups and supergroups in those situations? 

We'll have three short sets of remarks.  I'll go first, and then Frank Dukepoo will follow me.  I'll introduce them now, so that we won't have to take time during the remarks.  Frank teaches in the Department of Biological Sciences at Northern Arizona University.  He says in his bio, that he is a full-blooded American Indian of Hopi and Laguna heritage, and he's one of six Indians nationally who hold earned doctorates in the sciences, and the only Native American geneticist.  Our other speaker is Mike Bamshad.  He's the Assistant Professor of Pediatrics at the Eccles Institute of Human Genetics and the University of Utah.  His research interests include the identification of disease genes, limb development, and evolution. 

Well, my remarks will be short, and just to put my cards on the table, I guess what I want to suggest is that I think the second question of these questions.  ("Should researchers define the groups they study in terms of socially constructed communities in the first place?")  My answer would be, in a lot of cases, no.  I guess I've been convinced by the arguments of people like John Moore, who suggest that it's going about the study of human differentiation, and even the study of human migration, around the globe backwards, to start with our current set of culturally perceived, socially constructed groups, and sample them for DNA in order to reconstruct a pattern of lineages into the past.  He's agreeing, with Alan Wilson's original vision for the Diversity Project, which would have simply sampled randomly around the globe, and let the lineages and populations, the genetic demes, if you will, emerge from the data as they would, without reference to the peoples that they were sampled from.  Those kinds of correlations would come later, but they wouldn't be the premise, the starting point, the frame, for the study itself. 

Why should we go about it that way?  Well, John's arguments are scientific, he argues that it avoids the bias that starting with the culturally constructed groups imposes on the stories that we tell.  He says you'll get the answers you're looking for if you're looking for differences between groups.   Although modern population genetics tells us that the human demes that we could eventually construct, will look very different than the map of human communities that we live in, it's still possible, if you look for the right markers and use the right markers, to find markers that differ between most of the people in the socially constructed human communities.  So you can tell those stories; but your conclusions will be built in from the beginning. 

Moreover, I think it makes this comparative genomic research particularly hazardous.  We're used to thinking about genetic research as a two-edged sword.  The Genome Project has done a lot of work analyzing the relative risks of finding out an individual's genotype and weighing that against the relative benefits.  I think we're at the next level of difficulty here with the comparative group work; it's actually a four-edged sword.  Imagine how tricky this would be to wield in a safe and useful manner, without cutting yourself in the process.  The four edges are on two distinct blades of the research -- on the one hand, we're interested in the diversity, the differences, the variation, that we can find between people, and we're interested in that because of at least for the biomedical efforts, of the ways we might be able to tailor our health care interventions to people with differential genetic weaknesses and strengths. 

On the other hand, by the same token, as you identify people with differential susceptibilities, weaknesses and strengths, you also label them with those weaknesses and strengths, and as we have come to know that the Scots Irish are likely to go awry in one particular direction or another, then that's another burden to our reputation that we have to bear.  These are the reputational risks that Hank Greely talks about.  On the third hand, what the geneticists tell us, we'll mostly find in doing these studies, is a lot of similarity between human groups, much more similarity than difference.  That has the advantage of cutting against the social line we draw between ourselves and others, fighting against racism this will show that we're really one big human family.  On the fourth hand, it will also show that the identities we build for ourselves that are built on those differences may not jive with science.  Whether or not you want to buy the scientific story of your origins as opposed to the cultural story of your origins that you build your identity around, will become an issue for the communities whose origin stories are contradicted by the scientific discovery of similarity with other groups. 

Moreover, I'm afraid that we don't have any really good way to protect groups against the negative edges of these blades.  And I guess that's where it comes down to the discussion point: If we want to work with socially identified groups, can we do it in a way that protects them from these possible adverse consequences?  The two main protections we have to offer is that we can try to get their consent to participate, let them know what the risks are, and if they don't want to participate, presumably shield them from those risks by not having their genetic differences uncovered; and we can try to keep the results secret, by preserving their confidentiality by not identifying them as the social group we studied in order to get these genetic results.  Both of these can be useful gestures in dealing with a people respectfully; both of them, I think, are pretty limited in the protections they can actually provide for the members of the group.  Just to look at the group consent idea, my worries there are that it feeds what we'd like to avoid: the conflation of our social identities, these social groups, with a biological sense of a population.   When one says, "You are the group that's authorized for a study of the genetic population to which you belong," that suggests that you, the social authority group, and the population are one and the same, a view we don't believe, and we expect the research, actually, to eventually contradict. 

Moreover, as long as there are supergroups to which this group belongs, the protections are limited if another subgroup is willing to go along with the research.  And, of course, we also live in a world full of diasporas.  There are expatriates from almost all social groups, so that if we decide to study the unique genetic complexion of the Hmong, why go to the elders of the Hmong in Southeast Asia?  Why not go to the Hmong college student in Minnesota?  It would be odd to think we would let the elders in Southeast Asia prohibit an American citizen from joining a research study that's been duly approved by our own IRBs. 

So I'm not sure that the idea of containing the problem by requesting consent will take us very far.  At best, the group that refused can always say, "They didn't use our genes to make that outrageous claim about the supergroup," [or]  "It wasn't us when they said that Southwestern Indians carried the gene for alcoholism, that was another tribe."  But I'm afraid that's [a] pretty weak defense, when you are a Southwestern Indian, and now you do face the burden of being in that pool of people who've been labeled with the alcoholism gene.  That's one reason, of course, that sheer anonymity about the specific identity of the groups won't help.  We got the two articles in our packet about the genetic studies of alcohol dependence in a Southwestern Indian tribe, that wasn't identified in order to protect confidentiality.  Well, that's a good move, but as a result, the sphere of repercussions for that study, if there are any repercussions, is much broader now: it's the whole supergroup of Southwestern American Indians that will have to pay any price there is to pay for that research.  I guess what I'm wondering is whether we can get beyond looking at the phenotype and move into more of an Alan Wilson field biology approach, and work from the genes up to build our sense of what the relevant genetic populations are.  The map of those populations probably won't look much like our map of political and social entities. 

*This talk has been edited for web publishing by the author.