INTERVIEW WITH PILAR OSSORIO
edited transcript

Pilar Ossorio is a legal scholar, microbiologist and bioethicist who teaches at the University of Wisconsin - Madison. She is a leading expert on the ethical implications of genetic research.

What is race?

In our popular conceptions, we have a notion of race as being sort of simple divisions of people - divisions among people that are deep, that are essential, that are somehow biological or even genetic, and that are unchanging, that these are clear-cut, distinct categories of people. And that is not the case. All of our genetics is telling us that that's not the case. We can't find any genetic markers that are in everybody of a particular race, and in nobody of some other race.

Now, I would never say that race doesn't exist. One of the things we're hearing now is that the human genome project is somehow showing us that race doesn't exist. Well, what it's showing us is that the simple biological notion of race is wrong.

But just because race isn't something biological, that doesn't mean it's not real. There are a lot of things in our society that are real and are not biological. I think we would say marriage is something that's real. It's real because we have laws that we follow. We have social traditions that we follow. And those things are real. And race is real in that same sense. It's real in the sense that we create it in our relations between people. We treat people a certain way based on what race we believe them to be. Race, as we understand it as a social construct, has a lot to do with where somebody will live, what schools they will go to, what jobs they will get, whether or not they will have health insurance.

So race does play a very important role in our lives, and I think race is real as a social construct. It's not real as a biologically definable thing. But it's no less important because it's a social construct. It's still, in this country at least, it's a very significant social stratifying practice…And similarly, there is a way in which, regardless of genetics, we live in a social world in which we ascribe races to people, and we have race as part of our personal identity, and that's still very important for the world in which we live.

What makes us different one from another?

If we take the standard racial categories, and we ask: Are the people whom we call black more like each other than they are like people who we call white, genetically speaking, the answer is no. There is as much or more diversity and genetic difference within any "racial" group as there is between people of different racial groups. We also know that any two people will have a genetic difference between them approximately one time in every thousand DNA bases. So if you were just going along through the genome, base by base by base, in my genome and a friend of mine's genome, approximately once every 1,000 bases there would be a difference between us. If then we looked at my genome compared to some colleague of mine, a different friend, there would also be a difference between us once every thousand places, but it might be a different place. So people, we say, are about 99.9% genetically similar to each other.

Now, of course, when we study genetics, we're as interested in the differences as we are in the similarities. Of course, the .1 percent is - those are the genetic differences that make us unique from each other, genetically speaking. Those certainly aren't the differences that make us different as people, in the sense of, there is so much more than your genetics that makes you who you are. And right now in our society, there is a tremendous focus on genetics, and tremendous interest in genetics, but at the same time, we're sort of forgetting that there is biological development. There's psychological development. There are all of these other things that go into making a personality, even making the person's physiology. So what makes us different is both those genetic differences that we have between us, and also the interaction of that genome with the environment.

And the environment is a very, very complicated thing. So I sort of mean the environment writ large, everything from the environment in the womb to the environment in your school, toxins that you may be exposed to at various points in your life, in your development, and so forth and so on. There are many, many things that go into making us different from each other. One of those things happens to be the genetic differences between us.

There's another point that I think needs to be made: We tend to collapse or elide genetics and biology. But those are not the same thing. Genetics is a subset of biology, but biology is something much bigger. The relationship between genetics and biology is a complicated relationship. What you are biologically is only in part determined or influenced by your genetics. It's influenced by all kinds of other things as well, including what you eat, what your stress levels are, all kinds of other things. People forget that. We tend to talk about our genetic and biological make-up as though they're the same thing, and they're not. There's a lot more to biology than our genes.

The contingency of race

Race is also, to use the jargon, geographically and historically contingent. What do we mean by that? Well, I have a friend who's from the Bahamas; he grew up there. And he was white in the Bahamas. So he lived his life socially positioned as a white person. He came to the United States and instantly he became a black person. Because our understanding here of what it means to be white and black is different than the understanding of people in the Bahamas. So what we see and label as black is different than what other people see and label as black.

How a society constructs or understands race varies across societies, and even across parts of one society. There are a lot of things that go into ascribing race to a person. It's not just looks. It's not just physiognomy. It's also how they speak. Do they speak with an accent or not? Do they speak a certain style of English, or a different style of English, for instance? What religion are they, may have something to do with how we ascribe race to them.

So when we say race is geographically contingent, it means you can move from one part of the country, or one part of the world to another, and you can actually change race.

Likewise, we know that there is a certain percentage of people who are born one race, and die another race. And this is partly because there are different rules for how people's race is ascribed on their birth certificate from their death certificate. For a while, if any parent was black, then the child was labeled black. Sometimes it was just what the delivering physician thinks. Now I think in many states the rule is, the race of the mother as she reports it is the race that the child will be labeled on their birth certificate. On the other hand, very often when people die, the race that goes on their death certificates is whatever the coroner or officials ascribe to that person. Since obviously you're dead, you don't get to say what race you are when you die. And so there are a few percent of people who are born one race and die another race.

And historically, in order to be a naturalized citizen in this country, as an immigrant, you had to be categorized as white or black. And almost everybody who tried to naturalize - all but, I think, one case that went to the Supreme Court - all of them were people trying to be categorized as white. So the courts had to make decisions about who was white and who was not. Is an Armenian person white? There were a number of cases dealing with Asian people, and are they white or not white. And so one of the things that would happen is the person would come into court and say, "Look, my skin color is as white as anybody else's skin color in here who is categorized as white."

And it's very interesting to read these decisions and opinions, and have the court say, "Race is not only about the color of your skin. It's about other things too, like your attitudes about family, your attitudes about politics." The court often decided who was white and who wasn't based on whether they felt that the person would politically fit well into the kind of society we were trying to build. And sometimes it was pretty explicit that this was what the court was doing. And I don't think the court is different from other people in our society, even today, that we understand race in a very complex way that takes into account all of these other aspects of a person besides simply their skin color, or whether their nose is flat or narrow, or the shape of their eyes and how almond shaped they are.

Should race be used as a category in biological research?

I think as scientists we are still pretty confused about race, and what relevance race has to the things that we study, and whether or not race is a biological category. In part that's because scientists are human beings who have been socialized in this same society.

I think it's too simplistic to say that we should throw out the concept of race when we're doing science, because although race is not genetically definable, race is still, because of its social power, it still influences us in ways that may influence our health. So whether you live near a toxic waste dump of some sort, that's correlated with what race you are. Whether or not you have health insurance is also correlated with what race you are. And so there are many things in this society that are correlated with race that may influence our health. So I think it's too simplistic to say that we should just never use race as an analytical variable when we're doing science, because there are important things that we could learn, and that we should learn.

But the problem is that we shouldn't always assume that if we see racial disparities or differences, that those are due to genetic causes, as opposed to some other kind of cause.

I personally think that, usually, if we want to understand something about inheritance, we probably ought to be using categories that are much narrower and more clearly defined than race. We know, for instance, that on the continent of Africa, there is a lot of genetic diversity there. And so trying to understand something about ancestry, back to particular parts of Africa, might be very interesting. But if it is, then categorizing by "black" is not going to be a very helpful categorization.

And likewise, what does the category white mean? If you want to understand something about somebody's ancestry, and whether their ancestry goes back to northern Europe, or whether they have Celtic ancestry, and their ancestry goes back to the British isles, that might be much more important than the category "white." This is almost a non-category, as is the category black. The people who are captured within those categories are so genetically and biologically diverse, and their ancestries might be very, very different. There could be people in this country who are black, who might now be categorized as African American, whose ancestry actually is Australian indigenous. Very, very different ancestry. Yet, if we look at them, we might categorize them as black, the same way we might categorize somebody whose family is recently from Africa as black.

A lot of times when we're doing the science, we're not extremely careful about the descriptions of the populations that we're using, and so we will use 1,000 white people, people whose samples are gathered from various different points in the United States, for instance. But the ancestry of those people may be quite diverse. And then we will compare that to 100 or 50 samples of perhaps people from one town or small village from Africa. And then we will use the African sample to stand for "black people" and the sample of 1,000 white people from across the United States to stand for "white people." And you're comparing two quite different things when you make such a comparison, because you're comparing one group of people who may be fairly closely related to each other biologically, to another group of people whose ancestry goes all over the world.

And so it's not clear to me, necessarily, what the validity of those comparisons is. But however we can legitimately use those data, I think it's really important that we're very clear about who those data derive from. That one town in Africa should not necessarily represent all of Africa, or all black people, any more than northern Wisconsin should represent all white people. And we have a lot of confusion about that, and it's problematic for a lot of reasons. But one reason is because of how race and racism work. And one of the things that happens is that the less socially powerful group is perceived as being more homogeneous in general, as being less complex, less in need of individual attention, in some ways. And I think we play into that with the science when we use just a very small sample of African people, or African Americans, to stand for a whole "race" of people.

It's critical that we use terminology carefully and thoughtfully. If we don't, we confuse ourselves. We often use terminology, I think, in a somewhat sloppy way, and we actually use racial categorizations for people when we really are looking at something that is not race, but something much more geographically defined, much smaller, and more concise, if you will, than race…

And part of the problem with using race terminology when we really don't mean that is that we can reify the notion of race. And by reify I mean, take something that's an abstract notion and make it seem physically real. And that, I think, is very dangerous in the sense that, at the same time that the science is suggesting that these common-sense notions of separate, deep human racial categories are not really accurate or true, if we use racial terminology loosely and inappropriately we can reinforce people's notions that these racial categories are genetically real.

What do we make of newspaper stories about genetic differences between races?

We often see reported stories about genetic differences between groups of people. Now, what does that really mean? Usually it doesn't mean that in a certain group there's a version of a gene that's just absent completely from another group. That does occasionally happen, but it's very rare. Usually what it means is that, say there's one gene that has two different versions, or two "alleles." And in one group there might be 40% of allele A, and 60% of allele B. In another group, it might be 60% A, and 40% B. Those are statistically significant differences, but it doesn't mean that you could look at somebody from the first group and say, "I know that they have allele B," and somebody from the second group, you could look at them and say, "I know they have allele A." You can't do that based on these statistically significant group differences.

The other thing we need to think about is, the statistics change depending on the denominator. So how you draw the circle around the population, or each of those populations, could change the number that you get. So 40% A, 60% B, if you defined the population differently, you might get 45% A and 55% B. So those numbers, they're things that you can find. They're real data. They're statistically significant. But we have to be careful of attributing perhaps too much real world, concrete meaning to those numbers, or thinking that those numbers are going to be true for all times and all places and all people. Those numbers are true based on this statistical sampling in the population as you have defined and described it.

Pharmacogenomics: Why are drug companies doing population genetics research?

We're all interested in understanding why a drug is effective in some people and not others. Why do people respond differently to different doses? Some people might need a large dose of a drug. Some people might need only a small dose for it to be effective. Another thing that pharmaceutical companies, and others, are very interested in, is why do just some small proportion of people have a really bad side effect in response to a certain drug?

It has turned out though, that people who are studying what we call pharmacogenomics - i.e., the way genetic factors influence our responses to drugs - people who are studying pharmacogenomics are very interested in racial group differences in responses to drugs. And it's not really clear to me why they are.

Supposing you find that among those people that in this country we define as African American, that you have 40% A and 60% B, and B is an allele that helps you respond well to a certain drug, whereas maybe among white people in this country, however so defined, you have 60% A and only 40% B. Why would this matter to a pharmaceutical company?

Part of the reason it might matter is that they have to run all kinds of drug trials, all kinds of studies or research and experiments before they can get a drug to market. And if you can define a group of people who are more likely to be good responders to your drug, it may be that you can run fewer trials - that you need fewer people to get statistically significant results. You might be able to run faster drug trials. So there are reasons why pharmaceutical companies would like to know, as early as possible, who is likely to be a good responder to a drug. And those are legitimate reasons.

On the other hand, is this going to mean that once the drug is approved, it gets approved only for a specific population, one racial group of people? That's a lot more problematic, particularly if you've got the kind of difference that we often see, where there's a statistically significant difference in allele frequency, perhaps.

So say you find one of those differences. If a drug was only approved for use in one population of people, not everybody in that population is going to be a good responder to that drug. And not everybody in the other population would not respond to the drug. And so it's not clear to me that it would be really appropriate to say, "This is a drug for white people," or "This is a drug for black people," at least not based on genetics.

And in fact, we know there are a lot of nongenetic factors that can play into how people respond to drugs. Things in your diet can cause an enzyme in your body, which may metabolize the drug. Your diet can up-regulate an enzyme, cause your body to make more of it. Your diet can also inhibit an enzyme, cause it not to work as effectively. Those are things that can influence responses to drugs at the biological level, in measurable ways that may or may not map onto culture, more so than race. And those are biological responses. But they're not genetic responses.

So the differences we see in the way people respond to a drug - whether or not they have a side effect, how much of the drug is required before the drug is effective - these could be differences that have nothing to do with genetics, but may still relate to things like race or culture.

Genetic determinism - Correlation vs. causality

Looking for genes associated with any kind of trait is a really sort of sexy thing right now in science, and in the popular imagination of science. So there are all kinds of genes that people are trying to look for, genes that have to do with behaviors as well as genes that have to do with diseases, or genes that have to do with personality traits.

One of the problems with this kind of work is that it's easy to infer certain kinds of causality. If you find a relationship, a correlation, between a certain version of a gene and a behavior or a personality trait, it's very interesting to infer some strong genetic causality, which may be very inaccurate. And so some person could say, "Look, there must be a gene for incarceration." Well, you might find some gene associated with skin color that is also associated with likelihood of being incarcerated because we happen to be incarcerating a very high percentage of people of color in this country right now. So you might find some melanin gene associated with incarceration.

Now, does this mean that somehow this bit of DNA makes you march right over to jails and pop yourself in? No. It means that a certain degree of melanin in your skin, in this society, puts you in a certain place socially where you are more likely to be arrested, more likely to be convicted, more likely to have a long sentence. It's mediated through the social relations, not that there is an inevitable connection between this gene and incarceration.

And one of the fears is that if we think of genetics as being determinative in some sort of very fundamental way, where we have this sense of inevitability, that if something is genetic, then it's sort of unchangeable, that let's society off the hook. And even for basic biological traits, that's not necessarily the case at all. But certainly for these very socially mediated things, and psychological things, and personality traits, it's not, or very, very rarely, going to be the case that there is this kind of strong genetic causality between one or a few genes and some kind of social outcome, if you will.

One-drop rule and our common African ancestry

In some ways, we all have African ancestry. This is something that anthropologists have believed for a long time, and the genetics that we're doing is bearing that out. Modern humans came out of Africa, migrated out of the human population in Africa. Some people stayed in Africa. But basically, we all have African ancestry somewhere in our past.

Now, at certain points in this country, we've actually had laws trying to define who was Negro or black. And in essence, that is trying to define who is close to African ancestry. And so we had this "one-drop rule," this law that said basically if you had one drop of black "blood," you were black. And although those laws are gone, they still influence the way we think about races today. So in the United States, we talk about very light-skinned black people. And in a sense, that's the same thing as saying, "This is a person who has one drop." This person has maybe some African ancestry, and as long as we can see any African ancestry, as long as something is visible to us that looks like African ancestry, we tend to categorize that person as black.

And in truth, we all have African ancestry. The only question is how recent that ancestry has been. I mean, there is some way in which we all have one drop of African blood in us, if not more than that.

Actually, there have been times in this country, especially in the late 1800s and the early 1900s, when each state had its laws or its judicial decisions defining whether one was black or white. And that mattered a lot. It mattered because if you were black, then there were certain things, legally, that you could not do, certain places to which you did not have access, educational opportunities, etc. So it really mattered.

But interestingly, the different states defined black and white differently.

Genetic screening

One of the concerns about genetics and populations has to do with the targeting of at-risk populations. In other words, trying to understand when some group of people, however defined, has in that group a higher proportion of some allele that puts them at risk for a disease.

That kind of targeting can be useful for a couple of reasons. One is that you get fewer false positives if you have more real positives in the group of people whom you are testing. So one reason we want to target at-risk populations is to get better testing accuracy, and that's true whether you're talking about genetics or anything else. If you're going to do some sort of screening where you're testing everybody in a population, you don't generally want to screen a population where almost nobody has what you're screening for.

The other issue is economic. We don't have the resources in this world to do every single medical test on every person; we don't have the resources to do neonatal testing for every possible malady somebody might have. It's not just that it's economically inefficient. At some point it becomes perhaps immoral to use resources in that way. They're scarce, in a sense, and they could be put to much better use some other way.

So we don't just want to go around testing everybody for every conceivable thing that they might have or get. This means we have to make some choices, and trying to identify at-risk populations is one of those ways of making choices about when it's efficient enough that it's both economically feasible and perhaps morally acceptable to go around doing a kind of screening test on a group of people.

Then the question is, how do you define an at-risk population? And we have used race and ethnicity as ways of doing that. So we've used race as a way of deciding who might be at risk for sickle cell disease; ethnicity to decide whether or not somebody is at risk of having Tay-sachs, or other conditions that are inherited through the Ashkenazi Jewish ancestry. And so the question is, how do you define what the at-risk population is? How do you define that population for whom it seems acceptable to be doing screening tests? Are race and ethnicity the best ways of doing that?

Because it's always a kind of proxy. What you're trying to do is just narrow down a group of people in which there will be more than the average number of at-risk people in this group. Sometimes, in the past, race and ethnicity have been the only way we've had to do that, in a sense. And we still haven't done it very well. Because we could be asking people more specific things about their ancestry, rather than just looking at them and thinking, are you black, or asking them, "Are you Jewish?" as a way of putting them into the at-risk or not category. We could have been asking people more specific things about their ancestry now.

So it may be that in some cases we can ask people questions about their ancestry to narrow the at-risk population, to define it better, to define it more effectively than using race or ethnicity kinds of questions. Sometimes we might not be able to do that, because people might not, in general, know enough about the specifics of their ancestry for you to identify whether or not they are derived from people who are more likely to have sickle cell alleles, for example. But we do need to be careful. At least we need to be asking things to let us know if people have ancestors from India or a part of India or Greece where they might be at risk for sickle cell.

And similarly with Tay-sachs disease, there are other populations besides those descended from Ashkenazi Jewish populations. There are other populations where people might be at risk for Tay-sachs. We know this. We know that people descended from Acadians, French Canadians, have higher risk for Tay-sachs disease. So if we're trying to define at-risk in some way, if we just ask about Ashkenazi Jewish descent, we're still not doing a good enough job.

And it's always a tradeoff. If you can somehow define a population that is smaller than everybody who is born, and capture most of the people who are at risk in this more narrowly defined population, that has real advantages. But inevitably, some people will fall through the cracks. Just inevitably, there will be white people running around with sickle cell alleles.

<BACK TO TOP