Accueil - L'Observatoire de la génétique
Montréal • Beyrouth • Bruxelles • Cardiff • Marseille • Paris • Shanghai • Tunis • Leipzig • Toulouse
Accueil Le projet Zoom Cadrages Archives Hyperliens
No 29 -
septembre-octobre 2006

Tests génétiques et généalogie: dix raisons d’être sur ses gardes
Catherine Nash

Pour nous contacter

Titre - Zoom
Des articles un peu plus étoffés, présentant des synthèses ou des analyses. Une certaine manière de «centrer l’image» sur l’impact éthico-social.

Genetic Tests for Genealogy: 10 Reasons to Be Wary Imprimer

A case against the turn to genetic tests for answers to ancestry, identity and origins that foregrounds the limits of the methods and raises concerns about their social implications.

Catherine Nash*

Since my discussion of ‘recreational genetics’, race and relatedness appeared as a Zoom piece earlier this year (1), several people have contacted L’Observatoire de la Génétique troubled about the confusing or inconclusive nature of the results of their genetic tests for ‘deep ancestry’ and seeking advice about how interpret them. The help lines provided by most of the companies who sell tests that offer to supply information on ‘genetic ancestry’ or ‘ethnic or geographic origins’ must be busy responding to queries of this sort. For some people the tests provide, as the companies promise, a strong and significant sense of ancestral origin. But for many the experience of buying a test and then trying to understand the results and interpret their meaning for personal or family senses of origin and identity is very difficult and often unsatisfactory. For others, the results may be disturbing and unsettling. For others still, they can be only mildly interesting or end up meaning very little.

This article’s list of reasons to be wary about investing too much hope in these tests is a response to these requests for clarification and expressions of confusion. It is not meant to be a criticism of those who buy these tests nor of the desire to find out about one’s genetic background. The appeal of all kinds of self-knowledge, including knowledge about one’s genes, and the attractions of making connections with others are strong and widely shared features of the societies in Europe and North America where these tests are sold. It is also not meant to pre-judge how people use or interpret the results. It would be inappropriate to assume the role of expert and fail to appreciate the varied ways people deal with and make use of these new sorts of genetic knowledges. These comments are based on my own efforts to understand the basis of these tests and my reflections on their possible personal effects and wider implications.

This piece is, however, intended to be a direct counter argument to the heavy promotion of these tests by those who sell them as valuable, meaningful and unproblematic guides to genetic ancestry and origins. Critical discussions of these tests are beginning to emerge. In this short piece I want to add to those criticisms by presenting a strong and accessible case against the promotion of these genetic tests that will be of help to those considering buying them and those trying to understand their results. Some of these 10 reasons to be wary involve explaining the limits of the tests themselves; others raise questions about the social implications and the significance of genetic descriptions of identity or origins at a deeper level.

1. How much of an individual’s genetics do these tests describe?

‘Genes make you who you are as a unique individual’. Leaving aside the question of whether this is a useful understanding of genetics and identity and accepting, for now, the logic of this familiar viewpoint, it is worth considering the scope of the genetic information these tests are offering.

It is estimated that the total genetic component or genome of any one person differs from that of another by only 0.1% to 1%. This means that most of the genes that are said to make us who we are, are actually mostly shared with every other human. So these tests focus on the proportionally small amount of genetic material that differs between people. According to the companies involved, this is what makes the tests useful in ascertaining ‘ethnic or geographical origins’ (or in some cases fractions of different racial or ethnic ancestries).

Alternatively, it could be argued that the companies over-inflate the significance of these genetic differences by presenting them through the traditional importance of personal ancestral histories and the idea of individual human uniqueness. But even again, for now, accepting the claims of the significance of one’s genetic makeup to understandings of personal ancestry and individuality, it is also worth pointing out that these tests do not explore all the 0.1% to 1% but in the case of mtDNA and Y-chromosome tests, only focus on this particular chromosome and this particular form of DNA, and then only on selected segments of the strands of nucleotides that they contain. These are those areas or markers that have been identified by geneticists as areas of more rapid mutation and are therefore most likely to exhibit the most differences among people.

In addition, these areas of rapid mutation are located in what are described as non-coding portions of DNA. This means that according to current scientific knowledge they have no obvious function in terms of the production of proteins. So if we accept the tenet that ‘our genes make us who we are’, these tests are actually only telling us about a tiny proportion of our genome and about sections of DNA that have no known function (and for this reason cannot be called ‘genes’). For geneticists interested in what these sections of DNA might say about descent this is irrelevant; it is their narrowness and their form of direct transmission, from fathers to sons in the case of the Y chromosome and from mothers to children in the case of mtDNA, that makes them useful in exploring patterns of descent. But their narrowness also means that the tests only focus on a tiny portion of DNA that any individual inherits and they reduce the DNA that matters in terms of ancestry in any one genome to these selected portions. Autosomal tests, which I will discuss further below, focus neither on mtDNA nor on the chromosomes (X and Y) involved in sex determination. These tests are also restricted to the exploration of key markers on the autosomes. The genetic self-knowledge on offer is thus in no sense comprehensive. The results that are provided to customers are based on grouping people according to these narrow descriptions of descent. They say nothing about wider patterns of genetic similarity and difference.

2. What about all those other ancestors?

Y-chromosome and mtDNA tests only explore very limited portions of DNA. Furthermore, they focus on genetic material that is inherited directly. Thus, they cannot tell us anything about the complex mix of genetic material we have inherited from all the ancestors that precede any individual. In contrast to genealogy which at least in theory can provide a family tree that includes all the sets of great-grandparents, great great-grandparents and great, great, great-grandparents, and so on, whose genetic material has been mixed together and passed on to a present day individual, genetic genealogy largely, only focuses on narrow lines of direct descent. And while family trees can record the diverse geographies of birth, migration and residence of the individuals that represent ancestors, as well as all those other people who we aren’t directly descended from like aunts, uncles, cousins, in-laws, whose details enrich a family history, genetic genealogy can only offer one (if you are a woman) or two (if you are a man) points of ancestral origin. So, though genetic genealogy is presented as a way of overcoming the problem of lack of documentary sources, its solution to genealogical ‘brick walls’ is based on a narrow and reductive version of ancestry.

3. Where do those letters, numbers and names in my results come from?

So how are descriptions of ‘deep ancestry’, ‘ancestral origin’ or ‘geographical or ethnic’ origins derived from the analysis of selected segments of a customer’s genome? There are three main ways in which a customer may be given information about their ancestry, and there are specific problems about each method.

3.1. The use of haplogroups

Firstly, a customer’s Y-chromosome or mtDNA markers are given a label, usually in the form of a set of letters or numbers, that correspond to a recognised type, a haplogroup, such as the mtDNA haplogroups D, HV1 or K, or the Y-chromosomal ones like J2, R1A or J, that geneticists have identified as most common to particular parts of the world and that are the basis of their reconstructions of ancient human migration. Thus though the haplogroups are presented as personal results, and at least suggest their special link to the customer, as in announcements that ‘this is your unique genetic signature’, they are in fact shared with millions of other people.

These comparisons between a customer’s sample and these recognised, named and geographically mapped haplogroups are the basis of claims to be able to identify whether a customer has (direct line maternal or paternal) lineage to specific ethnic groups or that suggests specific geographical ancestral origins. Sometimes these ancestral categories can be broad. A customer may be told, for instance, that their haplogroup is one associated with African, Native American, Asian or European people. In some cases a customer is given a more specific ethnic or geographical origin based on more detailed regional studies of genetic diversity.

In addition because the haplogroups describe what are taken to be patterns of genetic diversity that are the product of patterns of ancient human migration and evolution, they only suggest supposed connections to very distant ancestral groups and their geographical locations. The time-scale of ‘genetic ancestry’ is not of recent generations but of tens of thousands of years back to ancient ancestors. Comparisons of the Y chromosome or mtDNA between individual customers can be used to determine whether they share ancestry at some point in the near or distant time. But the results of a test alone – with no such comparison – do not provide information on more recent ancestors who could have lived in many different places.

It must also be stressed that the haplogroup labels are the product both of individual genetic studies and a more recent international agreement on standardised naming. They are also based on current knowledge so a particular haplogroup could be further subdivided into several others as research progresses.

In addition to a named haplogroup, a man may also be given his results in the form of a set of numerical values for each of his Y-chromosome markers that have been sequenced.

In the case of mtDNA a customer may also be told of the difference between their rapidly mutating areas – or hyper variable regions – and that of the Cambridge Reference Sequence which is a composite sequence of mtDNA against which other types of mtDNA are compared. The use of this particular sequence is arbitrary, another reference sequence would produce other comparisons (2).

The basis of the test for Jewish or Cohanim ancestry is slightly different. In this case the test is based on examining a sample of men with the Cohen name and establishing which was the most common and therefore specific haplotype for the group (3). This appears straightforward but there is a crucial issue in this research and in the wider science of population genetics that informs genetic genealogy. It is the issue of sampling. In both the Cohanim study and in the naming and mapping of haplotype diversity more widely, these studies are not based on testing everyone named Cohen, nor everybody in a region but are based on samples of people that are taken to represent the wider group or population. This is a standard survey technique and scientists are aware of the issue of what counts as an acceptable sample size and composition. Nevertheless, a different sample, of Cohens for example, could produce a different most common Y-chromosome haplotype. (The results would also need to be compared with those of a ‘mixed group’ to see whether the statistical dominance of one haplotype is actually statistically significant. This also applies to genetic surname studies being promoted by several genetic genealogy companies.) In their surveys population geneticists commonly select their sample by screening out those people who are deemed to be ‘mixed’ or unrepresentative of a region from the study. This means that groups of people in specific places, that may be in fact be much more diverse because of the usual patterns of human migration, are instead represented as genetically homogenous .This means that the narrow focus on an individual’s direct maternal or paternal ancestry is coupled with a misleading representation of the genetics of the groups of people who are said to be genetically similar to a customer’s ancient ancestors.

3.2. Getting ‘matches’

Secondly, a customer may be offered sources of information on ‘ethnic or geographical origins’ in the form of ‘matches’ with other customers. Some companies invite customers to supply details of their own ‘ethnic or geographical origins’ to company databases that can be searched by the company or by customers for individuals with the same or similar results. (Mostly this is optional, but sometimes allowing one’s results and details to be searchable in this way is a condition of having the test done at all). A customer can then review the ‘ethnic or geographical origins’ supplied by other customers with whom they ‘match’. This is presented as a productive way of exploring ancestry especially for those whose haplogroup only suggests a very broad European ancestry that they had probably guessed already. It promises the possibility of more precise details on the places of ancestors and their ethnicity. However, this presentation of the value of the tests is often combined with a warning that the information provided by customers is subjective, based on their own ways of naming their origins and ethnicity.

But there is another fundamental problem with this method and it again involves the selectiveness of the database. It only contains the people who have bought the tests and allowed their results to be entered. This is thus in no way a representative sample of the people who could possibly match another individual. The information to be gleaned from the details of other matching customers is entirely dependent on the sort and number of customers who have contributed to the database. (This is also the case in  genetic surname studies.) A larger and more diverse set of participating customers would produce a different list of matches and their self-identified ethnic or geographical origins.

In addition, in some cases the ‘genetic matches’ are not based on other customers’ results but are derived from databases produced by collating the results of surveys conducted in the field of population genetics. However, although these are much larger databases, they are also the result of the selective sampling strategies of the researches and may contain many more matches to areas that have simply been surveyed in more detail. The issue here is the problem of aggregating information. Surveys may be done comprehensively in the particular regions being studied by population geneticists but the database that is produced by combining these surveys is not globally comprehensive. It is geographically patchy. A customer could, for example, find that they have ‘genetic cousins’ in Iceland whose population  has been intensely studied but none in other areas simply because those areas have not been subject to the same amount of research. Assuming an Icelandic ancestral origin in this case would be to ignore all the contingencies of the science that shapes these databases.

3.3. Autosomal tests

The third main method of providing ancestral information is used by the small but growing number of companies that sell autosomal tests. These are the tests which offer to estimate the proportion of a person’s ancestry that derives from each of four very broad groups. These are sometimes named as Sub-Saharan African, East Asian, European, and Native American, or in other cases named as West African, Native American, East Asian or West European. These tests are based on the use of databases too. In this case these databases contain statistics on the degree of prevalence of particular genetic markers (known as ‘ancestry informative markers’) in particular ‘populations’. They are also based on surveys conducted by population geneticists, using similar methods of sampling to study relatively ‘pure’ groups whose genes have not been ‘mixed’ by migration. In these tests, the customer’s ‘ancestry informative markers’ are examined and compared with the data on the greater or lesser occurrence of these markers in those surveyed groups. The results of this comparison are then statistically processed to produce a test result in the form of percentages of different ancestries. These percentages are therefore not a measure of the number of ancestors of different backgrounds within an individual’s genealogy but are a statistically derived estimate that depends both on the initial selection of survey participants, the examination of particular markers, complex statistical calculations and the convention of ordering patterns of human diversity into four main groups based on continental origins. Autosomal tests seem to offer a degree of statistical certainty about the proportions of different groups in one’s ancestry but the results are really an artefact of a whole series of estimates and approximations that have very little connection to the complex ways in which ethnical and racial identities are socially defined and experienced. At the same time they can easily been interpreted as supporting to the idea that racial or ethnic categories have a genetic basis.

4. Can race and ethnicity be described genetically?

In contrast to Y-chromosomal and mtDNA tests, autosomal tests do present an individual ancestry as one of mixture and multiple origins, but they do so by giving credibility to the idea that dividing humanity according to four racial types neatly corresponds to patterns of genetic difference. So while some geneticists argue that genetics refutes the existence of genetically discrete races, others readily make use of old racial categories. Those who do so often acknowledge that in reality the patterns of human genetic difference vary in gradients, but their fourfold description of humanity suggests much more distinctive groups and clear cut divisions (4).

Genetic genealogy is thus implicated in the contemporary resurgence of a racialised language of human difference. This may seem completely disconnected from an ordinary customer’s interest in what the test result can add to their family history. But buying a test effectively supports companies and geneticists that are influential agents in contemporary debates over the scientific credibility and social consequences of equating ethnicity, race and genetics.

This issue of the ways human genetic variation is correlated with race, or more usually ethnicity, in the presentation of many of these tests is one of the most significant causes of concern about this new development in genealogy. I put populations in inverted commas in Section 3.3. because the use of the term implies the existence of much more bounded and discrete groups of people than actually exist in reality. Though ethnic groups are often based on ideas of shared ancestry, human groups are always more mixed and fluid than pure and fixed in their composition. Attempts to achieve or maintain an assumed ethnic purity have been the cause of massive human suffering, manifest at worst in ethnic cleansing. In other cases the correlation of ethnicity/race with genetics produces much more stringent criteria for membership than usually operates. In the Cohanim case, for example, being Jewish is newly defined through the presence or absence of a particular Y-chromosome haplotype, though in practice Jewishness is traditionally inherited through maternal descent and can be achieved through conversion.

There is a further problem with attempts to define the genetic characteristics of ethnic groups. In some cases historic ethnic labels are projected on to contemporary people. The identification of a Viking haplotype, for example, is based on the assumption that contemporary Norwegians descend from and share the genetic patterns of Viking ancestors. In other cases contemporary ethnic labels are projected back into the past. The contemporary ethnic groups in West Africa whose genetics are used to locate the place and group of origin of a customer’s enslaved ancestors may not correspond to those that existed at the time of enslavement. Genetic tests in genealogy are based on over-confident assumptions of the historical continuity of genetic patterns and ethnic labels.

5. What happens to my sample and my personal data?

Beyond concerns about the scope and the limitations of genetic tests in genealogy, potential customers of companies that are selling these tests may also find it useful to reflect on what happens to their genetic information after the test has been conducted. Most companies now have detailed terms and conditions that customers are invited to read and accept before their sample is processed. Most stress the privacy of the test results (allowing however for different degrees of anonymity and accessibility on those customer databases) but also protect the company from any liability for any negative ‘social or psychological’ effects of the results. Most also say that the sample of a customer’s cheek cells and the genetic material derived from it will be destroyed after it has been examined, or destroyed after a period of storage to allow for re-analysis as the sophistication of the tests develops. However, destroying the physical sample does not destroy the information derived from it. And though under usual circumstances this information would not be disclosed, some companies acknowledge that they may be legally required to disclose the information in cases of criminal investigation. This may not seem problematic but the possibilities of future use of the information either in future genetic studies or in other unforeseen ways are worth considering. As in the case of genetic databases more widely, consent to the analysis of the sample for the sake of the current use does not necessarily ensure the customer has any control over the future of their genetic information and its possible uses.

6. Do my relatives want to know what I find out?

Those considering buying these tests may also want to reflect on the potential effects of their results on other family members. For though the tests give individual results they have implications for other relatives who in the case of Y-chromosome tests share paternal descent, and share maternal descent in the case of the mtDNA tests. An individual customer usually signs a consent form when completing the tests but the results are also relevant to other family members who are not offered the opportunity to sign or decline consent. The information that the results suggest about ethnic or geographical origins that one person receives thus pertains to other family members who may not have chosen to have this knowledge and for whom it may disturb their particular sense of cultural or ethnic identity. A man, for example, who may strongly identify with his Irish roots because of his maternal grandmother’s origins, may be told by his brother, son, or uncle that the Y-chromosome ‘matches’ suggest connections with Italy instead. Or as is frequently the case, Y-chromosome tests results can point to a white male ancestor for African-American or Black British men. These may be upsetting results for the individual who bought the tests but they may also have effects on those in the family that had not personally chosen to explore their ancestry in this way.

In addition, the ways these two most common types of tests focus on direct maternal and direct paternal descent alone can suggest different degrees of genetic similarity and difference that do not correspond to the patterns of closeness or distance within a family. The diagrams of family trees that many companies include in their explanatory material often highlight the direct paternal and direct maternal lines within the family tree to show to whom the test and the test results are applicable. Once these schematic diagrams of descent come alive with the names of actual relatives, the way the tests emphasise certain lines of the family over others becomes much more obvious. In my case, a focus on my mother’s line alone seems to exclude my father’s grandmother, as well as all those other women on both sides who are not direct maternal ancestors. When those who are logically deemed irrelevant to an individual’s genetic ancestry are still alive and interested in their daughter’s, grandson’s or nephew’s results, the discovery that they don’t share the haplogroup in question, can cut across their bonds of care and affection. A man who has had a close relationship with his mother’s father may have to explain to him that, actually, the results are only relevant to those men on his father’s side of the family, and that the results define his ‘deep ancestry’ in this way alone. In this case they do not share mtDNA either.

Of course close relationships and senses of identity may survive these new ways of thinking about descent. The genetic results can be combined with other ways of thinking of the family and other ways of defining the relationships and ancestral connections that matter. But an individual’s results can also have affects on other family members and on family relationships that are worth considering in advance.

7. Why should direct male-line and direct female-line ancestry and relationships matter most?

As I have already outlined above, some of the recent concerns about geneticized genealogy focus on the ways in which they connect ethnicity, race and genetics. But, these tests also suggest ways of thinking about descent in strongly gendered terms. This is not a matter of the simple inequality of men being able to take both Y-chromosome and mtDNA tests and women only being able to take mtDNA tests. Most companies respond to women's queries about the Y-chromosome test by suggesting that women could get a paternal male relative to take a test so that they can find out something about their father's line. However, this is not an ancestral line a woman is genetically connected to since she doesn't inherit the Y-chromosome from her father and since this paternal line is defined in genetic genealogy by the Y-chromosome alone. Yet, according to the idea of one’s own unique DNA mattering in terms of identity, these proxy Y-chromosome tests aren’t telling the women anything about herself but about her father, and his paternal line.

The gendering of ideas of descent has both more obvious and more subtle dimensions. Most obviously the promotion of Y-chromosome tests and especially genetic surname studies using the Y-chromosome, regenerate a model of genealogy dominated by the significance of patrilineage. The patrilineal naming system of many European countries and of the countries of European settlement reflects a social order in which women were subordinate. It is only relatively recently that genealogy has become more democratically focused on female as well as male ancestry. Some may argue that the existence of mtDNA tests for women offsets this renewed emphasis on surnames and male descent. Yet in some cases ideas of collective descent can come to be defined much more narrowly through male descent. Most Irish and Scottish clan organisations that are familiar expressions of ethnic affiliation in Canada and the United States allow membership of women and men who have a genealogical link to the clan name regardless of whether this is direct and paternal or not. But when the genetics of the clan are explored though Y-chromosome tests, membership, implicitly at least, becomes more narrowly defined as male and, implicitly at least is only open to men with direct paternal descent from a migrant ancestor who bore that name.

The subtle gendering of descent and relatedness through both the promotion of mtDNA tests and Y-chromosome tests also involves the ways in which relationships between  sons, fathers and grandfathers and backwards in time, and similarly between daughters, mothers and grandmothers and so on, are emphasised at the expense of other relationships. As suggested above, this emphasis on my mother’s maternal line makes my father’s mother somehow insignificant as an ancestor. Similarly, for my brothers the emphasis on direct Y-chromosome descent means that only one of two grandfathers, my father’s father, can count as genetically relevant. The explanatory material of many genetic testing companies acknowledges that this is a narrow version of descent but argue for the usefulness of being able to trace these forms of direct descent when the numbers of ancestors in an individual’s family tree multiplies with every generation backward in time.

Those popularising these tests also make the idea of focusing only on direct paternal and direct maternal genetic connections seem natural by drawing on the idea of the strength of the relationships between mothers and daughters and between fathers and sons. But in doing so they over emphasise these relationships at the expense of others - between grandmothers or mothers and sons and between daughters and fathers or grandfathers. This emphasis on these forms of genetic connection exaggerates the commonalities that supposedly underlie the so-called ‘naturally’ strong bonds between mothers and daughters and between  fathers and sons. In doing so, it also exaggerates the supposed differences between women and men. Geneticized genealogy can conjure up a world in which descent is only imagined in terms of separate and unconnected direct maternal and paternal lines, and origins, for men at least, in two regions and two ancient migration pathways. Maps of Y-chromosome and mtDNA haplotypes and their migration routes seem to suggest a human prehistory in which bands of women and bands of men separately left Africa and pursued their global journeys in isolation from each other. In this way genetic genealogy is part of a wider and problematic tendency to stress supposedly inherent differences between men and women and use these differences to justify social inequalities between women and men.

8. How to decide who genetically belongs?

Tests results can have implications not only for individuals and family relationships but also impact upon senses of belonging within wider social groups and in some cases formal inclusion or exclusion from groups that are defined through shared descent. For many people the possibility of the tests confirming senses of group identity – as of Scottish descent for example – is their attraction. But the results of tests that focus on the genetic profiles of members of social groups based on shared ancestry – Irish clan or surname groups, for instance – often differentiate between those who fit and do not fit, by referring to a haplotype taken to be typical of the group. As in families, senses of shared identity within groups can be challenged by test results that suggest that those affinities have no basis in genetics and shared descent. A man may find that a long held affinity with and interests in the clan histories and mythologies of his Irish or Scottish surname do not match his Y-chromosome haplotype since it turns out not to be the same or similar to the one that is most common within the group. These results may matter in terms of cultural identity but have no more formal consequences in the individual’s life. However, genetic test results could have legal and financial as well as personal and collective political and cultural effects if recent moves to insist that claims to being Native American, or of belonging to other indigenous groups, are supported by genetic tests become formal state policy (5). And as in families also, these ways of reckoning membership genetically do not necessarily correspond with customary ways of deciding who is a member and not a member of a community or group. Potential customers may want to consider how the promotion and consumption of these tests may support models of group membership based on genetics that can have divisive and damaging effects.

9. Is blood really thicker than water?

What are wider implications of the emphasis on genetic closeness and genetic connections within the culture of genetic genealogy? The possibility of making connection with others as ‘genetic cousins’ is often presented as a particularly rewarding aspect of the practice. In addition it is common to come across references to the natural senses of affinity that exists between those who discover that they share the same haplogroup. This extends the everyday idea of the significance of family connections, the idea that ‘blood is thicker than water’.

Framing genetic genealogy by ideas of family ties makes claims about the significance of finding ‘genetic cousins’ and celebrations of genetic connection appear benign and natural. But the idea that despite separation by thousands of miles, in many cases, and despite no previous contact, newly discovered ‘genetic cousins’ share a bond and sense of affinity based on shared possession of particular markers within their mtDNA or Y chromosome has as its corollary a more troubling model of human difference. For it is the logical extension of arguments about genetic similarity as the basis of social connection, is that senses of care, empathy, understanding and solidarity diminish with increasing genetic dissimilarity. The implication is that we are naturally bonded to those most genetically similar and naturally indifferent to, unsympathetic towards, or at worst hostile to those most genetically different. There are clear echoes here of claims about the ‘natural’ enmity between ethnic or racial groups that are used to argue against immigration, or used to explain social divisions or ethnic tensions in terms of ‘natural’ antagonism rather than in patterns of inequality that are the product of ethnic and racialised patterns of advantage and disadvantage.

Ideas of the genetic basis of empathy suggest a world of allies and enemies that could justify a whole range of state practices from the ethnic or racial segregation of housing, to anti-immigration legislation, to foreign policies designed in response to ‘natural’ patterns of antagonism or tension between those culturally, racially and genetically different. This may seem very far removed from personal tests for ancestral origins, but the claims made by genetic testing companies are part of a wider social domain in which the explanatory power of genetics is being explored, contested and put to work.

10. Am I just my DNA?

My second broad point and final one of this list concerns the relationships between genes and identity. The companies that sell genetic tests for personal or collective explorations of ancestry make strong and straightforward claims that the results will provide information on ‘who you are’ and ‘where you come from’. But what model of identity is at work here? In what ways does it correspond with or differ from wider understandings of identity?

On the one hand these claims pick up and reinforce existing understandings of the significance of ancestry to people’s sense of identity as individuals or as members of ethnic groups. They reproduce the idea that at some deep level identity corresponds to our biology, that it is predetermined, fixed and immutable within our genes. They imply that finding your genetic origins thus reveals something that we hitherto didn’t know but is nevertheless fundamental to who we are. But, on the other hand, in contrast to the emphasis on genetic inheritance alone within genetic genealogy, people often understand themselves not just through senses of what is genetically inherited but also in terms of what is culturally inherited, stories, traditions, attitudes for example, in terms of a much wider set of social relationships, and in terms of the impact of their childhood and lifelong experiences.  My sense of who I am is always a mixture of what seems pretty much set in place and my continuing and unfolding responses to the other people and events. Some people may add new knowledge of genetics to their diverse and sometimes contradictory ways of understanding themselves or give a relatively minor place to genetics in their senses of who they are. But to give priority to genetic descent and genetic origins over senses of the continuously unfolding and dynamic shaping of identity is to reduce understandings of ‘who we are’ to the most narrow and limited versions. Of course this may be comfortingly uncomplicated and that may be the attraction of the tests, but as this list has suggested, neat categorisations of people according to genetics in families, ethnic groups and in the world more widely, are more a cause for concern than satisfaction.

These ten reasons to be cautious about turning to genetic tests in genealogy obviously represent my personal perspective. Others will have different views of their worth and limits. But they are offered here as critical interjection and to provoke debate in an area that is currently dominated by the companies’ own versions of what the tests can provide and why a potential customer should ‘proceed to checkout’.

*Catherine Nash is Reader in Human Geography in the Department of Geography, Queen Mary, University of London. Her research interests are in feminist cultural geography and in particular in questions of embodiment, identity and belonging. Her current research project ‘Genealogy and Genetics: Cultural Geography of Relatedness’ is supported by an Economic and Social Research Council Research Fellowship (RES-000-27-0045). This paper also benefits from an additional British Academy Research Grant (‘Recreational genetics: an autoethnography, 2006).


(1) Nash C. “Recreational genetics, race and relatedness”. L’Observatoire de la Génétique September-November 2005, no. 24, online bulletin of the Centre for Bioethics, Clinical Research Institute of Montreal,
(Accessed October 31, 2006).

(2) The Cambridge Reference Sequence is a reference genome of human mitochrondrial DNA against which other mtDNA sequences are compared and named. It is a technique for exploring the differences between mtDNA sequences. Yet, another reference genome could serve the same function. It does not represent an individual (it is actually a composite based on sequencing the cell material from two women; nor is it somehow a ‘typical’ sequence. For a critical analysis of the technological and cultural processes that have produced the Cambridge Reference Sequence as a taken-for-granted reference sequence see: M’charek A. “The Mitochondrial Eve of Modern Genetics: of Peoples and Genomes, or the Routinization of Race”. Science as Culture 2005 14 (2): 161-183.

(3) The term haplogroup usually refers to a broad lineage composed of related haplotypes, the term used to describe more specific sequences. For a fuller critique of the Cohanim study, see: Marks J. “We’re going to tell these people who they really are: Science and Relatedness”, in Franklin S, McKinnon S (eds). Relative Values: Reconfiguring Kinship Studies. Durham, NC: Duke University Press, 2001, p. 370.

(4) See: Condit CM. ‘“Race” is not a scientific concept: alternative directions’. L’Observatoire de la Génétique September-November 2005, no. 24, online bulletin of the Centre for Bioethics, Clinical Research Institute of Montreal,
(Accessed October 31, 2006).

(5) There are reports that the current US Bush administration has ‘planned to introduce new legislation for new laws to require DNA tests to determine Indian blood’. Arizona Border Rights Foundation - Fundación de Derechos Fronterizos de Arizona press release regarding the Border Summit of the Americas, Tucson, Arizona, 29 September to 1 October, 2006 posted on (Accessed October 31, 2006). In contrast, the Australian Law Reform Commission’s report Essentially Yours: Protection of Human Genetic Information in Australia summarised the reasons against genetics tests over riding indigenous peoples own rights to self-determine group identity and membership. See its section ‘Kinship and Identity’ at
(Accessed October 31, 2006).



Haut de page

Copyright ©2006 - L’Observatoire de la génétique / Tous droits réservés.




Vers le site du Centre de bioéthiqueVers le site de l'IRCM