The Trouble with Race in Forensic Identification

Abstract

The capacity of contemporary forensic genetics has rendered “race” into an interesting tool to produce clues about the identity of an unknown suspect. Whereas the conventional use of DNA profiling was primarily aimed at the individual suspect, more recently a shift of interest in forensic genetics has taken place, in which the population and the family to whom an unknown suspect allegedly belongs, has moved center stage. Making inferences about the phenotype or the family relations of this unknown suspect produces suspect populations and families. We discuss the criminal investigation following the Marianne Vaatstra murder case in the Netherlands and the use of forensic (genetic) technologies therein. It is in many ways an interesting case, but in this paper, we focus on how race surfaced in science and society. We show that race materializes neither in the technologies used nor in the bodies at stake. Rather, race emerges through a material semiotic relation that surfaces in the translation that occurs as humans and things move across sites. We argue that race is enacted, firstly, in the context of legislation as biology reduced to bodily characteristics; secondly, in the forensic analyses as patterns of absent presence; and, thirdly, in society as a process of phenotypic othering.

Introduction

On the morning of May 1, 1999, a sixteen-year old girl, called Marianne Vaatstra, was found raped and murdered near the Frisian village of Kollum in the north of the Netherlands. Despite several and extensive investigatory efforts, the case remained unsolved for many years until it finally came to a closure in November 2012. The Marianne Vaatstra case is in many ways remarkable. It was the subject of numerous newspaper articles and documentaries and, most significantly, Parliamentary debates provoked legislation aimed at expanding the use of forensic DNA technologies in the hopes of solving the case. The proximity of the crime scene to a center for asylum seekers led to ongoing hostilities toward its residents who are mostly migrants from Iraq and Afghanistan. Their suspected involvement led to both racializing the unknown suspect and racism.

In this paper, we demonstrate how race surfaced in legislation, forensic investigation, and in society while the criminal investigation in the Vaatstra case evolved. We show that race neither materializes in the technologies used nor in the bodies analyzed, but rather it emerges in a relation that comes about in the translation that occurs as humans and things move across sites (M’charek 2013). Drawing on work in science and technology studies, we argue that race is enacted in three different ways: in the context of legislation as biology reduced to bodily characteristics, in the forensic analyses as patterns of absent presence, and in society as a process of phenotypic othering. Addressing the entanglements of legislation, forensic practice, and society in the context of the Vaatstra case, we demonstrate empirically this shape-changing capacity of race.

Over the years, the Marianne Vaatstra case has received considerable international attention and has by now become a paradigmatic case in debates about the regulation of forensic DNA technologies, in particular technologies aimed at identifying the unknown suspect (M’charek 2008; Sankar 2012; Jong and M’charek 2018). ¹ Forensic DNA technology played a key role in solving this case, which took nearly thirteen years, and investigative authorities strongly pushed for wider legal possibilities to deploy the latest forensic technologies. This process indexes an important change in the role of forensic DNA in the criminal justice system, a change from an identification tool aimed at including or excluding a suspect, into an investigative tool aimed at identifying traits of the unknown suspect. In practice, this constitutes a change from a focus on the individual suspect into a focus on the suspect population (see Cole and Lynch 2006). To make this point clear, let us have a brief look at the evolution of forensic DNA technologies.

Evolution of Forensic DNA Technology: From Individual Suspect to Suspect Population

Forensic DNA profiling was introduced in the late 1980s and early 1990s in various jurisdictions in Europe and the United States. Today DNA profiling has not only become a routine and everyday practice but it also figures as the preeminent technology for solving crimes. Adding to its status as a truth machine (Lynch et al. 2008) are innovations such as the introduction of polymerase chain reaction (PCR; Rabinow 1996). PCR makes it possible to compile a DNA profile with very few biological traces, enabling the development of forensic DNA databases and the comparison of a vast number of DNA profiles (Williams and Johnson 2008) facilitated by international agreements to share databases (M’charek, Schramm, and Skinner 2014; Toom, Granja, and Ludwig 2019). All of these possibilities and applications involve comparing individual DNA profiles: the profile found at the crime scene with that of a suspect or that of individuals already in the DNA database. However, a set of technologies has been developed aimed not at comparisons with known individuals but at the generation of clues about the identity of an unknown individual, the suspect. A key example is DNA phenotyping that makes it possible to infer some externally visible characteristics (EVCs) of an unknown individual from DNA. To do so, a collection of markers is targeted consisting of so-called biogeographical ancestry markers intended to determine geographical descent; pigmentation markers for traits such as hair, iris, and skin color; and markers for facial characteristics (see Kayser and Schneider 2009; Kayser and De Knijff 2011; see also M’charek, Toom, and Prainsack 2012; Toom et al. 2016; Wienroth 2018). Crucially, this information does not provide a portrait sketch but rather a descriptive profile that fits a group of individuals. Although facial characteristics might become relevant in specific cases, markers suggesting possible biogeographical ancestry of the unknown suspect often are valuable in criminal investigations. A second innovation, called familial searching, is based on a comparison between the DNA profile of a stain found at the crime scene and DNA profiles in DNA databases whereby partial matches, or near matches, suggest that the suspect (i.e., the person who left bodily traces at a crime scene) is a biological relative of the person in the database, for example, a brother, a father, or a son (Maguire et al. 2014). Familial searching can also be performed through examination of Y chromosomal markers. As the Y chromosome is passed on from father to son, this latter form of familial searching brings into focus male relatives. These forensic genetic innovations thus demonstrate that conventional DNA profiling was primarily aimed at individualizing a subject, and the more recent developments link biological traces lifted from crime scenes to populations and families.

Dutch Forensic DNA Legislation: Race as Biology

The Marianne Vaatstra case led to significant legislation that revised the Dutch regulation of forensic DNA. New legislation changed from restrictive regulation aimed at protecting the rights of the suspect into permissive regulation aimed at empowering criminal investigative authorities. In 1994, the Netherlands was the first country in the world to introduce special legislation for the use of forensic DNA to compare DNA from the crime scene to that of the suspect. At that time, submitting a suspect of a (capital) crime to DNA research was deemed a severe violation of the integrity of the body and of the legal rights of the suspect who, after all, might be innocent. However, technological advancements as well as the successful use of DNA analyses to solve crimes helped to relax Dutch legislators’ earlier restrictive stance. Whereas the 1994 law restricted the use of DNA to capital crimes, such as rape and murder, a 2001 amendment made it possible to use DNA profiling in so-called high-volume crime, such as (car) burglary and theft (M’charek 2008; Toom 2010). This amendment also regulated the compilation and extensive use of DNA databases.

The Marianne Vaatstra case has played a crucial role in the development and implementation of forensic DNA phenotyping and familial searching in the Netherlands. Her murder took place in the proximity of a center for asylum seekers whose residents were mainly from Iraq and Afghanistan. Because the local population viewed them as suspect, and because the police did not have any evidence against the asylum seekers, the latter were looking for ways to exclude the residents as suspects. The police thus contacted the Forensic Laboratory for DNA research inquiring whether it was possible to infer the biogeographical ancestry of the unknown suspect. In April 2000, the laboratory analysis of sperm cells collected from the victim’s body concluded that the biogeographic ancestry of the suspect suggested a probable origin in a northern European or Dutch population, and not the Middle East. However, these results were unlawful at the time and thus their use in the case was prohibited. Nevertheless, the results precipitated amendments to the 1994 law, effectively widening the scope for forensic genetic technologies from identifying individuals to producing a population of interest, a focus for the criminal investigation.

Changes in the Code of Criminal Procedure are often provoked by a problematic, high-profile case. Soon after Marianne Vaatstra was murdered, the Public Prosecutor’s Office started lobbying for a legal provision for DNA phenotyping that tested both the legal and forensic genetic possibilities (Toom 2011). In 2003, an amendment to the Dutch law regulating the use of DNA phenotyping was passed and reads: “DNA research can only be applied to determine the sex, race or other externally visible traits to be pointed out through an Order in Council” (Staatsblad 2003: 1–2). ² How did race become a feature of this legislation, in particular because its use stands in contradiction to other laws? Take, for example, Article 1 of the Dutch Constitution:

All persons in the Netherlands shall be treated equally in equal circumstances. Discrimination on the grounds of religion, belief, political opinion, race or sex or on any other grounds whatsoever shall not be permitted. (Ministry of the Interior and Kingdom Relations 2005: 5)

When the draft legislation was discussed in Parliament, Member of Parliament of the Green Party, Femke Halsema, responded critically to the inclusion of race:

I want to make a remark concerning the comparison that you have drawn with other regulations in the law with respect to race. For there is really something else going on. The provisions on race are so far protective. They prohibit making a distinction according to race. This is, as far as I know, the first Dutch law that will in fact permit differentiation on the basis of race. (House of Representatives 2002a, 5774)

In this debate, the Minister of Justice was asked why he had decided to use the category of race rather than population:

The notion of population is just not specific enough, because therewith also a social distinction can be made. Race refers to physical characteristics and that is what we are talking about here. This is the notion that we always use—also in all other treaties—to indicate physical personal characteristics. This is also the reason why the notion of race is to be found in treaties about racial discrimination, because racial characteristics [raskenmerken] are characteristics which you cannot do anything about. (House of Representatives 2002a, 5776)

Through reducing race to a straightforward marker—immutable physical personal characteristics—the Minister of Justice solved a complicated issue, namely, the legal prohibition of race discrimination. In this seemingly facile gesture, the Minister moved away from the then dominant view in human genetics that there are no biological races and that physical differences are statistically clustered into populations (see, e.g., Serre and Svante Pääbo 2004) ³ Moreover and important for our argument here is that race is enacted as biology and reduced to bodily characteristics. Race as biology, we want to suggest, did not happen as a simple misreading of legal texts. The Minister himself had indicated that:

The Dutch law does not have a definition of the concept of race. This concept is primarily used to determine in which cases citizens should be protected against discrimination. To this end […] the concept is broadly defined and also includes: skin color, ancestry and national or ethnic descent. (House of Representatives 2002b, 7)

The juxtaposition of DNA, EVCs, and the biogeographical origin of an unknown person to the broad definition of race in legal treaties have provoked a translation of race, thereby reducing race to biology. This work of translation also played a crucial role during the forensic investigation. So let us return to the Marianne Vaatstra case. ⁴

The Forensic Investigation: Race as a Pattern of Absent Presence

At the crime scene where Marianne Vaatstra’s body was found on May 1, 1999, the police forensic team and the coroner secured thirteen biological traces found on and around her body. These traces, including blood, pubic hair, and sperm, were sent to the Netherlands Forensic Institute. Of the large amount of forensic research that was conducted in the course of this investigation, we highlight three analyses aimed at finding clues about the identity of the unknown suspect. As will become clear, these forensic techniques produced different accounts of identity that interacted and resulted in the production of race.

Hair Analysis

Pubic hair was among the several traces secured from Marianne Vaatstra’s body. These hairs were anthropologically analyzed to produce an estimate about the ancestry of the unknown suspect. Microscopic analysis examines variations in human hair based on markers such as color, pigmentation distribution and density, and shaft form (Ogle and Fox 1999, 12). However, given the high variation of these characteristics between individuals, “[e]ven combining a number of characteristics does not reduce the potential pool of donors of the hair to a single person” (Cole and Duster 2016, 31). Thus microscopic hair analysis does not constitute an identification tool, it does not individualize but instead produces rough probabilistic clustering of the unknown person in a racial typology (Smith and Goodman 1996; Cole and Duster 2016). An example of such racial typologies can be found in a detailed study on the physical (microscopic) comparison between hairs and their forms in the Atlas of Human Hair by Ogle and Fox (1999). Referring to this study as an atlas already suggests that hair characteristics are indicative of someone’s geographical origin. This is further reified by the use of three geographical categories in the section where they discuss pubic hairs of “East Asian, African, and European” origin:

Individuals with East Asian heritage have pubic hair that is typically black and wavy. Individuals with African heritage have highly curled pubic hair, with a form near that of the scalp hair. The pubic hair of African heritage individuals is typically black in color but may have other colors owing to admixture with individuals of European descent. Individuals with European heritage have pubic hair with a wide variation in form that may be curly and convoluted but lacks the wavy form seen in East Asian pubic hair or the high degree of curl seen in the African pubic hair. The color of European heritage individuals’ pubic hair ranges widely and is correlated with the scalp hair color. (Ogle and Fox 1999, 60-61)

While Ogle and Fox use the categories “East Asian, African, and European” in forensic practice, in case reporting these categories are usually translated into “Asian, Negroid, and Caucasian.” ⁵ In June 2000, the Office of Public Prosecutor announced that the hair analysis conducted on the pubic hairs indicated that the “perpetrator does not belong to the negroid or Asian race” (Openbaar Ministerie 2000a), suggesting that he belonged to the “Caucasian” race. While the category “Mongoloid” is not common in the Dutch practice, Caucasian and Negroid are used in Dutch forensic anthropology. ⁶ To be sure, the conclusions presented by the public prosecutor are highly probabilistic and can only be used as leads in an investigation. Yet they were nevertheless important because they suggested that the perpetrator was probably not someone living in the center for asylum seekers. Microscopic hair analysis helped to exclude the individuals suspected by the local population in the north of Frisia, but it also produced a particular version of race. Race, through the analysis of hair, is enacted as typological difference. The long history of physical anthropological research aimed at determining racial types (Stocking 1990) is folded into the current forensic practice of hair analysis.

DNA Research on Y Chromosome

The biological material left by the perpetrator on Marianne Vaatstra’s body provided valuable sources of DNA. Based on the sperm cells, a DNA profile was compiled and uploaded to the national DNA database of the Netherlands to search for a possible match. In absence thereof, the DNA was further subjected to a new forensic DNA technology concerned with the inference of EVCs. From the mid-1990s onward, publications appeared on frequency distributions of Y chromosomal DNA collected from 3,825 unrelated males from forty-eight subpopulation groups from Europe, America, Asia, Africa, and Oceania (De Knijff et al. 1997). Results of that research demonstrated that Y chromosomal short tandem repeats (STRs) were suitable for establishing male lineages (Kayser et al. 1997) and for determining the ethnic origin of a person (Jobling, Pandya, and Tyler-Smith 1997). In subsequent years, a database was created to catalog Y chromosomal STRs from unrelated males from the globe’s many different (sub)populations. This database, the Y Chromosome Haplotype Reference Database (YHRD), had been maintained for several years when Marianne Vaatstra was murdered. One of the YHRD project leaders, Dr. Peter de Knijff, a professor of population genetics, director of the Leiden Forensic Laboratory for DNA Research, was an expert witness in this case. He was approached in the Fall of 1999 by one of the criminal investigators involved in the Marianne Vaatstra murder case, who requested De Knijff’s assistance. The investigator asked whether it was “possible to generate a clue regarding the geographical descent of the perpetrator by means of DNA research” (De Knijff 2006, 2-3). Although the DNA law, introduced in 1994, did not allow for the inference of geographical descent from a crime stain sample, De Knijff (2006) agreed to the request.

The police had heard that we had been working on such technology. If that research would indicate that the perpetrator was not an allochtoon [migrant] but, for example, a Fries [Frisian, autochtoon], then there was a fair chance that the anxiety would fade away. The alternative scenario, it was an allochtoon, was something I was not willing to consider. (pp. 2-3)

He would later recount that he saw this as an act of civic disobedience to help a good cause, namely, to alleviate the burden placed on the residents of the asylum seekers center—an act of anti-racism (De Knijff 2006, see also Bliss [2015] for other examples). Based on the sperm DNA, an YSTR profile consisting of eleven markers was compiled. In April 2000, the laboratory had drafted the case report and sent it to the public prosecutor. The report explains that the DNA profile was compared to the YHRD database which at that time consisted of more than 3,500 males of western and central European descent. This profile was also compared to YSTR profiles of males of Kurdish descent (98), North-African (183), and Central-Asian (172) descent. This comparison had shown a match (for all markers) with three profiles in the YHRD database, males of German origin. ⁷ The report elaborates further:

This Y type fits within a group of Y types that are common in the Netherlands. This supports our theory that it is probable that the Y type originates from a Dutch man. If I would e.g. add general Kurdish or North-African profiles in this figure [network] they would clearly differ from the groups of Dutch profiles. ⁸

A few months later, the public prosecutor announced that, based on the Y chromosomal research, “it is ‘very plausible’ that the originator is from Dutch or northwestern European descent” (Openbaar Ministerie 2000a). The continual focus on geographical descent in the forensic genetic research is remarkable with reference to race. The microscopic hair analysis helped to enact race as a typological difference and to bring about racial types. By contrast, in the genetic research, race seems to disappear from view and difference becomes a matter of probability and geographic descent (see also Fujimura and Rajagopalan 2011).

However, the laboratory results that were based on a statistical analysis of geographic descent were translated overnight in the media. The newspaper headlines reported: “Murderer of Marianne is white man in the vicinity.” ⁹ This work of translation was not specific to the media but happened in various sites in the forensic work. We will therefore briefly consider the psychological offender profiling.

Offender Profiling

Thus far, the unknown suspect was enacted as a male who did not belong to the “Negroid or Asian race” and was likely to be of Dutch or northwestern European origin. Yet a third perpetrator identity was established based on offender profiling, which is a behavioral psychological assessment of the nature of the crime. Deducing the identity of the suspect based on the modus operandi is a common part of the routine of criminal investigation. For example, the coroner examining Marianne Vaatstra’s body suggested that based on the specificities of the homicide, especially the fact that her throat was slit, the suspect “might well be a foreigner (‘buitenlander’). It did not look Dutch to me, have never seen such a thing in my whole career.” ¹⁰ While this observation was made in mid-1999, another offender profile was broadcast on Dutch national television on June 27, 2000. According to the latter offender profile, the murderer was likely to have extreme fantasies about sex, violence, power, and control, be between twenty to forty-five years, live within a fifteen-kilometer perimeter of the crime scene, and be a white male (Openbaar Ministerie 2000b). It is not trivial that the offender profile was published two weeks after the media reports about the YSTR profile, which hinted at the Dutch geographic descent of the unknown suspect. It is crucial that the Dutch word for white that was used in the media as well as for the offender profile was not “wit,” a more or less technical term for the color white, but “blank,” which is a normative and aesthetic word, connoting shiny, goodness, and innocence. It is a racial category, particular to the Dutch colonial history and racial practice (see Wekker 2016). Thus, the specific translation from geographic descent into a white/blank offender not only produced correspondence between offender profiling and DNA profiling but, crucially, situated the forensic work in a Dutch racial discourse.

We are not suggesting that all genetic probabilistic knowledge will necessarily be translated in such a way. Rather, our point is that translation is part and parcel of forensic work. Across the chains of evidence and custody, from the crime scene to court, multiple translations must happen to make forensic evidence relevant to the criminal investigation and crime solving (see Bal 2005; Lynch et al. 2008; Kruse 2015; Van Oorschot 2018). And it is in such processes of translation that even forensic practices indifferent to race might contribute to racialization. Our analyses underscore that race is a pattern of absent presence (see Law 2004; Jong and M’charek 2018; M’charek 2020). We have encountered it in the practice of microscopic hair analysis as a racial typology with its specific history in racial physical anthropological research; in the practice of forensic genetics with its genealogy in population genetics and its probabilistic approach to population differences where race seemed to have become irrelevant; and then race appears with the word “blank” as a Dutch colonial classification, in the offender profiling practice. The different difference-making practices do not map onto each other but are interrelated through the very practice of forensics and thus produce race as a (topological) pattern of absent presence (M’charek, Schramm, and Skinner 2014).

The “Marianne Vaatstra” Murder Case: Race as Phenotypic Othering

In the previous sections we examined the “multiplicity” of race (Mol 2002) by taking a closer look at the practice of legislation and that of forensic research. Whereas race was reduced to biology by simply taking the body as a matter of fact in legislative practice, a more complicated pattern emerged in forensic practices. Different technologies produce different versions of race and recent forensic genetic technologies might be said to be indifferent to race, leading us to conclude that race is rather a pattern of absent presence. In this final section, we want to add to the complexity of race in practice by attending to how the murder case was engaging local citizens and other actors outside the legal and forensic practices. We use the notion of “phenotypic othering” as an analytical device to demonstrate how a different kind of suspect population was generated in Dutch society. The phenotypic other is an effect of processes of othering in which differences in appearance play a crucial role. Importantly, these differences are not immediate differences that belong to bodies out-there. The phenotypic other hence goes:

beyond the somatic body and include[s] markers such as hairdo, dress, or beard style. Thus, differences ascribed to specific bodies that are deemed to belong to targeted groups are made visible and readable in specific sociopolitical contexts through specific scientific and technological practices. (M’charek, Schramm, and Skinner 2014, 471)

Phenotypic othering then is specific to place and time, but it also assumes an attention and a skill to read differences. Bodily features of those who are othered become readily visible and readable. An example would be “the Muslim other” in the post-9/11 terrorist threats era: “we” have become educated to read these bodies by linking certain markers to certain bodies that are deemed to belong to targeted groups. Whereas such targeted groups and their bodily markers seem to be “in your face” and easily readable, other groups, no matter their skin color, lifestyle, or religious inclination, can remain unnoticed. Thus, phenotypic othering is not about the differences associated in certain bodies but about the differences that are made relevant or urgent to act upon. We here mobilize empirical material to advance the notion of the phenotypic other and demonstrate its relevance for an analysis of the dynamics of racism.

The Center for Asylum Seekers

Marianne Vaatstra’s body was found in a pasture not far away from the center for asylum seekers near the village of Kollum. The proximity of the center was mobilized by the local village population to cast suspicion on its residents: the perpetrator is not one of us, but one of them—a resident of the asylum seeker center must be the murderer. ¹¹ In addition to the location of the center, the manner of death—slit throat—entertained the imagination of many, to become a central device of phenotypic othering. Across various media, commentators pointed to the method of murder as revealing key information to the ethnicity of the perpetrator. In a newspaper article, an interviewee who was said to make a slicing gesture along the throat was quoted saying: “This is not part of our culture. And hence people say: [the perpetrator lives in] the center for asylum seekers.” ¹² Slitting a throat was also framed as “a non-Dutch manner of murder” by, among others, the late populist politician Pim Fortuyn. ¹³ The key role of the slitting of a throat as a device of phenotypic othering was expressed twelve years later in a TV show of the famous Dutch crime reporter Peter R. de Vries. Revealing new insights about the manner of death, he articulated the misleading, yet commonly held belief about the murderer:

The perpetrator was well prepared. Like a predator looking for a prey he was waiting to attack Marianne from the bushes. After that, he killed her by cutting her throat. Given this modus operandi, the suspect must be an inhabitant of the asylum seekers center. ¹⁴

In the local newspaper Dagblad van het Noorden of October 30, 2000, a reader commented that slitting a throat resembles Arabic and Islamic ritual slaughter practice. The fact that this claim did not require further explanation suggests that it resonated with a tacit notion of a threatening phenotypic other in Dutch society.

In this context, the mayor of Kollum issued a press release in which he stated that rumors about involvement of residents of the center were not based on any facts—forensic or otherwise. Despite the statement by the mayor, it was reported that the local population threatened to tear down its walls if the perpetrator lived in the center. ¹⁵ In this rather explosive situation, the Kollum municipal council decided to grant the center for asylum seekers permanent status. At an information meeting about the new center, protestors pelted the mayor with eggs and referred to the center as “a hotbed of human trafficking, drug dealing and other criminal activities.” ¹⁶ This evening marked the start of an ongoing mood of discontent and distrust between a very vocal part of the local population on the one hand, and the police, Kollum administration, and center for asylum seekers on the other.

The Symptomatic Ali

Although they had no reason to focus on the asylum seekers, the police singled out one resident as a person of interest. Although he left the center already before the murder, Ali H. became vox populi’s perpetrator. Although the police had determined that Ali H. had already left the Kollum area before Marianne Vaatstra was killed, as an effect of the media pressure, the police had officially marked him as a suspect issuing an international arrest warrant for him in July 1999. ¹⁷ The announcement reinforced the incrimination of asylum seekers and migrants; the suspect status of Ali H. thus came to legitimate the criminalization of a whole group.

Ali H. was arrested in Istanbul in October 1999. His DNA cleared him from suspicion. “Disappointment dominates in Kollum,” a national newspaper announced. ¹⁸ Although Ali H. was now exonerated, a substantial segment of the local population persisted in believing a resident of the center for asylum seekers committed the rape and murder of Marianne Vaatstra. One of the protesters in Kollum questioned the arrest: “this is either a big mistake or a tactical move. I wonder whether they arrested the right Iraqi.” ¹⁹ Such allegations were not resolved by the YSTR research indicating that the perpetrator was most likely of northwestern European descent. For example, Marianne Vaatstra’s father, Bauke Vaatstra, said in a documentary entitled Afscheid van Kollum [Farewell to Kollum]: ²⁰

If the murderer of our daughter is not caught, never is going to be caught, for everyone, including for us, it will forever be an asylum seeker. I don’t mean to say that it is one, but I want to point out that if he is never going to be caught, he will always be one [of them].

Others pointed out that some residents seeking asylum were not excluded by the DNA results referring to that the suspect is probably of northwest European descent. Potential suspects could originate from Northern Africa or former Yugoslavia or so was reported by the Dutch RTL broadcast company in 2009.

Ali continued and continues to figure as the murderer in conspiracy theories about the Marianne Vaatstra case, suggesting that the police had arrested “another Ali”—the wrong Ali—in 1999. The “Ali” who had been investigated by the police was much taller and slimmer than the Ali who was supposedly involved in the murder. In an attempt to put an end to the controversy, the public prosecutor published a statement in January 2011 in which they presented an explanation of “how Ali became short and fat” (Openbaar Ministerie 2011). Ali obviously did not stand for himself, an individual, but signaled and figured as the condensation of a whole group and associated views about it; a group that in the context of the Marianne Vaatstra case could be criminalized and excluded. “Ali” thus became the symptomatic Ali. Through him the suspicion was directed toward a whole group; a group that was phenotypically othered through characteristics ascribed to the unknown perpetrator.

Peter R. de Vries and the Arrest of Jasper S.

In 2007, the Office of Public Prosecution commenced a new enquiry in the Marianne Vaatstra case. Results of that enquiry were presented by crime reporter Peter R. de Vries on national television in May 2012 and included a new scenario about the unknown suspect. Instead of a “well-organized predator who had been waiting for his prey,” the new scenario portrayed the perpetrator as an “occasional offender” because he seemed unprepared, which was evidenced by him carrying only a pocketknife and leaving many traces. ²¹ Moreover, the reanalyzed traces from the field suggested that Marianne Vaatstra might have voluntarily walked into the field with her murderer, possibly before things got out of hand. Based on this scenario, the Office of Public Prosecution announced that it was planning a familial DNA mass screening in September 2012. This episode of Peter R. de Vries was instrumental in mobilizing the men in north Friesland to participate by donating their DNA.

Journalists went to Friesland again and reported about the case, focusing in particular on the supposed new perpetrator identity, the planned mass Y chromosomal familial DNA screening, and the reactions of the local population. In one report, a journalist narrates her meeting with four men in a local pub:

It seemed logical, they think. Jan: “It is one of us.” For a moment it becomes silent at the table. One of us. It sounds sinister. Jan: “I think that he was closely involved with Marianne. That he is very jealous. That he is a smart guy.” Piter: “Otherwise he wouldn’t have slit her throat.” Jan: “Because of that everyone thought it was a foreigner.” ²²

Slitting the throat here becomes a strategic move of a smart guy; one of us who had intended to make people believe that the perpetrator was a foreigner. This allows for both configurations—slitting a throat as a nonwestern way of killing with popular suspicion put on the phenotypic other, and the investigative scenario in which the analysis of the perpetrator DNA suggests that he is of northwestern European descent—to coexist in the conversation. This bar conversation illustrates how the reality of phenotypic othering in the Marianne Vaatstra case existed parallel to the reality of forensic investigation that we sketched earlier. Importantly, despite their intention to defeat the racialization and criminalization of a social group, the forensic facts did not unsettle the reality of phenotypic othering.

Before going into the response to the conclusion of the case, let us briefly pause with the DNA familial searching technology. In May 2012, a government bill allowing familial searching came into effect and the Marianne Vaatstra case was the first Dutch case in which this technology was employed. The traces left behind by Marianne Vaatstra’s murderer were presented as particularly apt for this technology. They showed relatively rare genetic characteristics found exclusively in Europe and underrepresented in the Dutch DNA database. Forensic experts suggested that these characteristics might be more common among the Frisian population. ²³ This led one person to refer to the samples as a “Golden Rolls-Royce.” ²⁴

Familial searching in the Dutch legal system can be deployed in two different ways. The first is in accordance with a classical DNA database familial search action where the DNA trace is compared with DNA profiles of known individuals to search for genetic relatives, in particular brothers, fathers, and sons. But in the Marianne Vaatstra case, this did not lead to any near matches that indicated genetic kinship between the trace and the profiles of known individuals included in the database (Kal 2012). The Office of Public Prosecution subsequently decided to organize the second type, familial DNA mass screening, in September and October 2012. A population of 8,080 men who at the time of the crime were between sixteen and sixty years old and lived within five kilometers (3.1 miles) from the crime scene were asked to volunteer a biological sample (i.e., saliva). Almost 90 percent of the men complied (Openbaar Ministerie 2012).

On Monday, November 19, Peter R. de Vries tweeted: “Man arrested. White Blanke suspect. Frisian, lived 2.5 km away from crime scene. 100% DNA-match!” ²⁵ The accused Jasper S., a local farmer from the Frisian village Oudwoude, remained unsuspected for thirteen years yet participated in the familial DNA mass screening. The people from his village were shocked when he was identified as the suspected murderer. A headline in a local newspaper stated “The suspect is white blank, so not an asylum seeker.” ²⁶ A woman reacted: “When everybody here believed that the perpetrator was a resident of the asylum seeker center, at least that gave us some peace. You don’t want it to be a father from here.” ²⁷ The local community was in dismay. Jasper S. was portrayed as a hardworking farmer and father, embedded in the small community life through his occupation and ties with the church:

Well, DNA doesn’t lie, mumbles Nycklo de Vries. But it is still hard to believe this. He knew the man who was arrested. Just like everybody here in Oudwoude. A very normal, social man. With quite a bit of land and cattle. ²⁸

But few in the village can paint a picture of him. Tall, small, long or short hair, big or lean, sporty, socially engaged. “Well, what does he look like?” Churchgoing [“Kerks”], ²⁹ just like his parents. ³⁰

After his arrest, neighbors recalled an incident of joyriding for which the farmer was convicted in 2009. Jasper S. was then allegedly diagnosed with “dissociative fugue,” amnesia, hereby limiting his accountability for the crime.

Psychologizing Jasper S. underlines his individuality. In contrast to Ali H., he kept being treated and seen as a person, a farmer, a man with a family, a churchgoer, friendly, a normal man whose appearances are hard to describe. Rather than expelling his family from the village or tearing down their house, the dominant response was one of compassion. “Oudwoude won’t let the family of the suspect down” read the header of a national newspaper. ³¹ They received an overwhelming amount of supportive messages from people in the village community. The contrast between the mood after the DNA match and before, when an asylum seeker was still believed to be the murderer, could not have been stronger.

The rarity of the DNA profile, the DNA mass screening for familial searching, and the geographical delineation of the population of interest have together contributed to the racialization of a Frisian community. However, as we have shown, this racialized group identity could easily dissolve, making room for individuality. By contrast, race as a process of phenotypic othering subsumes the individual in an allegedly homogenous group. In our example, phenotypic othering materialized in the figure of the symptomatic Ali who folds together different elements—the knife, the center, physical appearance, religion—to produce a racialized category. The elements that help to configure the symptomatic Ali are specific to this case. In a different context, a knife does not necessarily make race neither does religion. Moreover, the phenotypic othering arose from not only the suspicion placed on Ali but also the continued doubt that the right Ali had been exonerated. This phenotypic othering contributed to the consolidation of the identity of the other as a group.

Discussion

In the post WWII era, the United Nations Educational, Scientific, and Cultural Organization statements on race and the Human Genome Project (HGP) critically interrogated the scientific evidence for the concept of biological race (M’charek 2005; Lipphardt 2012; Selcer 2012). The HGP articulated the sameness of human beings by stressing the 99.9 percent commonality in the genes. ³² Yet subsequent research focused attention primarily on the 0.1 percent difference. This interest in difference is at the heart of current forensic genetic technologies (Kayser and De Knijff 2011; Kayser 2015). To unravel the multiple ways race is made relevant in practice, we have focused on the Marianne Vaatstra case in which novel technologies and legislation were pioneered.

The focus on difference in genomic and genetic research has attracted attention and critique. An important strand of critique has shown convincingly that, despite the promise of commonality in the gene, genetics is contributing to a “genetic reinscription of race” and to the molecularization of race (Duster 2006; Skinner 2006; El-Haj 2007; Fullwiley 2007). Although this process of molecularization of difference is highly important and will require ongoing attention in the context of big-data and data-mining endeavors, here, instead of attending to the molecular, we remained at the surface, as it were, to map the spatial distribution of race across practices. Based on this forensic case, we examined race in the practice of legislation, the forensic investigation and in society where the case had a considerable impact.

In the practice of legislation, race came about as a biological and physical characteristic. In parliamentary debates about the Dutch legislation for the inference of EVCs of the unknown suspect based on DNA, race was treated as a matter of fact, written on the body. It is that which is unchangeable about a person’s appearance, “characteristics which you cannot do anything about”––biology. However, in the practice of forensic investigation, it was precisely the biology of difference that was problematized, as was the delineation of race. There, race was better attended to as a pattern of absent presence. Whereas microscopic hair analysis relying on the historic racial knowledge of physical anthropology helped to enact race as a typological difference, DNA profiling has a provenance in population genetics that contributed to the trivialization of race. Via behavioral psychology, race came about through the language used in the offender profile, in particular with the Dutch colonial classification blank. Finally, attending to race in society, especially to how the Marianne Vaatstra case moved the people in Frisia, and how they engaged with it, we have shown that race emerged as part of a process of othering. The case contributed to an antagonistic and hostile relation between the local population and the residents of an asylum seekers center near the crime scene. Focusing on how suspicion was placed on what we have called the symptomatic Ali, and the contrast between this suspicion and the disbelief that the murderer was Jasper S., we have argued that race has come about as phenotypic othering. Crucially, phenotypic othering works through a focus on the body but involves the work of connecting and stitching different elements to bodies. In the case, elements including the knife, the center, physical appearance, and religion helped to produce a racialized category.

Race, then, is not solid nor will it melt into air. Its ghostly presence is volatile precisely because it is not singular. It does not wander about as a pregiven entity because it is, as M’charek (2013) has said it, a material semiotic relation. While moving between practices, race continues to be configured differently, revealing its capacity to shift and change. This simultaneous movement and change constitute the politics and power of race.