Equal before the Law

On Equality

“All are equal before the law and are entitled without any discrimination to equal protection of the law.” Thus reads article 7 of the Universal Declaration of Human Rights. In this article, we argue that equality is not an abstract principle on which a criminal justice system is based but rather a goal that must be practiced during the criminal investigation and legal process. In a similar vein,  Van Kampen argues in her study of expert evidence under the European Convention that the concept of fairness does not so much depend on the specific rights of the defendant laid down in this Convention, but rather on the “procedure as a whole” (2000, 186). “Fairness cannot be defined outside the dialogue between the affected parties” (Brants and Field 1995, quoted in Van Kampen 2000, 187). More in general, “it is the method of resolving the conflict, more than the law itself, which seems fundamental—the judge more than the law, and criminal procedure more than the criminal law” (Soulier 1995, quoted in Van Kampen 2000, 188). Enacting equality is a matter of methods, procedures, and the actors involved rather than the principles of law. Equality is thus better seen as a practice.

European continental law is inquisitive: equality before the criminal law is concerned more with the preconviction phase than the conviction phase, as in US law (Whitman 2009; for the difference between these legal systems, see below). The focus is less on the equal threat of punishment and more on the truth-finding process and equal threat of investigation (Whitman 2009). In this context, standards in legal procedures and the investigative process are crucial to create equality before the law in at least two respects. First, standards specify what sorts of evidence are allowed. Second, standards specify in technical terms how such evidence should be obtained and prepared for use in the courtroom. To establish equality before the law, “sameness” between investigative practices, with respect to principles and rules regarding the collection and analyses of materials, needs to be established.

In the N-case discussed below, the charges against the suspect are rape and murder. Identifying the suspect as the perpetrator, proving guilt, must be “beyond reasonable doubt.” As we show, sameness had to be established between two laboratories that examined the DNA to create legal equality. Only then could the suspect be identified as the originator of the biological trace found at the crime scene. This article thus attends to the making of various kinds of sameness and asks, what is the machinery of sameness and how does it work?

Introducing the N-case

On July 2, 1993, the body of a seven-year-old boy was found in the Maaswaal Canal near the city of Nijmegen in the Netherlands. The boy had been missing a couple of days. Upon the discovery of his body, the public prosecutor requested an autopsy in the Netherlands Forensic Institute (NFI). ² Besides bruising suggesting sexual abuse, sperm cells were found in smear samples from the boy’s mouth. The trace samples were in a bad condition since the body had been in water, ³ prompting the question whether they would make it possible to produce a DNA profile of the perpetrator.

One month later, a suspect was arrested. He denied any involvement and voluntarily submitted blood samples for DNA analysis. Although the materials had already been send to the NFI, the defense consented to send the samples to the Institut für Rechtsmedizin der Universität Münster (Lab M) ⁴ because it was developing a new method to extract DNA from mixed biological samples such as those found in the N-case. Such a method could help establish whether or not there was a match between the suspect and the material found on the victim.

A link had already been established between suspect and victim, based on eyewitness accounts and other circumstantial evidence. Children who had been playing with the victim testified they saw the suspect walking with him the day he went missing. Another witness had seen the suspect in the neighborhood a couple of times on previous days. Other evidence came from CCTV recordings of a cash machine in the neighborhood, which the suspect had used the day the boy disappeared. The CCTV recordings confirmed the children’s description of the suspect (color of clothes). However, these circumstantial accounts did not add up to a direct link between perpetrator and victim at crime scene. DNA profile comparison would be the only way to include or exclude the suspect. DNA evidence thus became the key “witness” in the criminal investigation. However, it would take more than two years before Lab M could deliver its results of the DNA analysis.

Science Meets the Law

In her Science at the Bar, Jasanoff (1995) shows that facts and expert knowledge are not beyond contestation when they enter the courtroom. She examines a number of cases to show that in the process of debating the scientific facts, scientific knowledge and legal facts are coproduced during the legal process. “At their most effective, legal proceedings have the capacity not only to bring to light the divergent technical understandings of experts but also to disclose their underlying normative and social commitments in ways that permit intelligent evaluation by laypersons” (Jasanoff 1995, 215). ⁵ Putting scientific facts and the role of expertise to the test contributes to public understanding of science and technology and the closure of controversies.

Jasanoff argues that in coproducing scientific expertise in legal procedure, the inquisitorial model’s “major drawbacks are the uncritical acceptance of mainstream views, both judicial and scientific, and the failure to acknowledge or question the socially constructed aspects of scientific testimonies” (Jasanoff 1997, quoted in Bal 2005, 54). Yet Roland Bal (2005) has shown that an inquisitorial system may take on adversarial aspects in how it deals with forensic knowledge. Technologies may become controversial, opening up scientific facts and questioning the practice and process that helped to produce them. ⁶ Even in a relatively calm environment as the Dutch inquisitorial approach considerable work is required to produce and check the relevant facts and to produce sameness.

Making a DNA Profile

On August 6, 1993, Lab M received one sample from the suspect, one sample from the victim and two swabs taken from the victim’s mouth containing a mixed sample from victim and perpetrator. On the basis of microscopic analyses, the swab that contained the most biological material was selected for the complicated analysis ahead. The other swab was stored and could be used if counterexpertise analysis would be requested at a later stage.

 Before producing the DNA profiles from the obtained material, researchers at Lab M first wanted to develop and test novel DNA extraction technology to increase the chance of success when applied to the case at hand. ⁷ However, it took Lab M more than two years to routinize the technique. ⁸ Only in February 1996, its report reached the NFI. Lab M reported that according to their analyses, the suspect could not be excluded; there was a match between the suspect’s profile and the biological trace (sperm cells) found on the victim. The Münster lab expressed the matching likelihood as a ratio (1:165,000) of the chance that any random individual in the reference population could have left that biological trace at the crime scene. ⁹

 In the Netherlands, the NFI compared the DNA profile to a “white Dutch population” and adjusted the matching likelihood (to 1:45,000), arguing in its report and testimony that the statistical difference between the two figures was negligible. The difference had to do with the genetic markers used to produce the DNA profile. These showed specific clustering in different populations. Since the German reference population was far larger than the Dutch reference population, it was important to adjust the probabilities. The Lab M report, containing additions made by the Dutch forensic scientists, was sent to the Dutch Public Prosecutor responsible for the N-case.

On Identity

To ask what is a DNA profile amounts to asking what is identity. Just like any other piece of evidence, the profile links places, humans, and events to each other. But the profile is also an identification tool aimed at establishing a relation between a biological trace and a specific person, the suspect. A profile is a collection of genetic markers that together are said to be unique for a specific individual given the estimated low probability that more persons in the same population might have the same set of marker characteristics. In laboratory practice, a marker is not simply the object of research but also the technology to visualize it and a methodological tool to study it. Let us pause briefly with these three features of a marker for they are instructive in how identity is enacted in forensic practice.

First, the textbook definition: “an identifiable physical location on a chromosome whose inheritance can be monitored” (Kevles and Hood 1992, 381, emphasis added). The definition makes clear the interest of genetics in the way marker fragments are passed down from one individual to the other. But it especially underlines that a marker is in the DNA. It is thus an object of research that scientists can uncover. But this is only the beginning.

As STS scholars have shown, many interventions are necessary to establish and study a research object in the laboratory (Latour and Woolgar 1979; Rheinberger 1997). In everyday genetic practice, extracted DNA first has to be copied. There are various ways of doing this but since the 1990s, the Nobel prize-winning polymerase chain reaction (PCR) technology has become the standard (Rabinow 1995). In contrast to other techniques, PCR makes it possible to study even very small biological samples and has thus contributed to the prominent role of forensic DNA in criminal investigations (M’charek 2008a; Toom 2011). In PCR, various chemicals are added to the solution containing the extracted DNA, such as chemical groups (nucleotides) that will make up the DNA building blocks (A, C, T, and G) of the copies, and thermostable enzyme polymerase, primers (needed to copy fragments of specific interest) as well as chemical groups aimed at visualizing marker fragments. From DNA extraction onward, handling is aimed at visualization. To produce a “selective perception,” DNA has to “be upgraded” (Lynch 1990). This entwinement of DNA and technology suggests that a marker is part and parcel of the visualizing technology used in the laboratory. Not merely an object of study, the DNA marker is also the very technology used to visualize that object.

To illustrate the point of a marker as a methodological tool, let us go back to its role in the production of a DNA profile. The N-case used six markers. A DNA match is not enough to identify a suspect. It still has to be compared with a population with respect to the distribution of marker variation. The goal of comparison is to determine the matching-likelihood probability that a randomly chosen unrelated individual in the population could carry the same profile. The matching-likelihood probability of 1 in 165,000 that Lab M delivered in the N-case was based on a comparison of the suspect’s profile and reference population. Markers must be polymorphic (contain different alleles) ¹⁰ to calculate matching likelihood. This does not mean that all individuals differ for all genetic markers. For marker X, they may carry different alleles, whereas for marker Z they may carry the same allele. Applying several markers, say six, may result in an individually specific profile.

Differences and similarities for a given set of markers are crucial to establishing the individuality of a DNA profile. This accentuates the importance of the reference population. First, comparing a DNA profile of an individual from Central Africa to a German population may lead to biased matching likelihood if that individual carries an allele that is common in the population of origin but rare in the German population. Second, the polymorphism of a genetic marker is theoretically investigated. A genetic marker with only two alleles spread evenly throughout the population might be interesting for forensic purposes. Yet another genetic marker, also with two alleles, but where almost all individuals carry the same allele, may not pass. In the second case, the chance is greater that two individuals will look alike than in the first. Therefore, a genetic marker should reveal a specific rate of differences and similarities contributing to the goal of forensic identification, and to the statistical models applied for that purpose. ¹¹ To put it differently, good genetic markers should contribute to the analysis of what they reveal, identity. A genetic marker is then not just an object of research and the technology to visualize it, but very much a methodological tool as well.

This brief excursion into the concept of a genetic marker shows that identity does not inhere in someone’s DNA. Rather, it is best seen as a spatial relation established between obtained biological material, various chemicals, technical devices, procedures, and statistical models. The established identity is not ready or stable. Once a profile starts moving from the laboratory to the courtroom, many more entities will become crucial to establish or maintain identity.

The N-case: Changes in the Legal Setting

Initially, the suspect had been held in custody but this was suspended at the request of the defense when a court hearing in August 1993 made it clear that the results of the DNA analysis would not become available before 1995. Although it initially accepted the delay, the defense would later try to get the case dismissed on the basis of “reasonable term.” This legal principle states that criminal investigations should conclude within a reasonable time, generally determined on a period of two years. According to the defense, the prosecution’s “inactivity” was in conflict with the terms of the European Court for Human Rights and the presumption of innocence. ¹² The court honored the plea to lift custody until Lab M could deliver the results of the DNA analyses, but not the claim to dismiss the entire case. ¹³

 Meanwhile, in September 1994, a new law regulating the use of forensic DNA in court proceedings came into force in the Netherlands. ¹⁴ The defense used this to question the admissibility of the DNA evidence that was expected to come from Lab M on three counts. First, the defense argued that according to the new law, only labs appointed through an Order by Council are entitled to conduct DNA analyses. ¹⁵ Only two Dutch labs, the Netherlands Forensic Institute (NFI) and the Forensic Laboratory for DNA Research (FLDO) in Leiden, were authorized in accordance with the law by the Sterlab/Sterin Board of Accreditation. The defense argued that according to the law, DNA evidence should be delivered by either one of these labs and, furthermore, that the lab conducting the research should also be accredited. Lab M in Münster was not accredited and its conduct had not been subjected to the standardizing procedures of the Board of Accreditation. This put the admissibility of evidence in question. Although the defense had originally consented to sending the samples to Münster, in February 1996 it requested that an independent expert from the other Dutch accredited institute (FLDO) should review Lab M. The court agreed and appointed an expert from FLDO to do the review.

 The second objection concerned the responsibility for the DNA part of the criminal investigation. In the N-case, DNA research was the responsibility of the Public Prosecutor while the (new) law prescribes that the Examining Judge should lead the investigation. ¹⁶ The defense drew attention to this as a procedural problem, but did not ask to pass the case to the Examining Judge as suggested by the court.

 The third objection concerned the two-year adjournment. The defense argued that the prosecution evidence was at best circumstantial. Despite having the suspect’s full cooperation, not obligatory at the time, the prosecution had failed to come up with solid evidence within a reasonable time. The dismissal of custody in 1993 attested to the circumstantiality of the evidence. Thus, the defense pled for dismissal of the case on the basis of a transgression of principles of “reasonable term.” The court did not grant his request.

The N-case: Problems with the DNA Evidence

On February 13, 1996, Lab M finally sent its report to the NFI. After reviewing the results, the NFI compiled a Dutch-language report which was presented in court on February 29, 1996. Since the results suggested a DNA match, the court decided to take the suspect into custody. Yet, the queries about the admissibility of the results under the newly enacted law prompted two questions: could the results produced in Münster be considered scientifically sound and would they be legally admissible. The court decided to hand over the case to the Examining Judge and request a review of Lab M: “Contrary to what the defence counsel  suggests, it cannot be stated that the DNA research conducted in Germany should be considered unlawful. Further investigation is needed by the Examining Judge.” ¹⁷ In other words, could Lab M be considered similar to the accredited Dutch labs and would it comply with the standards required by Dutch law?

Since NFI forensic scientists were already involved in the case, the head of the FLDO in Leiden was called in as an expert witness. His assignment was interesting. Not only was he asked to review the Lab M report (testify on how they put down scientific evidence and the reliability of their analyses), he was explicitly asked if the results produced abroad were legally admissible. Answering these required the expert to assume two identities at once, legal and scientific.

On Standards

Preparation of the Dutch DNA law adopted in 1994 had taken almost five years. A committee chaired by legal expert Moons had evaluated the penal code and studied the available literature on the use of DNA evidence. The Moons report, delivered in January 1991, discusses several controversies in the United States that had pivoted around the validity of DNA evidence as a means of identification and the reliability of the technology. Validity was challenged especially in relation to matching-likelihood probability, and more specifically on the issue of population substructures. As noted, the probability based on Lab M’s German reference population was lower than on NFI’s white Dutch population. Such differences become significant when the reference population is not representative of the suspect’s profile. ¹⁸ Reliability of technology was a concern in the paradigmatic case of People vs. Castro, in which laboratory misconduct stood center stage (Lander 1992, Aronson 2007; Lynch et al. 2008; M’charek 2008b). In the wake of this case, the National Research Council (1992) issued its first report aimed at standardizing US laboratory procedures and methodological interpretation of data.

The Moons Committee recommended codifying technical procedures aimed at standardizing forensic DNA, ranging from bodily material used for DNA research, how and from whom this material should be taken, and which markers to use. An external auditing agency was to ensure that laboratories were accredited before the enactment of the law.

These controversies were also a source of scholarly engagement, specifically focusing on closure. Whereas some argued that the current mundaneness of DNA evidence is the result of technical improvements (e.g., Aronson 2007), others see closure not so much as a result of technical advancements but as a bureaucratic and administrative achievement (Lynch et al. 2008). Lynch and colleagues argue that in this process, the “sources of uncertainty” are black-boxed and that closure is better understood as “administrative objectivity,” an effect of standardized procedures aimed at concealing individual interpretation and subjectivity (2008, 230-33).

Here we want to propose another take on the contingencies of forensic DNA and standards. Rather than seeing “ambiguities” and “instabilities” as threats to standardized procedures, we draw on Vicky Singleton’s (1998) analysis of the British Cervical Screening Program (CSP), and suggest that ambiguity is part and parcel of applied DNA practice and contributes to its stability. In her analysis, Singleton shows that instabilities are built into laboratory practice and into the whole CSP. “[T]he CSP emerges as composed of a series of interacting, complex, decentered identities and as characterized by ongoing instability” (1998, 100). ¹⁹ Ambivalence and multiple identities do not jeopardize the network, Singleton argues, rather they contribute to its stability. By accommodating instability, the laboratory can arrive at and help maintain the stability of the overall CSP process. Moreover, the laboratory’s ability to deal with uncertainty (e.g., in results) and ambiguity (e.g., of samples) increases its scientific status. “It becomes an important complex component carrying out difficult and lengthy procedures and hence worthy of increased status and resources” (1998, 101).

Let us take this counterintuitive suggestion, that accommodating uncertainties and ambivalence contributes to stability, to investigate how standards are established in practice. We will examine the concerted efforts of the expert witness and the court to make the German laboratory similar to accredited Dutch laboratories and order their results under Dutch law.

The N-case: Doing Science and the Law Together

The FLDO expert witness was asked to review the results delivered by Lab M. Were they in accordance with the Dutch law? In his report, the expert argued that it may seem to be against the law to have a DNA analysis conducted abroad. But he had read other legal documents, especially Recommendation 92 of the Council of Europe and hit upon a discrepancy between the formulation of the Dutch law and this document. ²⁰ R(92), signed by all European ministers of justice, talks about the standardization of forensic DNA technologies across Europe and encourages collaboration across European Union (EU) borders, as long as the laboratories work according to the standards laid down.

The expert suggested to take R(92) into account in the N case. But he added, “It is beyond my capacity as a population geneticist and molecular biologist” to answer why the Dutch legislator had not taken R(92) into account earlier. He argued that it was up to the court to decide which of the two regulations should be given more weight.

To answer the question of Lab M’s reliability, the expert enlisted a number of publications on forensic DNA published by Lab M. He pointed out that its director was a renowned member of the International Society of Forensic Haemogenetics and that Lab M played a central role in introducing and standardizing technology in Europe. Its reputation was beyond doubt. However, since German law does not require the accreditation of forensic laboratories, it was difficult to conclude on the basis of the DNA report whether Lab M should be treated similarly to Dutch laboratories.

With reference to the soundness of the data, the analyses and conclusions in the German report, the expert witness was confronted by a problem. In his report (May 7, 1996), he quoted clarifications provided by the NFI about the matching-likelihood probability of 1:165,000.

This frequency is calculated under the presupposition that all fragments of the profile based on the sperm cells have been visualized.

The expert witness was puzzled by this statement because as far as he could see one allele was missing. For one specific marker (HUMFES), two alleles (fragment lengths) were found for the suspect: 10A and 13 and two for the victim, 10 and 11. However, for the biological trace containing the DNA of both victim and perpetrator, only three alleles were reported: 10, 11, and 13. What had happened to allele 10A? The expert witness sketched three scenarios for the disappearance of this allele:

10A was not in the biological trace, which would mean that the suspect could be excluded as the perpetrator.

Because of these ambiguities, the expert witness requested permission to travel to Münster and review the raw data. In Münster, he conducted a blind review of the material and data. He checked all films, Polaroid pictures of the gel runs, and the gels on which the markers had been visualized. To assess Lab M’s research routines, he reviewed the protocols and interviewed colleagues about procedures and measures taken to reduce the chance of contamination. The Dutch expert witness confirmed the presence of allele 10A. He found that allele 10A is faint, due to the visualizing technology (silver coloring), but it was there. On request, he was shown other research in which alleles 10 and 10A had been successfully separated.

In conversations about the case in Germany, he learned that Lab M members had also identified allele 10A in this very case but had three reasons for not reporting the find. First, they argued that the evidence from the other five markers was strong enough to include the suspect. Second, they wanted to prevent unnecessary discussion on the basis of one faint allele. Third, they were hesitant because allele 10A was not only faint, the difference between 10A (suspect) and 10 (victim) was only one nucleotide (one DNA building block). Given the facts that more victim’s than perpetrator’s DNA was available and that the visualizing technology leaves its own trace (silver coloring), Lab M could not indicate with certainty where allele 10A began and where it ended.

By contrast, the FLDO expert argued that he and other Dutch colleagues held another attitude: “Either you believe in your technology and procedures, in which case you report even faint alleles, or you don’t but then you should not report at all.” ²¹

Back in the Netherlands, the expert witness reported to the court that Lab M had employed all safety measures and that it should be considered an achievement that they had retrieved so much information from the difficult material. He concluded that the presence of allele 10A was confirmed by the data.

On July 17, 1996, he elaborated on his findings at the court hearing. Again the defense argued that the expert witness did not answer the question whether Lab M met the criteria in the Dutch Law. The following conversation took place:

Then the defense asked, “Can a lay person read allele 10A?” The expert witness answered, “I cannot answer that in general. The pictures and the original gels are available. If you are experienced in reading such material, you would know what to pay attention to, what you would need to read. A trained eye tends to see more than an untrained eye.” ²³

At the end of July, the court confirmed the validity of the DNA evidence, which had become even stronger now that allele 10A was taken into account. The suspect was charged on the basis of article 244 of the Penal Code which punishes acts which consist of or also consist of, sexual penetration of the body with a juvenile younger than twelve years. ²⁴ The suspect was reminded of his right to a counterexpert research. The court stated, “If he does not attend to his right to a counter investigation, or in case the second sample does not contain enough cell material for counter research, then the court takes the DNA research of Münster to be the same as research intended in article 195a article 2 Sv” [emphasis in original]. ²⁵

The defense decided to appeal against the decision on the case in response to which in December 1996 the Court of Appeal ruled that both the charge of sexual penetration (article 244) and the charge of murder could be proven. The suspect was sentenced to twelve years imprisonment and Involuntary Commitment (TBS). Next, the case was taken to the Supreme Court which ruled in 1997 that the suspect was granted the right to have a counterexpert research conducted on the DNA. The case was referred back to the lower court for this purpose, complying with the judgment made in December 1996. ²⁶ The DNA counterexpertise research that followed was conducted in yet another German laboratory (since both Dutch laboratories were already involved in the case). The second available swab was used for DNA extraction which yielded no DNA and therefore made no further analysis possible.

Managing Stability and Instability

The dissimilarity between Lab M and the Dutch laboratories was the initial reason to send the evidence to Germany. However, as we have seen, with the introduction of the DNA law of 1994 the equal treatment of the suspect as well as his identification depended paradoxically on making Lab M similar to the Dutch forensic laboratories.

The expert witness review of Lab M made clear how this laboratory accommodates instabilities to contribute to standardized DNA evidence practice. This is in accordance with Singleton’s (1998) analysis mentioned above. The Lab started out with a difficult biological sample that contained very little DNA material from the alleged perpetrator. The lab had to develop novel technologies to handle such a difficult sample. And then it had to deal with an ambiguous allele. Did this allele stem from the victim, from the suspect or was it an artifact? To tame this uncertainty and prevent it jeopardizing the DNA profile, Lab M decided not to report on the “missing” allele. Also, Lab M tested its technologies for more than two years before applying them to the biological evidence in this case. This contributed to the potential stability of the facts and to the overall scientific standing of the institute.

However, on its way into the Dutch legal practice, the DNA profile was nevertheless called into question. The Lab M’s report carried ambiguities and uncertainties into the courtroom. Hence, the puzzlement of the appointed expert witness and his eventual visit to Münster. To bring the DNA profile back into court and allow it to pass, Lab M’s methods and conduct had to be stabilized and presented as similar to those of the Dutch labs. Ironically enough, the expert witness’ ambivalent status (he was asked to answer legal and scientific questions) contributed to establishing the similarity of Lab M. Since equal treatment of the suspect depended on the fairness of the criminal investigation and extent to which Lab M’s conduct could be said to meet the standards, establishing Lab M’s similarity not only made the facts stand up in court but also enacted the equality of the suspect before the law.

While arguing that Lab M’s procedures did meet the standards, the expert witness also underlined instructive differences in what standards entail in practice and the implications for reporting. That Lab M’s forensic geneticists did not report allele 10A is one such interesting difference. According to the Dutch expert, the findings on 10A were relevant and should have been included. This can be viewed as a difference between a routine treatment and one in which the case is viewed in isolation. The Dutch expert not only relied on his evaluation of allele 10A in the N-case, he also took into account other experiments where Lab M had successfully visualized this allele. Its presence in the N-case was made dependent on an alignment between current and previous practices. Rather than an isolated fact, it had become part of the laboratory’s routines. By taking routines seriously, the ambiguity of allele 10A was tamed and its presence represented as a fact.

To bring this back to the issue of identity and sameness, one could say that identity is not just a matter of the technologies applied, but also of emerging and changing routines. Identity depends on the data found as well as the treatment of such data and its embeddedness in routines. All these aspects of scientific work can be made into key issues when establishing identity, both in the laboratory and in the court. As the Dutch expert witness indicated in his first report: If allele 10A could not be confirmed, it might indicate that the suspect should be excluded as the perpetrator.

The treatment of the missing allele has underlined yet another difference between German and Dutch forensic practice. The position of the German Lab M can be characterized as: scientific facts should speak for themselves. They should stand the test when they appear outside the laboratory and are subjected to the scrutinizing eyes of their reader. If there is possible doubt or ambiguity in the results for one of the markers it should not enter the report. Dutch practice does not draw the boundary between what counts as forensic evidence or not in this restricted way as it gives far more attention to the procedures by which evidence is assembled. For the Dutch court, the burden of proof does not lie in the scientific facts as such. Facts are considered part and parcel of routinized and standardized research practices. They are supported by a (protocolized and judicially normalized) report and are intimately linked to the scientific practice that helped produce them. More specifically, facts do not come from just any laboratory, but are produced in laboratories that are standardized and embedded in the legal system. In court, the expert witness’s accounting for the details of a standardized scientific practice is still crucial to the status of the facts. ²⁷ Not only does he have a trained eye for the facts and the methods applied to produce them but also for the legal criteria for forensic DNA evidence. His evaluation of the scientific measures taken by the German laboratory embody more than criteria for good scientific conduct. Routines, methods, and safety measures are emblematic of the hybrid practice of science and the law. ²⁸

Conclusion

In the Dutch context, the N-case is extraordinary in many ways. Not only was it subjudice for such a long time, it also spanned two different legislative periods during which DNA forensic technology developed distinctively. For these reasons, it is a paradigmatic case that helps us to see both how legal norms change and how practices oscillate (Knorr-Cetina 1981) around rules and principles as expert views and practical experience come into play. In this article, we have attended to practices of sameness in forensic DNA identification. Taking equality before the law as a practice rather than a formal principle, we have shown that establishing similarity in the form of demonstrative compliance with standards is crucial for the equal treatment of the suspect as well as for his identification.

Our focus on procedural standardization is particularly relevant since equality before the law in the continental criminal justice tends to focus more on the preconviction phase and the fact-finding process. Rather than viewing standards as more or less rigid modes of working, inspired by Singleton, we have shown that standardized procedures in forensic identification depend on the ability of laboratories, and other actors involved to accommodate instabilities and to assume an ambivalent identity. The complexity and heterogeneity of forensic identification as well as the entwinement of science and the law require such management of instabilities and fluidity of identity.

It is crucial to underline that the Dutch hybrid practice of science and the law does not always work the way it did in the N-case. Our case is not an illustration of arriving at an “administrative fix” or a mode of “black-boxing the sources of uncertainty” (Lynch et al. 2008) that is made durable beyond and across individual cases. Although the expert witness held a crucial role in this case, his knowledge and the laboratory practice can become matters of dispute on other occasions. Uncertainties and instabilities are never fully tamed or neatly managed. For example, M’charek (2000) shows in a forensic DNA case in 1996, involving the same expert witness, how the DNA profile failed to pass despite various attempts to rescue it. We therefore conclude that the interaction between science and the law in the N-case is a temporary local achievement in which a standardized practice was performed. The ambiguous identity of the expert witness, requested explicitly to assess legal rules as well, made it possible to treat his findings as more than scientific conclusions. The testimony became a legal argument for treating Lab M as similar to the Dutch laboratories and to treat its conduct in accord with the aim of the Dutch DNA law. This in turn helped to identify the suspect and in process establish his equal treatment before the law.