Abstract
New and emerging forensic genetics technologies offer significant insight into personal information, changing the way that policing and criminal justice uses of such technologies are being considered and legitimized. In this article, based on data from Central and Western European countries and the United States of America, we analyze how the compounding, interdependent effects of four such technologies—massive parallel sequencing, forensic epigenetics, forensic DNA phenotyping, and forensic genetic genealogy—facilitate the dissolving of boundaries between forensic and medical, as well as between commercial and non-commercial domains. Mobilizing social epistemology and epistemic culture as dual analytical lens, we argue that we can witness the emergence of an increasingly complex forensic genetics assemblage, fostering dependencies between policing agencies, research scientists, and commercial companies. At the heart of this assemblage lies the transformation of central knowledge claims and distinct roles and responsibilities defining the legitimate application of genetics data and information in policing contexts. The dissolving of boundaries and deepening of co-dependencies within the assemblage encourage increased self-governance between the key stakeholders, to the detriment of the field's societal accountability and legitimacy. The discussion in this article provides a necessary starting point for reframing the discussion of forensic genetics’ governance.
Keywords
Introduction
Many criminal justice systems have legitimized the use of forensic genetics by limiting DNA use to comparing two or more profiles to identify a person. Key to this understanding has been that forensic genetic markers must not reveal personal information beyond confirming a (mis)match, biological sex, and parentage (Cole 2007). This assumes that marker-corresponding alleles on the genome do not contribute to gene expression, i.e., they are “non-coding” for how genes contribute to traits such as hair, eye, or skin pigmentation. For policing and the judiciary this has meant that forensic DNA is deployed for comparing or matching profiles (as trial-relevant evidence) rather than for attributing characteristics (intelligence informing investigations). Policing legislation and codes of criminal proceedings in European jurisdictions are often still based on an interpretation of a matching focus for using DNA in the criminal justice system (Samuel and Prainsack 2018; Williams and Wienroth 2017). This interpretation is increasingly challenged by clinical and nascent forensic genetic research efforts (e.g., Benecke 2002; Kayser and Schneider 2009). Accompanied by lobbying from various sectors, including the forensic genetics community, corporate actors, policing representatives, and policy-makers, changes have been made to legislation to enable both matching and attributing in policing within Switzerland, the Netherlands, Germany, and France. Apart from the Netherlands, new legislation tends not to introduce oversight, training, or necessary mechanisms to accompany this significant stepchange in using genetic material for security and justice purposes. This introduces further complications into the ethics and practices of making persons known through forensic genetics.
Dissolving Boundaries
Inspired by this change in the uses of DNA in the criminal justice system in Europe, here we analyze how the compounding, interdependent effects of four new and emerging technologies drive the dissolving of practical, governance, and ethical boundaries between forensic and medical genetics, as well as between commercial and non-commercial domains, while fostering dependencies between stakeholders. We shed light on the increasing application of health-related information in the forensic field and for criminal justice purposes, while observing the growing exploitation of the so-called “genomic potential” in commercial service provision in justice and security contexts. This culminates in the practical difficulty of demarcating forensic and medical domains in the development of technologies and, due to their increasing informativity, also in their deployment outside the laboratory. Yet, medical and forensic domains are subject to differing ethics, aims, practices, and expectations. Similarly, state-run DNA databases are embedded within an entirely different framework than commercial ones; within the profit to non-profit continuum, we see very different ethical models across recreational (including ancestry), health, research, and forensic databases. These challenges occur simultaneously, and at times they also intersect with significant implications for the positioning of forensic genetics in criminal justice and society more widely, and for its governance. Specifically, the dissolving of boundaries we address here has far-reaching consequences for data ownership, data uses, cost of forensic services, research capacities, and the independence of forensic analyses. It is, therefore, important to understand the old, new, and still emerging configurations of forensic genetics, to enable the field's informed ethical governance. We explore the dissolving boundaries and strengthening dependencies between stakeholders in four technologies:
- - - -
Forensic Genetics as an Assemblage
Since the introduction of “DNA fingerprinting” in the 1980s (Aronson 2010; Jeffreys, Brookfield and Semeonoff 1985) forensic genetics has developed in waves of technosocial innovations (Wienroth, Morling and Williams 2014): Introduction and contestation of forensic genetics in the 1980s and 1990s (first wave); its consolidation and integration into criminal justice via enabling legislation, the introduction of digitized DNA profiles and databases, and cross-border and cross-database interoperability since the mid-1990s (second wave); the increasing use of attribution via familial searching, forensic DNA phenotyping, and biogeographic ancestry testing to search for unknown suspects since the early 2000s (third wave); and, vitally for this article, the emergence of a fourth wave that relies on increased sequencing capacities and the dissolving of boundaries between medical and forensic data and uses since the 2010s. While the waves’ temporality is fluid, cresting differently in jurisdictions across the globe, they help understand how forensic genetics has been changing, expanding, and is being negotiated.
Social studies of forensic genetics have analyzed the initial legitimation of forensic DNA profiling (Aronson 2010; Derksen 2000; Lynch et al. 2008) and the ethical, social, and legal aspects of its sociotechnical innovations since (Hopman 2020; M’charek 2008; Skinner 2013), presenting insightful approaches to understanding how forensic genetics is developed, mobilized, and governed. However, this scholarship still lacks a conceptualization of the entanglement of, and its effects on, academic research and criminal justice with other logics in the context of forensic genetics, including commercial dynamics and policing priorities.
In this article, we take inspiration from Deleuze and Guattari's (1987) concept of assemblage as a relational and productive articulation of multiple, heterogeneous parts linked together (Müller 2015, 28). The forensic genetics assemblage consists of research-related institutions, law-enforcement agencies, oversight bodies, specialized companies, and affected publics through which discourses, materiality, and practices (e.g., contestations of reliability, utility, and legitimacy; biosamples, biomarkers, equipment; and forensic services) are produced, distributed, applied, and negotiated. As there are no pre-determined hierarchies or a single organizing principle underpinning assemblages (Müller 2015), the concept is particularly useful to draw links between the pragmatic and political concerns related to crime control—identifying victims and suspects, producing evidence for prosecution—with global processes of knowledge production, capital accumulation, and technoscientific governance. The concept of the assemblage helps us understand how new territories “emerge and hold together but also constantly mutate, transform and break up” (Müller 2015, 29), such as sudden moments of expansion and intensification in forensic genetics that make it harder to govern the field from the outside, while seemingly enabling some parts of this assemblage to develop a mode of self-governance as counterpoint to external or independent oversight. We therefore explore how increasing the geneticization and scientification of human beings creates scenarios of ungovernability within the new forensic genetics’ assemblage of research scientists, policing agencies, and commercial actors.
Analytical Approach: Social Epistemology and Epistemic Culture
Starting with a literature review of Global North texts from genetics and social sciences about the four technologies and criminal justice law and codes of criminal conduct in various European jurisdictions (including Germany, Portugal, Switzerland, UK) and the United States of America, and informed by previous interviews and informal contacts with forensic geneticists and observations at conferences of the forensic genetics community, we have studied the types of knowledges (operational assumptions and expectations, genetic [re]production of personal characteristics, socio-legal and ethical understandings) drawn from and generated for their development and deployment, and their critiques. Thus, we have identified sources and ways of knowing, and changes in the ways that new and emerging forensic genetics developments are rationalized and negotiated—made known—in moving from matching individuals (e.g., to be used as evidence in court) to attributing group-based personal characteristics including familial relationships, appearance, ancestry, health, and so-called lifestyle as forms of intelligence informing investigations. We have studied the technologies’ material networks (databases, corporate websites), media discussions, and governance documents (policy, legislation), to inform our understanding of the dependencies between forensic genetics researchers in academia and case work, policing agencies, and corporate stakeholders.
Based on these sources, we mobilize a dual analytical lens of (1) social epistemology and (2) epistemic culture. Social epistemology is “the conceptual and normative study of the relevance of social relations, roles, interests and institutions to knowledge” (Schmitt 1994, 1), it helps us explore understandings of genetics within forensic technologies, the data they can provide and how these data are interpreted—social epistemology helps us account for the emergence of forensic genetics as a technology of policing. Epistemic culture, as “those amalgams of arrangements and mechanisms—bonded through affinity, necessity and historical coincidence—which, in a given field, make up how we know what we know” (Knorr Cetina 1999, 1, original italics), structures our analysis of user communities, their assumptions, purposes, and governance (regulation, codes of conduct, professional ethics).
The first component of our analytical lens starts from the forensic genetics epistemological claim that emergent research and development render a clear distinction between “coding” (contributing to the expression of traits such as hair or eye color) and “non-coding” (previously also referred to as “junk” DNA) parts of the genome unfeasible. The key aspect here is that legislation built on this distinction has provided legitimacy for comparing DNA traces on the basis that these genetic markers cannot offer information on inheritable illnesses or specific traits. Contesting the capacity to strictly distinguish between parts of the genome that can be used to only compare two or more traces, and those that can be used to infer personal information about the donor of a trace, has undermined the building block of much regulation in the field of forensic genetics. As forensic geneticists Kayser and Schneider (2009, 158) explain: One of the main distinctions made in this context is the use of “coding” vs. “non-coding” DNA markers for forensic purposes. This apparently clear definition is supposed to provide a limit for the extent to which the human genome may be analyzed in the forensic context. However, the (human) genome is not organized in such static way but rather consists of blocks of DNA that are inherited mostly intact….Thus, a non-coding DNA marker that exists in close physical proximity to a marker which codes for a phenotype (including an EVC phenotype) can reveal the same information as the coding marker itself.
The second component of our analytical lens attends to the elements that engender the forensic genetics community. Up until recently, the role of commercial actors was mainly to provide materials, equipment, and specialized analysis, but we are witnessing a shift whereby commercial actors have become more relevant in curating databases with forensic interest, discussed below in the section titled Commercial and non-commercial dissolving boundaries. The growing role for commercial databases foregrounds the differences between commercial and non-commercial frameworks—for data and research ethics, approaches to data collection, curation, and sharing, service provision and communication—and new actors that sociotechnical innovations bring to the field.
To shed light on the dynamics of dissolving boundaries, we mobilize four new and emerging technologies as case studies. Rather than reiterating existing analyses of individual issues posed by each of these technologies in isolation, we attend to their compounding effects. Figure 1 provides a summary overview of our analytical approach.
Summary of Analytical Approach.
Dissolving Boundaries Between Forensic and Medical Genetics
Research in medical and forensic genetics has co-emerged. Some researchers work in both, for example in the development of geno- and phenotyping of facial features for diagnostic purposes (e.g., Baynam et al. 2015; Dominguez-Alonso, Carracedo and Rodriguez-Fontenla 2023; Roosenboom et al. 2016) and for forensic DNA phenotyping. This can be partly explained by research funding being more readily available to medical than to forensic genetics. In response, the forensic genetics community has anchored the need for further research support and a continued raison d'être for the academic research field to three main arguments. Firstly, further research is vital to address evidentiary issues arising from problems such as mixed DNA traces, and from a lack of understanding about what DNA at a crime scene may say about the activities leading to its deposition (e.g., analyzing which body fluid the trace originates from). Secondly, increasing sensitivity in finding and analyzing tiniest DNA traces aggravates existing and creates new issues (e.g., DNA transfer, contamination) that can lead to miscarriages of justice. Thirdly, matching DNA profiling may be insufficient in difficult cases, and new tools are needed. Especially the first and third arguments drive the four technologies discussed in this article, thus facilitating the step-change to potentially extensive personal information being made available to criminal justice stakeholders and corporate forensic service providers. At the same time, much of these analyses generate group-based intelligence rather than individualizing information.
This development evidences the deployment of biopower (Foucault 1976), firstly because physical and behavioral features of the human body are given social meaning based on a medical framing of the body that works as a disciplining force; and secondly through the power of population-level regulatory forces, specifically criminal justice and wider forms of policing—such as surveillance—to which these technologies contribute (see Rabinow and Rose 2006). This takes place within the context of efforts to control crime and to evidence the effectiveness and efficiency of forensic genetics to legitimize it.
Forensic biological research is often seen as part of the discipline of legal medicine—see overlapping publications in national and international journals for legal medicine and for forensic science, but also in descriptions of forensic research groups within academic institutions of legal medicine, e.g., in Austria and Germany—denoting the epistemological and epistemic relatedness of medicine and biological forensics. However, medical and forensic applications take place in ethically, socially, and professionally divergent domains: The rationales for genetic analyses, the use of information, their operational and oversight contexts matter. In fact, medical ethics has been an essential part of the professionalization of medical research—consider the Nuremberg Code, the Helsinki Declaration, the Singapore Statement on Research Integrity; and the emergence of bioethics in biomedicine (e.g., Beauchamp and Childress 2001; Childress 2020). In the forensic genetics domain, a focus on research ethics as part of the field's professionalization has only recently started to emerge (Wienroth et al. 2021), but still falls significantly short of a professional ethos (Wienroth and McCartney 2023). Despite the epistemological and epistemic differences of these fields, some of the developments in the medical realm are gradually diffusing into forensics, thus dissolving boundaries. In this regard, one of the key step-changes in forensics is the move from using only short tandem repeats (STRs)—used to match/compare different profiles—to also analyzing polymorphisms. That is, insertions and deletions of nucleotides in DNA sequences (indels) and single nucleotide polymorphisms (SNPs) that provide information on genetic variations and are deployed to associate gene changes with disease susceptibility and occurrence (Bradbury, Köttgen and Staubach 2019; Lin et al. 2017). This renders forensic uses of genetic material potentially much further-reaching in terms of its informational power, not only trying to match a crime scene trace with a suspect, but developing profiles of unknown suspects, and generating insight into health and behavioral aspects of persons of interest.
A similar development can be observed in research on the microbiome, expanding from developing health-relevant uses to identifying investigative applications. This includes using microbial DNA and RNA for estimating timescales of body decomposition; body fluid identification to make statements about the action level of a stain (its relevance to the crime and how the crime may have occurred); touch DNA analysis (minuscule traces can offer more microbial than human DNA), and for the purpose of identifying unknown persons by analyzing health-relevant, behavioral, and location information drawn from the microbiome (National Institute of Justice 2021). In this section, we focus on two examples of forensic genetic technologies to observe instances of boundary dissolving between forensic and medical genetics: Massive Parallel Sequencing and Forensic Epigenetics.
Massive Parallel Sequencing
Massive Parallel Sequencing (MPS) is a technological platform capable of running diverse tests on the same genetic sample simultaneously. MPS first emerged in genetic medicine (Rogers and Venter 2005), with the promise of enabling faster and cheaper genetic sequencing. From early on, MPS effectiveness has been linked to the need for increasing analytical capacities and greater professional responsibility for the production of new and more forms of personal data (Tucker, Marra and Friedman 2009). In the forensic context, MPS has emerged since the 2010s (Børsting and Morling 2015), where it is seen a means to address common sequencing issues with samples that contain DNA from different people, generating greater sensitivity by simultaneously trying to match an unknown with a known DNA profile (e.g., developing a crime scene profile that can be compared against profiles on policing databases). MPS is also used in forensic contexts to analyze DNA traces to attribute traits such as appearance and ancestry to a sample of an unknown/unmatched person. 1 Other types of genetic markers might also be tested at the same time, potentially increasing the level of data and of informativity available from a single biosample. Higher sensitivity of analysis is relevant, for example in rape cases, where samples tend to contain large quantities of the victim's DNA and smaller, harder to detect, amounts of the perpetrator's DNA. MPS is expected to contribute to greater individualization in forensic genetics and higher value to investigations: “the distinctive character of the profiles increases and therefore also the value of the evidence” (Bruijns, Tiggelaar and Gardeniers 2018, 2647). However, the information made available by MPS is highly dependent upon specialized training. Thus, scientists’ capacity to turn genetic data into information useful to an investigation varies widely.
MPS is portrayed as an enabler for improving the investigative relevance of other new and emerging forensic genetics technologies (Børsting and Morling 2015). Furthermore, higher sensitivity (meaning smaller sample sizes) and simultaneous analysis (meaning materials need less time in labs) are likely useful in addressing biomaterial storage as a significant limitation to data-rich genetic analysis—due to costly refrigeration and space considerations. Indeed, MPS is portrayed as offering a higher level of efficiency for using even very small, degraded, and/or contaminated samples (as is usual in criminal investigations). However, not all MPS systems seem to operate equally well and show significant limitations, including when testing traces of less than two nanometers (two millionth of a meter), a usual forensic sample size (Truelsen et al. 2021). Therefore, forensically useful MPS depends on competently applying current techniques of copying DNA that can produce large enough quantities to run a reliable, effective analysis (De Knijff 2019; Gorden, Sturk-Andreaggi and Marshall 2021). Interestingly, MPS has also been considered useful for validating emerging technologies: In the case of environmental DNA (eDNA)—of skin cells, fungi, bacteria, pollen etc.—high-throughput capacity has been described as vital in quickly generating reference data (Young and Linacre 2021).
MPS capacities have been developed with a view to commercialization, also in forensics (e.g., see Illumina 2017), with a focus on technical improvement capacity arising from competitiveness between commercial developers (Børsting and Morling 2015) and an emphasis on “cost-effectiveness” (Gorden, Sturk-Andreaggi and Marshall 2021). As with the examples of forensic DNA phenotyping and forensic genetic genealogy discussed below, commercialization emerges as an issue of dissolving boundaries when forensic services are provided without sufficient validation and oversight of proprietary forensic products. This has also been raised as a key issue for why MPS has not yet found much application in the forensic domain, despite the perception that it could be particularly helpful in rape cases (De Knijff 2019; Hopman, Oorschot and M’charek 2023). A survey of European forensic institutes reflected users’ concerns about the lack of validation and clear standards (Alonso et al. 2017).
Of particular interest for MPS is the dissolving boundary between forensic and medical uses and informativity (epigenetics provides an example of MPS application) which is indicated by the platform capacity of MPS to generate a more comprehensive and complex analysis of genetic material. This depends on the choice of biomarkers applied to the genetic analysis. Some markers will have both medical and forensic uses. Most commonly this is RNA (ribonucleic acid), which plays a vital role in gene expression. SNPs also represent localized mutations that can be indicative of health-relevant data, so-called off-target phenotypes (Bradbury, Köttgen and Staubach 2019).
Forensic Epigenetics
Epigenetics studies the molecular mechanisms regulating the function of genes, for example through DNA methylation (Eccleston et al. 2007, 395). The prefix epi- means “on top of” or “above” genetics (Moffitt and Beckley 2015). This field has received intense research interest and media attention (Pickersgill et al. 2013) due to its (presumed) ground-breaking implications leading to reframing the debate of nature versus nurture (Machado and Granja 2020, 37–40). Headlines such as “The Famine Ended 70 Years Ago, but Dutch Genes Still Bear Scars” rendered the field a hot topic in the science section of news media such as The New York Times. 2 Studying the relevance of epigenetic changes in the development of cancer, obesity, and other medically relevant conditions is a booming field within medical research (Eccleston et al. 2007).
Four main principles are shared among experts in the field: epigenetic mechanisms can be (i) the cause for a change in gene function but do so without changing the gene sequence; (ii) established in initial development, particularly during pregnancy, and their effects may manifest throughout life; and they are (iii) inheritable but also (iv) reversible. However, most insights so far are based on small-scale, experimental animal studies exploring the impact of maternal behavior on methylation patterns during pregnancy and infants’ early development (Weaver et al. 2004). Consequently, “the implications of research in epigenetics for science, health and society are unclear” (Pickersgill et al. 2013, 434). Nonetheless, as evidenced for genome editing, knowledge production continues to be perceived as a good in itself and made distinguishable from the future application of that knowledge. In other words, “this distinction effects a discursive separation of the ethical debates about research…from any potential medical applications of such knowledge” (Wienroth and Scully 2021, 7).
Epigenetic studies have driven the notion that biological mothers’ life conditions and behaviors during pregnancy and children's early development significantly affect epigenetic markers, with far-reaching consequences. This framing holds the power of activating and augmenting a range of moral discourses, thus enabling (increased) scrutiny of female bodies (Pickersgill et al. 2013; Richardson 2021).
The innovative potential of epigenetics has not passed unnoticed in criminology (Moffitt and Beckley 2015), with some authors starting to explore how the modulation of gene expression might impact crime. Some studies address the epigenetic influences of crime hotspots (Leshem and Weisburd 2019), while others analyze how epigenetics may be used to explain the developmental origins of chronic physical aggression (Tremblay and Szyf 2010).
Despite the enormous investigative value ascribed to forensic post-genomics research, forensic applications of epigenetics have mainly been restricted to body fluid identification, sex determination, differentiation between monozygotic twins, and biological age prediction (Haddrill 2021; Sabeeha and Hasnain 2019; Vidaki, Daniel and Syndercombe-Court 2013). Since the late 2010s, considerations of the potential investigative uses of epigenetics and epigenomics have grown with expectations that improved understanding of how environmental factors impact human genetics would increase its value for investigative intelligence. In particular, forensic geneticists argue that: While the genome is typically non-informative regarding lifelong environmental influences on the body, which can provide forensically relevant information, the epigenome acts as an interphase between the mostly “fixed” genome and the principally “dynamic” environment. (Vidaki and Kayser 2017, 1)
Commercial and non-Commercial Dissolving Boundaries
There are two key developments contributing to the dissolving of boundaries between commercial and non-commercial domains in forensic genetics. These include the expansion of forensic service provision by private companies to policing agencies, and the use of commercially held DNA databases to search for criminal suspects.
Different kinds of public–private partnerships have emerged in the forensic genetics field. In most continental European jurisdictions, routine forensic analyses for criminal investigations, (disaster) victim identification, and migration purposes tend to be conducted by public laboratories closely working with and/or managed by police forces, and by often academic, research-oriented laboratories. More specialized or emergent types of analyses tend be forwarded to private providers. For example, in Austria, five academic forensic institutes (Granz, Innsbruck, Linz, Salzburg, Vienna), provide the police with case work analyses, furthering the capacity of police laboratories like the one run by the Federal Crime Agency. Interviews with forensic scientists, observations at conferences, and collaboratively published scientific literature show that researchers at academic institutes work—at times directly (for example in developing and testing technologies), at times indirectly (exchanging research insight for access to otherwise expensive materials and equipment)—with corporate entities in developing further forensic knowledge and equipment. The Netherlands Forensic Institute (NFI), part of the Ministry of Justice and Security, covers most national case work needs in the Netherlands, but also offers for-fee international services such as training and expertise, to supplement its state-funded income. The NFI used to develop highly specialized new forensic analyses (e.g., body fluid identification), but budget cuts since the late 2010s seem to have led to a reduction in developmental research capacity and an increase in collaboration with industry and universities. In Germany, academic researchers and commercial institutes provide analyses in cases where police laboratories cannot due to cost or capability limitations. Commercial service providers can, if they receive block contracts for specific services, offer these at reduced prices, thus making them more competitive. This has led to criticism from academic forensic scientists that the move of public case work funding to corporate providers threatens basic research capacities.
In such contexts, the main role attributed to the forensic market and corporate actors (large R&D companies such as Illumina, QIAGEN, and Thermo Fisher), tends to be the provision of materials and equipment for forensic genetics research, as well as the development of forensic equipment and technologies. Corporate involvement is indirect, while smaller commercial companies tend to offer specialized forensic services.
Another model of public–private partnership is more common in the US, UK, and Australia, which have a tradition of outsourcing services since the 1990s. In these countries forensic services are mainly tendered by private providers while regional police-owned laboratories—at times run by commercial forensic providers, such as the laboratory of the Metropolitan Police in London, UK—provide the most common analyses, thus splintering the forensic services market and reducing the capacity of individual providers to develop and validate novel technologies. In the UK, three large companies (Eurofins, Cellmark, Key Forensic Services) share the market of block-commissions. In order to stay profitable, they outsource many jobs to tiny one or two-person commercial forensic service providers. A review by the UK's House of Lords Science and Technology Select Committee (2019) identified lack of funding as a threat to the commercial viability of key service providers, alongside inconsistency in outsourcing practices among the 43 police authorities in England and Wales; and inconsistency in the regulation—including validation—of commercial providers such as specialized genetics services. In the US, forensic service providers offering specialized forensic analyses to forensic practitioners and police investigators have grown significantly. A particularly illustrative example is the number of cases outsourced to Parabon NanoLabs, particularly their Snapshot service that combines FGG, FDP and kinship inference.
Commercial providers have recently also become more relevant within the forensic landscape for owning and curating DNA databases. Since the mid-1990s there has been a move to establish and grow forensic DNA databases that can be used to search for profiles. Such databases involve the “collection, storage and use of DNA profiles from nominated suspects, convicted offenders, victims, volunteers and other persons of interest to criminal investigation work” (Machado and Silva 2015, 820). Databases are usually established and regulated by the state, with varying legislative frameworks across countries (Santos, Machado and Silva 2013). They are, therefore, associated with particular regulatory frameworks and, in some cases, subjected to specific oversight mechanisms—see for example the UK's National DNA Database (Amelung and Machado 2019, Skinner and Wienroth 2019). Until recently, with a few concessions related to the exceptional authorization to access non-forensic DNA databases, state-run DNA databases were the main instrument used by law enforcement. Yet recently there has been a change: recreational DNA databases used by citizens to voluntarily upload DNA data originated from direct-to-consumer testing, are also starting to be used to search for criminal suspects (Granja 2021). In this section we focus on two examples of forensic genetic technologies outlining the increasing blurring of boundaries between commercial and non-commercial DNA databases: Forensic DNA Phenotyping and Forensic Genetic Genealogy.
Forensic DNA Phenotyping
The ambition to infer a person's appearance from genetic material in forensic genetics research goes back to the 1990s, starting with hair pigmentation and facial morphology (Motluk 1998; Valverde et al. 1995). Forensic scientists have framed further development of this technology—referred to as forensic DNA phenotyping for its emphasis on visible characteristics such as eye, skin, and hair pigmentation, biological age, head hair shape, and facial morphology—in terms of knowledge claims: as adding value to criminal investigations where profiles generated from traces do not match existing DNA profiles on databases; when eyewitnesses are not available or cannot provide further clues (Kayser 2015; Kayser and Schneider 2009); and/or their reports may be unreliable (Schneider, Prainsack and Kayser 2019). These epistemological claims suggest that knowing an unknown person by using FDP is more trustworthy than an eye witness account. This claim is based on mathematical calculability of probabilities of either confidence in data, or the comparison of contrasting hypotheses (also known as a likelihood ratio), suggesting an epistemic preference for quantified over experiential data.
Commercial service providers also draw on this argument of making visible what more traditional DNA profiling and databasing cannot shed light on, emphasizing the role of FDP in urgent and cold case investigations alike. However, some commercial actors are more promissory about their services’ benefits, describing them, as generating a useful and reliable DNA “snapshot” of criminal suspects (see Parabon NanoLab's Snapshot service). Such ideas echo aims for FDP research previously articulated by leading academic forensic geneticists (e.g., Frudakis 2010 on “photofitting”). Although such promises are no longer widely echoed by the research community, the strong emphasis on marketing by commercial actors means that such services continue to be promoted in ways that exaggerate their potential. Notably, Parabon NanoLabs, Identitas Inc., and Verogen have been commercial leaders in the direct marketing of FDP services to criminal investigators. Much of the scientific basis for their products has been developed by academic scientists, often supported by corporate developers of lab equipment, and then adopted by commercial service providers (Granja and Machado 2023; Wienroth 2020), who then advertise products as useful in cases that offer “no suspects or database hits, to narrow suspect lists, and to help solve human remains cases” 3 in order to “keep investigations alive [by] recover[ing] more information from a single sample.” 4 Parabon's suggestion that their FDP product Snapshot can “‘reverse-engineer’ DNA into a physical profile” mobilizes epistemic claims such as precision, scientific reliability, and operational confidence (e.g., by showing confidence intervals for characteristics). 5 This is of particular interest because proprietary products do not need to publicize their validation process, for example in academic journals. Although used for longer than FGG, a recent article in Slate magazine referred to FDP as “Genetic Genealogy's Less Reliable Cousin” (Mak 2019), which is likely to say more about consumers’ confidence in FGG than about their distrust for FDP.
FDP is associated with collectivizing and racializing suspicion (e.g., Bartram, Plümecke and Schultz 2021; M’charek, Victor and Lisette 2020) by generating suspect groups of people who share certain characteristics. For providers and users of FDP, this operational aspect of the technology renders it useful for police agencies issuing public calls for information on crime suspects. So while the use of FDP and the related technology of biogeographic ancestry testing is growing in commercial forensic service provision to criminal justice actors in the USA (and elsewhere, but extensively there, see Parabon NanoLabs), these proprietary products tend to lack peer-reviewed technology validation, shoring up contestation of reliability between academic and commercial forensic scientists (Granja and Machado 2023; Wienroth 2020).
Forensic Genetic Genealogy
Forensic genetic genealogy refers to the investigative searching for criminal suspects and/or persons of interest on recreational DNA databases by constructing family trees. In 2010, the non-profit database GEDmatch was set up by two amateur genealogists. The database enables customers to upload (their personal) DNA profiles generated by commercial ancestry and health companies (e.g., 23andMe, Ancestry.com, MyHeritage) to identify genetic connections with other uploaded profiles, with the goal of developing a family tree. Although there has been forensic interest in other genetic databases before, 6 forensic genealogy marks the emergence of using non-forensic DNA databases for systematic and routine criminal justice searches. Since its role in the detection of the Golden State Killer in 2018, GEDmatch specifically, and forensic genealogy in general have become popular with law enforcement agencies (Granja 2021, 2023; Phillips 2018; Ram, Guerrini and McGuire 2018).
Rather than relying on state-owned forensic DNA databases alone, which contain short tandem repeat profiles and are subjected to specific regulations (Santos, Machado and Silva 2013), there is growing interest in using DNA databases built and expanded through direct-to-consumer genetic testing to search for criminal suspects. Such databases hold profiles inclusive of information derived from SNPs, so they have a much wider informativity potential. Recreational and corporate databases are not subject to state regulations that prescribe including and excluding criteria, retention periods, and access controls. Those privately or commercially owned databases voluntarily offering data access tend to do so based on the (alleged) consent of citizens who uploaded their data (for a problematization of the issue of consent see Samuel and Kennett 2020). Law enforcement access to such databases tends to be decided upon by their custodians, usually private companies. This means that while many databases restrict access and require law enforcement to present a warrant to enable searches, others such as GEDMatch and FamilyTreeDNA publicly state user policies that allow law enforcement agencies to conduct extensive searches. Since these commercial databases are based on opt-in/opt-out consent, using them enables investigators to bypass the stricter data-sharing, data protection, and data retention rules that apply to government police databases.
Recognizing the potential of publicly accessible ancestry databases for policing purposes, in late 2019 the commercial forensic services provider Verogen (2019) acquired GEDmatch, evidencing a “New Phase in Law Enforcement's Use of Consumer Genetic Data” (Bala 2019). Shortly after, a new service for police agencies was created, named GEDmatch Pro: GEDmatch PRO™ is a dedicated portal designed to support police and forensic teams with investigative comparisons to GEDmatch data. The portal separates police comparisons of GEDmatch data from standard genealogy activities and offers a range of tools most relevant to help further investigations.
7
FamilyTreeDNA also makes its database available for criminal justice searches. According to the company website, law enforcement or any third-party representative working with them is required to register prior to uploading to the FamilyTreeDNA database. Permission to use the service may be granted to identify the remains of a deceased individual and/or to identify a perpetrator of homicide, sexual assault, or abduction. 11 The creation of such specialized portals consolidates a new avenue for serious crime investigations, increasingly used in the US (Dowdeswell 2022) but also explored in Australia, Norway, Sweden, and the UK (Scudder et al. 2020; Thomson et al. 2020; Tillmar et al. 2021).
FGG therefore completely reconfigures previous agreements and understandings about what kind of genetic data can be used for law enforcement searches and how, as well as the type of information it can provide. In addition, FGG technology also reconfigures the epistemic cultures of forensic genetics as there are two new sets of central actors within the forensic genetics assemblage: companies curating the databases being used to conduct forensic genealogy, and genealogists. The latter are a highly diverse group of individuals, from various educational and professional backgrounds, who due to their interest and investment in genetic genealogy have become a ready source of expertise for forensic genetics. Their role shows how the emergence and development of FGG has relied on citizen science—citizens actively and voluntarily contributing to scientific development (Granja 2023).
Forensic Genetics Assemblage and its Dependencies
The forensic genetics assemblage, characterized by dissolving boundaries and deepening dependencies between forensic and medical fields, as well as between commercial and non-commercial domains, materializes along relational axes between the three core groups of (1) forensic research scientists; (2) policing agencies; and (3) commercial companies (see Figure 2). The latter combines corporate suppliers of materials for forensic genetics research and case work to academic researchers and other forensic users, with commercial forensic service providers offering specialized forensic analyses to forensic scientists and police investigators. In this article we focus solely on the social epistemology and epistemic culture of the forensic genetics’ assemblage. Although we cannot explore in detail all governance aspects, we acknowledge that further study of legislation and regulation, including best practices, accreditation, and oversight bodies, as well as self-governance of the forensic genetics community, are vital in the analysis of forensic genetics (Wienroth 2018).
Forensic genetics assemblage and dependencies.
Regarding social epistemology, the previously held distinction of coding and non-coding regions on the DNA is set aside. Indeed, development and use of new forensic genetics technologies explicitly build on abolishing the “non-coding” notion to legitimize further investment of academic, policing, and commercial resources. The technologies addressed in this article, which are at different stages of implementation in criminal justice settings, are discussed as holding informational potential well beyond matching identification purposes, toward attributing appearance, ancestry, disease predisposition, “lifestyle” factors, and biological bonds between individuals. Claims about the informative potential of (epi)genetic/genomic data, and the ever-expanding possibility to explore and capitalize upon it emerge as an epistemological model for legitimizing the diffusion of boundaries in the development and deployment of new and emerging forensic genetics technologies. However, in many jurisdictions, this step-change in the perceived informativity potential of genetic material, as well as its investigative interpretation, is proceeding without reframing the governance of genetic material in the criminal justice system. This is particularly apparent for FGG and FEpi, where there is no sustained discussion about whether this type of information is desirable or even useful for policing purposes. Debate about the genetic inference of appearance and ancestry, and the logistical and wider operational adaptations required for using MPS infrastructure and information in police work, tends to be limited to the widely discredited balancing of human rights/freedoms with security, as observed in discussions of FDP and BGA in Germany (Amelung and Machado 2021; Lipphardt 2018; Weitz and Buchanan 2017).
The forensic genetics assemblage also sheds light on a range of new and renewed epistemic cultures that take different forms. First, it outlines the increasingly integrated collaboration of academic scientists and policing laboratory practitioners with commercial organizations in the development and validation of new forensic technologies. For example, companies such as Thermo Fisher and Illumina work with and give materials and equipment to scientists to evidence the scientific quality (that is, the feasibility and reliability) of products. 12 Second, current configurations bring into focus a set of actors that are either taking new extensive roles or are completely new to the field. Companies, sometimes consisting of only one or two specialized service providers, sometimes in the form of international biotechnology corporations, are significantly expanding their field of action in forensic genetics, increasingly providing specialized services to police work. Simultaneously, specialized commercial products tend to be offered to police forces ad hoc rather than as part of a tendering process for block-contracts. Additional providers have entered the field of new and emergent forensic genetics analyses, too. For example, policing agencies in the UK, Germany, and Australia include police-run forensic laboratories who at times are involved in validating the technologies we have discussed, enabled by personal relationships between employees of policing agencies (e.g., lab scientists) and academics, or commercial research companies.
Specifically for forensic genetic genealogy, we observe citizen science impulses (Granja 2023) that shed light on the role of forensic genetic genealogists in advancing complex cases. Although this emergent role is still under discussion—with steps toward certification only beginning (Gurney et al. 2022)—it already embedded in the forensic genetics assemblage. Genealogists who started off, at least initially, as amateurs (Granja 2023) are hired and incorporated into commercial companies and police agencies, and genetic genealogy is beginning to be recognized as an autonomous scientific discipline (Durie 2017).
This proliferation of roles and actors has helped create liminal spaces of responsibility, accountability, and transparency where their allocation and extent become diffused. In the same ways that actors shift from the public to the commercial domain (e.g., investigators leaving to work as consultants to specialized companies), data and materials are now also travelling across different domains (research-police-commercial) without the particular safeguards that used to protect this kind of sensitive information in state-run DNA databases.
Concluding Remarks
We have analyzed four genetics technologies at different stages of development and implementation to show how their compounding, interdependent effects foster the dissolving of boundaries between forensic and medical fields, and between commercial and non-commercial domains. By analyzing aspects of social epistemology and epistemic culture, we present evidence of an emergent and increasingly complex forensic genetics assemblage, based on the generation of new and deepening of existing links and co-dependent relations between policing agencies, research scientists, and commercial companies. At the heart of the forensic genetics assemblage lies the transformation of central knowledge claims and distinct roles and responsibilities defining the legitimate application of genetics data and information in policing contexts (e.g., the distinction between “coding” and “non-coding” regions on the genome as a knowledge boundary). This is accompanied by a legitimizing strategy based on claims that new and emergent forensic genetics technologies have a unique informational nature and growing applicability.
Our analysis of dissolving boundaries offers an example of the deployment of biopower (Foucault 1976) at the levels of human body and population—by identifying, technically enabling, and arguing for integrating new features of the human body into a system of policing and criminal justice in order to enhance the governmentality of crime. At the same time, FDP, BGA, FGG and FEpi generate a challenge to oversight of the field itself by contesting key features of external governmentality (e.g., the scientific basis of legislation and regulation), by deepening co-dependencies (e.g., producing ethical and jurisdictional questions of data and database ownership between research, commerce, and policing), and by shifting the goalpost for accountability (e.g., by exploiting less stringent or non-existent governance for private citizen science and corporate service provision, thus avoiding more proscribed governance for criminal justice and policing, or medical genetics). The dissolving of boundaries and deepening of co-dependencies within the assemblage, however, seem to encourage increased self-governance between the key stakeholders, to the detriment of the field's societal accountability and legitimacy. And yet, these developments are taking place against a backdrop of data infrastructures developed in an ethically dubious manner, informed by colonial legacies of forensic and policing approaches to identification of suspect populations and their practices, and colonial logics of data collection and retention, as well as interpretation (e.g., Lipphardt et al. 2021; Moreau 2019; Van Oorschot and M’charek 2021). Indeed the forensic genetics assemblage is part of the sometimes fraught history of forensic genetics (Derksen 2000; Lawless 2012; Lipphardt et al. 2018). These are global and political issues and as such require an international and interdisciplinary approach to address them.
Arguably, these developments and considerations are a necessary starting point for reframing the discussion of forensic genetics’ governance. More particularly, by moving from frameworks prescribing good ethical practice in the criminal justice system toward approaches that recognize “ethics as lived practice” (Wienroth et al. 2021) and incorporate new actors, materials, and infrastructures. This approach overlaps with the need to professionalize forensic science, which cannot be done at the local level (Robertson 2011; Willis 2023), an evaluation that applies even more so to forensic genetics. Taken together, the increasingly global nature of genetics knowledge production and databanks, non-forensic reference and recreational databases (which are nonetheless used for forensic purposes), and cultural tropes of knowing and identity call for an international effort to professionalize forensic genetics. This is a political act that requires inter- and transdisciplinary efforts, which places ethics at the heart of professional practice (Wienroth and McCartney 2023), allowing problematic issues to be contested and scientific ideas and roles to be related to cultures of knowledge, values of justice, fairness, equity, and societal goods such as criminal justice and security. Although this political act may not be sufficient in itself to address all the particularities and challenges arising from the governance of forensic genetics, it is a necessary first step that fosters the emergence of a generative space for more ethically informed decision-making and practices, learning from the past, reflecting on the present, and anticipating for the future.
Footnotes
Acknowledgements
We would like to thank three anonymous reviewers and the editor for their valuable comments that have supported us in developing this work. We are also grateful to our network of colleagues whose conversations and advice over years have contributed to our understanding of the subject.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work has received funding from the Fundação para a Ciência e Tecnologia (FCT - Portuguese Foundation for Science and Technology) under the Programme Scientific Employment Stimulus—Institutional Call, attributed to Rafaela Granja (https://doi.org/10.54499/CEECINST/00157/2018/CP1643/CT0003). In addition, this work is also supported by FCT under the project UIDB/00736/2020 (base funding) and UIDP/00736/2020 (programmatic funding).
Fundação para a Ciência e a Tecnologia, (grant number CEECINST/00157/2018, UIDB/00736/2020).
