Introduction
In this blog post, we hope to bring public awareness and engagement to a current debate in the field of forensic anthropology— the use of ancestry estimation. The biological profile is a list of identifying, demographic parameters estimated from the skeleton by forensic anthropologists. Ancestry is one of its most contemplated aspects. Along with ancestry, forensic anthropologists estimate sex, age, and stature, which can provide critical information of the unknown individual. Many of our readers may be familiar with how the term ancestry is used in a different context to denote one’s ancestors or the reflection of such within one’s genetic makeup, as many people have discovered through programs like AncestryDNA or 23andMe. In forensic anthropology, ancestry is used as an identifying parameter for missing individuals due to the non-zero correlation between skeletal morphology, geographic origin, and social race (Dunn et al., 2020). While race is not biological, it is clear across various medical fields that systemic racism is biologically embodied through a combination of environmental, evolutionary, and social influences, as well as historical events, such as forced migration, assortative mating, and institutionalized racism in the U.S. (DiGangi & Hefner, 2013; Gravlee, 2009; Kilroy et al., 2020; Krieger, 2005; Ousley et al., 2018; Plemons et al., 2018; Risch et al., 2009; Sauer, 1992; Spradley, 2016). Although race is a sociological construct, it has strong, systemic impacts on individuals. However, in forensic anthropology, given that race or gender are not determined by biology, the anthropological context must also be considered when deriving meaning from analyses (Chrysostomou & Thompson, 2016; Jones, 2014; Sauer, 1992; Sauer et al., 2016; Schall et al., 2020).
Race, the comparative correlate to ancestry, is utilized in reporting outlets, such as missing person reports and systems like the National Missing and Unidentified Persons System (NamUs), which ultimately constrain reporting by forensic anthropologists. These systems work within our racialized society to achieve a primary goal: identifying the unknown. Each additional parameter that forensic anthropologists can estimate from the skeleton, including ancestry, helps us further narrow down missing persons cases for comparison, moving one step closer to a positive identification. Biological profile estimations are important not only in unidentified/missing persons cases but also for repatriation efforts of pre-historic Indigenous remains. However, the application of ancestry in forensic anthropological investigations has come under considerable scrutiny (Bethard & DiGangi, 2020). Researchers question whether the use of the term “ancestry” properly reflects methodology or if another term, such as “population affinity,” might better reflect this parameter. Two questions asked are: 1) Should we be estimating ancestry?; and 2) If so, are we doing it in a way which helps or hinders identification?
The Ancestry Debate
A recent letter to the editor in the Journal of Forensic Sciences (Bethard & DiGangi, 2020), brought up a lot of questions that our lab wants to reflect on including the following: Should ancestry be estimated by forensic anthropologists? What role do macromorphoscopic (MMS) variations play in this process? How does ancestry estimation affect bias in the identification of missing persons? By posing such questions, Bethard and DiGangi (2020) challenged forensic anthropologists to evaluate ancestry estimation in general and its overall benefit or harm to society as a whole— especially those already affected by systematic racism. Bethard and DiGangi (2020) make several major claims. First, they push for the field to abandon morphoscopic assessments of ancestry, specifically MMS variations which are a suite of morphoscopic traits used in forensics, based on the argument that this method is rooted in typology, lacks proof of heritability, and has not established the necessary evolutionary and ecogeographical background. Relatively new in comparison to metric methods that rely on measurements, MMS analysis uses visual morphological characteristics of the skeleton to estimate ancestry of an individual. This method has been standardized by researchers in the field within the last two decades (Hefner, 2009; Hefner & Ousley, 2014; Plemons & Hefner 2016). Although both methods of ancestry estimation exist (metric and MMS) using various parts of the human body (cranial, dental, postcranial), Bethard and DiGangi (2020) argue that cranial MMS variation is the only problematic method. Beyond methodology, the authors are concerned about the effects of reporting ancestry estimation and suggest it might contribute to bias in law enforcement investigations. While they state there is no evidence for such claims, they argue that such a possibility should not be overlooked.
Since the release of the letter to the editor, Stull and colleagues (2020) have authored a response to address the arguments posed by Bethard and DiGangi. In reference to the use of MMS traits, Stull and colleagues (2020) argue that anthropologists understand that skeletal variants are not equivalent to race and their use does not serve to reinforce typological notions, but to correlate ancestry and social race to make predictions about a person’s identity for identification purposes. Individual variations are not associated with populations or social races, rather, they are analyzed within statistical frameworks that rely on worldwide population samples with known frequencies to estimate (not determine) the ancestry of an unknown individual. Furthermore, Stull and colleagues (2020) point out the ample research into the evolutionary framework of variants used in ancestry estimation that support the use and accuracy of the method (Hefner & Linde, 2018). Next, Stull and colleagues (2020) address the concerns about law enforcement bias. While there have been publications that address some issues of law enforcement bias, such as Missing White Woman Syndrome or other identification biases, there is no evidence that the practice of estimating ancestry plays a role. And if there is no evidence for such a bias, then it is unfair to halt ancestry estimation in light of its benefits to missing persons. While it is important to verify the accuracy of ancestry estimation methods and to understand the role these methods play in society, removing ancestry estimation without proof of its potentially harmful role is not conducive to scientific processes.
Morphoscopic Ancestry Assessments
By bringing their opinions to the forefront of the field, these two articles allow forensic anthropologists to reflect and encourage us to examine the impact of estimating ancestry at individual case and societal levels. We reflect on these questions regarding ancestry estimation and personal perceptions of heritage, ancestry, and ethnicity while including examples from research and casework carried out at the MSUFAL. Our use of MMS methodology in our casework at the MSUFAL has caused us to reflect on the use and further development needed for this method and other ancestry protocols. Significant progress has been made in addressing the typological history of the MMS method and providing consideration of continuous human variation by 1) establishing a large, diverse reference dataset to capture variation within and between populations, 2) creating quasi-continuous variations to remove the notion of discrete traits, and 3) providing statistical frameworks to provide reliability and validity for the method (Hefner 2018). Recognizing that the typological trait list (an approach that assumes individuals from one ancestry group must exhibit a certain list of traits) used in forensic anthropological ancestry assessments was outdated and inaccurate, Hefner (2009) created a procedure for scoring cranial MMS variations. This method allows forensic anthropologists to collect standardized, non-typological data (Figure 1). Standardizing the data permitted the application of robust statistical methods to estimate the ancestry of an unknown individual with probabilities and likelihood statements (Dunn et al., 2020).
Figure 1. MMS Data Collection Tool
While there have been considerable efforts to accommodate for continuity in human physical characteristics across space, further research into the evolutionary and genetic underpinnings of MMS trait variation, in the same manner that craniometric, dental metric, and dental morphology have been investigated, is required. Studies have measured heritability in craniometric, morphoscopic cranial traits, and soft tissue (Adhikari et al., 2016; Relethford and Harpending, 1994; Roseman and Weaver, 2004), so it is unlikely that MMS traits contain no aspect of heritability. Large, publicly available genetic datasets allow for more research to be carried out on heritability of MMS variants, as well as the evolutionary and ecogeographical background. MSUFAL member, Amber Plemons, is conducting her dissertation research on these areas of concern. Her research aims to measure the influence of genetics and climate on MMS traits using a global population dataset. This project will reveal the evolutionary processes responsible for variation in MMS trait expression providing the necessary theoretical background to validate its use for ancestry estimation.
We know that we inherit physical characteristics from our parents which creates a genetic component to our physical appearances. As people move across the landscape, they may exchange genetic traits with other populations creating overlap in genetic structures between the groups. However, our bodies also physically adapt to our environment to ensure that we will survive and reproduce. For example, our nose is responsible for modifying the air we inhale to protect our respiratory system from damage from air that is too cold or too dry. A population may demonstrate particular physical traits that have been subject to natural selection over generations to improve the survival rate for that population. These genetic and environmental forces of physical trait expressions create geographic patterning in human variation. Anthropologists rely on these geographic patterns to estimate ancestry. The level of refinement that may be obtained for these estimations depend on the evolutionary background established for the method being used, the available reference samples for comparison, and knowledge of the population histories (e.g. migration patterns). The MMS Databank provides various levels of refinement for ancestry estimation, including 3-Group Race, Ancestry, Geo-Origin, and Population (Figure 2).
Figure 2. Ancestry Classification Refinement (Hefner, Kamnikar & Plemons 2018)
Researchers are actively working to understand which level of refinement is the most accurate and useful in reporting for medicolegal investigations. Researchers argue that 3-Group Race (Asian, African, and European) has been the traditional reporting method but should be abandoned as it groups populations together that have not been geographically adjacent for thousands of years, reducing the genetic and environmental correlations necessary for accurately estimating ancestry (Hefner & Spradley, 2018). Instead, these groups could be used as a starting point and should be refined as data allow. The assignment of Ancestry more aligns with typical categories for legal missing person reports in the United States, but these classifications are still very broad and do not account for distinct population histories, lumping together inappropriate populations which may reduce the ability to accurately estimate ancestry. Geographic Origin seems to be the most appropriate level of refinement for forensic investigations to improve identification efforts. However, these assignments do not overlap appropriately with the check boxes on legal documents. For example, in NamUs a missing person is assigned to one of six Race/Ethnicity categories or marked as “Other”, “Uncertain”, or “Not provided” (Figure 3). There is no space for entering additional useful information that may align more appropriately with the individual’s population history. Herein lies the problem of ancestry estimation in forensic anthropology.
Figure 3. NamUs Missing Persons Inputs (NamUs.gov)
One example of a successful and useful ancestry estimation in our lab is the 2019 case of Charles Oppenneer. In this case, law enforcement brought a cranium they believed to be a White female to the MSUFAL for a biological profile and trauma assessment. Plastic deformation and rodent gnawing obscured many features of the cranium, contributing to an overall gracile appearance of the cranium and making assessments of sex and ancestry challenging. MMS variations typically used for ancestry estimation, including the inferior nasal aperture and anterior nasal spine, as well as morphological traits used for sex estimation, were damaged and not useful for estimating any aspects of the biological profile. Additionally, measurements were limited due to plastic deformation and traumatic damage to the cranial vault. Despite law enforcement’s initial observations that the remains were a White female, craniometric analyses of sex and ancestry using FORDISC consistently classified the remains as most similar to an Asian male. Upon consulting NamUs, there was only one missing Asian male whose demographic and geographic details matched those of this case. Dental radiographs of the missing individual from NamUs were compared with postmortem radiographs of the unknown decedent, and an identification was made. If ancestry was excluded from this NamUs search, upwards of 90 individuals matched the circumstances of the case. By estimating ancestry using validated methods, an identification was made quickly and accurately. Had ancestry not been included, the identification process would have taken significantly longer and cost substantially more time and money than necessary. This case illustrates the utility of estimating ancestry and the ways in which an accurate, scientifically-based ancestry estimation can benefit members of the medicolegal community.
Moving Forward
As a discipline there is a need to educate law enforcement and reporting outlets on the theoretical processes of evolutionary and genetic biology that allows forensic anthropologists to estimate a biological affinity of their ancestors, without perpetuating the outdated typological ideas of certain traits correlating with certain populations. DiGangi & Bethard (2020) argue for the halt of MMS research, but isn’t the goal of science to support or disprove hypotheses through the scientific method? We need to continue to collect data to reinforce or refute the hypotheses surrounding the biological underpinnings of MMS research, not stop the research altogether. The use of ancestry, on the whole, is extremely important regarding the other pillars of the biological profile (e.g. stature, age, and sex) which rely on these population-specific data to more accurately determine the identity of an unknown individual (Adams et al., 2019; Cattaneo et al., 2018; García-Parra et al., 2014; Kenyhercz et al., 2017; Liebenberg et al., 2019; Tallman, 2019). The question is, can we stop estimating ancestry when trying to identify unknown individuals? Research, time and time again, has shown that population-specific methods are ideal for estimating the other three parameters— how do we use population-specific methods if we cannot estimate ancestry as well? Do we rely on historical methods that were created solely using American Black and American White assemblages and bias our results towards those populations? Refinement of ancestry has been done using craniometrics and has continued with MMS traits. This is especially true for groups clustered together under the term Hispanic. Researchers have demonstrated variability within Latin American populations, which has been applied to the current migration crisis at the U.S.-Mexico border (Hughes et al., 2013; Monsalve and Hefner, 2016; Spradley, 2014; Spradley, 2016; Tise et al., 2014). Kelly Kamnikar’s research aims to improve upon identifying variability in samples from the Yucatan Peninsula of Mexico and Guatemala. This important data will contribute to the accurate assessment of region of origin of unidentified migrant remains that are recovered along the southern U.S. border, which can help expedite the identification and repatriation process. To hear about future MSUFAL research regarding this topic and further discussion on these questions, listen to The Sci-Files podcast interviewing MSUFAL PhD students Kelly Kamnikar, Amber Plemons, and Micayla Spiros on their dissertation research.
Conclusion
Ramifications of this debate (i.e. the continuation or halt of ancestry estimation) impacts forensic anthropology, but those impacts extend to the public at large that forensic anthropologists serve. Anthropologists understand that certain skeletal traits do not equal race or even gender, but there is a correlation between human variation and the skeleton due to culture, environment, and evolutionary processes, which allow us to make these predictions.
Forensic anthropologists cannot be the only stakeholder participating in this discussion. We are not the only body in the medicolegal community – nor are we the only field to utilize the term ancestry. If we plan to truly move forward with this discussion, we must reach out to our colleagues in other forensic disciplines to understand and consider their perspectives as well. This discussion is much larger than forensic anthropology, and even the medicolegal community, as it really impacts those in the community we are working to serve. No methodology is perfect, and we have yet to fully understand the range of human variation. It is fair to say that our system is not perfect either. However, the primary goal for forensic anthropologists is identification and it seems appropriate to use any tools (properly backed by the scientific process) that we can in order to assist in that process.
Anthropologists are aware of difficulties in terminology, which is why there is a further discussion in the field surrounding ancestry terminology (Maier, Craig and Adams 2020). Bridging the biological aspects of the skeleton with social constructs such as race and gender is a hurdle that we need to continue to understand due to the complex relationship between different standards of reporting and self-identification. It can be complicated to understand the overlap between group division, different tiers of ancestry estimation, self-identification, and ethnicity – and how such estimates of biological parameters by forensic anthropologists should be most appropriately characterized to promote the identification of decedents. Changing a term, such as ancestry, within the discipline might be beneficial but engagement with broader audiences, like medicolegal practitioners, law enforcement, and the public at large, is lacking. If we want to make meaningful change, we need to engage with diverse audiences about these issues and work with them to formulate solutions to improve our practice.
Authored by: Amber Plemons, Micayla Spiros, Rhian Dunn, Kelly Kamnikar & Alex Goots