NAGPRA, Repatriation and the Ethics of Skeletal Reference Collections

The Hamann-Todd Collection at the Cleveland Museum of Natural History numbers approximately 3,100 individuals, accumulated between 1893 and 1939. The primary intake source was the Cuyahoga County workhouse and the hospital wards of the city's public institutions, where Ohio's body-provision law at the time permitted medical schools to claim unclaimed bodies before burial. The demographic composition of the collection reflects the demographic composition of Cleveland's unclaimed poor: approximately 38 per cent African American, a proportion substantially higher than Cleveland's Black population at the time, because social and economic marginalisation correlated with both institutionalisation and unclaimed-body status. This composition has important implications for any biological-profile method validated on the collection: the Hamann-Todd is not a representative sample of the US population.

The Robert James Terry Collection at the Smithsonian Institution's National Museum of Natural History numbers approximately 1,728 individuals, accumulated between 1898 and 1967 by Terry at Washington University in Saint Louis and by his successor Mildred Trotter, who later developed the foundational Trotter-Gleser stature regression equations from the collection. The Terry Collection's intake came primarily from the St Louis City Hospital and its associated institutions. Again, the demographic over-representation of poor, Black, and institutionalised individuals is well-documented in the collection's own accession records.

The Bass Donated Skeletal Collection at the University of Tennessee Knoxville (UTK) sits on a different ethical foundation. William Bass established the Anthropology Research Facility (ARF) in 1981 and began accepting donated cadavers from individuals who had consented during their lifetime to donate their remains for forensic research. The Bass Collection now numbers over 1,750 individuals, all with documented ante-mortem consent. It is demographically broader than Hamann-Todd or Terry (though still predominantly European-American and from the US Southeast) and is the primary reference for the Body Farm's postmortem interval research. The consent-based model of the Bass Collection has become the template for all subsequent outdoor forensic research facilities.

In South Africa, the Pretoria Skeletal Collection at the University of Pretoria and the Raymond A. Dart Collection at the University of the Witwatersrand together represent approximately 2,000 or more South African skeletal individuals, primarily accumulated through anatomy department intake during the apartheid era. The demographic distribution, the intake mechanisms, and the community consent status of these collections have been the subject of ongoing ethical review by South African bioanthropologists and by the Health Professions Council of South Africa.

In India, the AIIMS New Delhi anatomy department and the DFSS (Directorate of Forensic Science Services) Gandhinagar both maintain skeletal reference material used for method validation and casework comparison. The AIIMS collection includes forensic anatomy cases and donated remains; documentation of individual consent histories is variable. The DFSS collection includes skeletal material from medicolegal cases and from institutional donation. These collections are referenced in Indian forensic anthropology literature on sex estimation and stature equations (Mukherjee 1955, Pan 1924, Singh-Bhasin equations) but have not been the subject of the same legislative repatriation scrutiny as the US collections, in part because the donor demographic and the colonial-extraction pattern are different.

The Native American Graves Protection and Repatriation Act (NAGPRA), signed into law on 16 November 1990 (Public Law 101-601), is a US federal statute that creates obligations for federal agencies and institutions receiving federal funding. Its core provisions are: first, the protection of Native American human remains, funerary objects, sacred objects, and objects of cultural patrimony from disturbance on federal and tribal lands; second, the requirement that museums and federal agencies holding such items conduct an inventory of their collections, consult with culturally affiliated tribes and Native Hawaiian organisations, and offer repatriation of the items on request.

The Act defines "cultural items" to include human remains, associated funerary objects (items buried with the individual), unassociated funerary objects, sacred objects, and objects of cultural patrimony. The determination of "cultural affiliation" is the operative question: an institution must determine whether a group of human remains can be linked, by a preponderance of evidence, to a present-day tribe or Native Hawaiian organisation through geography, kinship, biology, archaeology, anthropology, linguistics, folklore, oral tradition, historical evidence, or other relevant information.

The 2010 amendments to NAGPRA regulations (43 CFR Part 10) addressed a significant gap in the original 1990 statute. The 1990 Act required repatriation of culturally affiliated remains but left unclear what happened to remains whose cultural affiliation could not be determined. Thousands of skeletal individuals in institutional collections, acquired from contexts where tribal affiliation was unclear or documentation was absent, fell into a "culturally unidentifiable" category and remained in collections for twenty years after the Act's passage. The 2010 amendments established that culturally unidentifiable human remains should be offered for repatriation to the tribe or organisation with the most likely affiliation, based on geography or other available evidence, rather than being indefinitely retained.

Compliance with NAGPRA has been uneven. The Kennewick Man case (known as the Ancient One by the coalition of Columbia Plateau tribes) was in litigation from 1996 to 2016: a 9,000-year-old skeleton found in Washington State was claimed for repatriation by a coalition of Columbia Basin tribes, while a group of physical anthropologists sued to study the remains on the grounds that the ancient remains were not demonstrably affiliated with any contemporary tribe. The Ninth Circuit Court of Appeals ruled in 2004 that the remains were not "Native American" under NAGPRA because their cultural affiliation to contemporary tribes could not be established. Congress passed an amendment in 2016 directing the Army Corps of Engineers to facilitate repatriation; the remains were repatriated and reburied on 18 February 2017.

The United Nations Declaration on the Rights of Indigenous Peoples (UNDRIP), adopted by the UN General Assembly on 13 September 2007 (A/RES/61/295), is the most comprehensive international statement of indigenous peoples' rights. It was adopted by a vote of 143 in favour, 4 opposed (Australia, Canada, New Zealand, and the United States, all of which subsequently endorsed UNDRIP by 2016), and 11 abstentions.

UNDRIP Article 11 affirms that indigenous peoples have the right to practise and revitalise their cultural traditions and customs, and that states shall provide redress "through effective mechanisms, which may include restitution" in respect of their cultural, intellectual, religious, and spiritual property. Article 12 affirms the right of indigenous peoples to maintain, protect, and develop the past, present, and future manifestations of their cultures, including human remains. Article 31 affirms their right to maintain, control, protect, and develop their cultural heritage, traditional knowledge, and traditional cultural expressions. Together these articles provide the international normative framework within which national repatriation policies in Australia, New Zealand, Canada, and the UK have been developed.

Australia has moved furthest in legislative implementation. The Aboriginal and Torres Strait Islander Heritage Protection Act 1984 (Commonwealth) provides protection mechanisms for Aboriginal areas and objects. The National Museum of Australia, the Australian Institute of Aboriginal and Torres Strait Islander Studies (AIATSIS), and most state museums have adopted repatriation policies that go beyond what any single statute requires. The AIATSIS Return of Cultural Property Programme has facilitated the return of over 1,000 sets of Aboriginal and Torres Strait Islander ancestral remains from Australian and overseas institutions since the 1990s.

In New Zealand, Te Papa Tongarewa (the Museum of New Zealand) operates the Karanga Aotearoa Repatriation Programme, which since 2003 has facilitated the return of Maori tupuna (ancestral remains) and kaitiaki taonga (cultural treasures) from institutions in the UK, France, Germany, Sweden, and other countries. The Programme is notable for the quality of its bilateral negotiation model: it does not simply demand return but engages the holding institution in a dialogue that typically results in voluntary repatriation.

In the United Kingdom, there is no statutory equivalent of NAGPRA. The Human Tissue Act 2004 governs the holding and use of human remains in England, Wales, and Northern Ireland. Section 47 of the 2004 Act, as amended by the Human Tissue (Scotland) Act 2006, allows the trustees of nine national museums (including the Natural History Museum, the British Museum, and the Pitt Rivers Museum Oxford) to deaccession human remains up to 1,000 years old if they consider it appropriate to do so. The default position for remains over 1,000 years old, which includes most archaeological material, remains retention with case-by-case review.

In 2018, De Montfort University in Leicester returned the skeletal remains of an individual known as Asaru-Ojo to the Ogoni community in Nigeria. The remains had been held in the university's anatomy collection and were believed to be those of a man brought to Britain during the colonial period, likely in the late nineteenth or early twentieth century. The university determined that the remains could not be assigned a specific identity but could be linked by provenance to the Ogoni region of southern Nigeria. After a consultation process with the Ogoni community and the Nigerian government, the university facilitated the return and reburial.

The case attracted attention because it was handled without a legislative mandate. The Human Tissue Act 2004 permitted but did not require deaccession of remains of this age. De Montfort acted on the basis of an institutional ethics determination that the collection of the remains, under colonial circumstances, without community consent, created an obligation to return them when a plausible community affiliation could be established. The case has been cited in subsequent UK institutional reviews of colonial-era anatomical collections as a practical model for how returns can be effected under the existing legal framework without waiting for statutory reform.

The Natural History Museum London, which holds one of the largest collections of human remains in the UK (approximately 20,000 individuals, predominantly from the nineteenth and early twentieth centuries), has a formal Human Remains Policy that commits to the respectful curation of all remains, consultation with source communities, and case-by-case consideration of repatriation claims. The Museum has returned remains to Australia, the United States, and other countries; claims from African communities are at a less advanced stage of institutional engagement.

The Pitt Rivers Museum Oxford completed a project in 2019 and 2021 to review the provenance and community connections of its human remains collection (approximately 2,000 individuals). The review found that the majority of the remains were acquired without documented consent and that many were from colonial contexts. The Museum announced in 2020 that it would accept repatriation claims from any community with a demonstrable connection to its holdings. The Pitt Rivers case is significant for its public acknowledgment that the collecting practices of the museum's nineteenth-century donors, including General Augustus Pitt Rivers himself, were ethically unjustifiable by contemporary standards.

The dependence of forensic anthropology method validation on the Hamann-Todd, Terry, and allied historical collections is not incidental. The foundational age-estimation methods (Suchey-Brooks pubic symphysis phases, Iscan-Loth sternal rib end phases, Lovejoy auricular surface phases) were all validated on Hamann-Todd and Terry specimens. The admissibility implications of using these collections in casework are directly addressed in the ABFA, SWGANTH, ENFSI FAWG quality frame, which covers Daubert, BSA 2023, and ISO 17025 requirements for method validation transparency. The first generation of discriminant-function sex-estimation equations were built from these same collections. The Howells craniometric reference database, which underlies FORDISC, was assembled from skeletal collections including Hamann-Todd, Terry, and various archaeological assemblages, some of which also have contested provenance.

This creates a specific problem for practitioners who must defend their methods in court. Under the Daubert reliability standard in the United States or the analogous standards in the UK, Australia, and India, the validation sample must be representative of the population from which the case specimen comes. If the validation sample is overwhelmingly European-American or African-American from early-twentieth-century Cleveland, methods validated on it may not have established error rates for South Asian, South American, East Asian, or sub-Saharan African populations. This is not a hypothetical concern: FORDISC has been the subject of published studies (Ousley and Jantz, 2012; Hefner, 2015) documenting the limitations of its reference sample coverage for non-US populations.

The practical response has been the development of alternative and supplementary reference collections assembled on an informed-consent basis. The Bass Donated Skeletal Collection at UTK is the largest US consent-based collection. The Lisbon Identified Skeletal Collection (Museu Bocage, Portugal), the Coimbra Identified Skeletal Collection (University of Coimbra, Portugal), and the Scheuer Collection (now at University College London) provide European reference samples with documented individual identities. The Pretoria and Witwatersrand collections in South Africa are the primary reference for sub-Saharan African population methods. The AIIMS and DFSS collections in India provide the primary reference for South Asian stature, sex-estimation, and age-estimation equations.

Researchers building methods on historical collections that have contested provenance, particularly methods that will be cited in NAGPRA-related proceedings or in legal cases involving Indigenous remains, now routinely address the provenance question directly in their publications, acknowledging the circumstances of collection, noting the consent status of the sample, and identifying the limitations that follow from those circumstances. The Journal of Forensic Sciences and Forensic Science International have both published editorial guidance encouraging this contextualisation.

The willed-body donation model, in which individuals consent during their lifetime to the use of their cadaver for medical or forensic research, is the current ethical standard for reference collection building. In practice, practitioners and bioethicists have identified several complicating factors.

First, the demographics of willed-body donors are not random. Individuals who donate their bodies tend to be older, more highly educated, more European-American or, in the Australian context, more European-descent, and more likely to have prior positive interactions with academic medicine than non-donors. The Bass Collection, for example, is demographically skewed toward European-American males from the US Southeast. Methods validated on a consent-based collection with this demographic profile may be no more valid for non-European or non-American populations than methods validated on the Hamann-Todd. The consent is genuine; the representativeness is limited.

Second, the specificity of consent matters. Some donors consent broadly to "forensic research"; others consent to specific uses and exclude others. A donor who consented to outdoor taphonomy research at the ARF in 1990 did not consent to having their skeletal elements 3D-scanned and placed in a digital morphometric database accessible to researchers worldwide. Institutions managing consent-based collections are navigating, with varying degrees of care, the question of what the original consent authorises as the technical capabilities available to researchers change over time.

Third, donated skeletal collections increasingly include individuals from communities whose relationship to institutional medicine is historically fraught. Native American individuals and African-American individuals have begun donating to some US collections. The institution's management of these donors' remains requires attention to whether tribal or community affiliations are documented and whether NAGPRA or equivalent frameworks apply to donated remains (NAGPRA does not currently cover voluntarily donated remains from enrolled tribal members, but this has been the subject of policy discussion).

In 2020, two overlapping institutional statements shifted the public position of American anthropology on the use of historical skeletal collections in research that invokes biological race.

The American Anthropological Association's 2020 statement "Race and Racism in Anthropological Research" reaffirmed the AAA's longstanding position that biological race is not a valid scientific concept for human populations, extended it to the specific context of forensic anthropology's use of race-based skeletal collection data, and called on the discipline to adopt population-affinity language and methods that do not reproduce racial categorisations. The statement explicitly cited the demographic circumstances of the Hamann-Todd and Terry collections as an example of how racial categorisation at intake has propagated through the methods built on these collections.

The Society for American Archaeology (SAA) issued a separate statement in 2020 on NAGPRA implementation, specifically addressing the culturally unidentifiable remains category and calling on institutions to accelerate compliance with the 2010 amendments. The SAA statement noted that many institutions had invoked research value as a justification for retaining culturally unidentifiable remains rather than offering them for repatriation, and that this justification was, in the SAA's view, insufficient to override the communities' interests.

The practical implications for casework are nuanced. The AAA statement does not prohibit forensic anthropologists from using Hamann-Todd or Terry specimens in their research. It calls for explicit acknowledgment of the collections' provenance and demographic limitations, for population-affinity framing that does not invoke racial typology, and for active investment in building consent-based, demographically representative collections. The ABFA and OSAC have both updated their guidance to reflect population-affinity language in ancestry estimation, replacing terminology such as "race" or "racial group" with "population group" or "ancestral group" in casework reports.

The repatriation question and the population-affinity question intersect in one specific practical context: forensic cases involving the remains of Indigenous individuals. An osteological analysis of remains that may be Indigenous faces the NAGPRA question (are these remains subject to repatriation before they are analysed?) simultaneously with the population-affinity question (which reference sample and which methods are valid for this individual?). OSAC guidance and the ABFA position advise that when Indigenous remains are encountered in a forensic context, consultation with the relevant tribal authority should begin before laboratory analysis proceeds, and that the analysis should use methods validated on populations with relevant geographic and biological affinity.

Contemporary forensic anthropology casework guidance from ABFA, OSAC, and ENFSI FAWG converges on a set of ethical practices that extend beyond the legal minimum of NAGPRA compliance.

Before conducting laboratory analysis on skeletal remains from a context that may involve Indigenous, historically marginalised, or community-affiliated individuals, the practitioner should determine whether the remains are subject to NAGPRA, UNDRIP-implementing national legislation, or equivalent frameworks in the relevant jurisdiction. If they are, consultation with the relevant tribal authority, community, or national government should precede analysis. This is not simply a legal precaution; it produces better forensic anthropology, because community knowledge about an individual or context often provides information that skeletal analysis alone cannot.

The selection of methods should account for the population affinity of the case specimen. Using methods validated exclusively on European-American reference samples to analyse remains that are likely to be East Asian, South Asian, sub-Saharan African, or Indigenous North American creates error rates that are unknown and likely higher than the published error rates for the validation sample. Practitioners should select methods validated on a reference sample as close as possible to the population represented by the case specimen, document the choice, and address the limitation in the uncertainty quantification section of the report.

Casework reports should include provenance information for the methods used: the reference collection on which each method was validated, the consent status of that collection if known, and the known demographic limitations of the sample. This information is not routinely requested by investigating agencies, but it is relevant to the Daubert reliability assessment and to any subsequent peer review of the report.

In jurisdictions without NAGPRA or equivalent legislation, including India and most African states, forensic anthropologists who encounter skeletal remains from context that suggests Indigenous, tribal, or community-affiliated origin (burial sites, archaeological contexts, sites of known historical conflict) should apply the ethical framework as a professional obligation even without a legal mandate. The same population-affinity considerations apply when choosing reference samples for DVI commingled-remains analysis in mass-disaster contexts. The ABFA Code of Ethics and the ENFSI Code of Conduct both require practitioners to respect the cultural and spiritual significance of human remains and to engage with communities of interest where such communities can be identified.

1. Determine legal status of the remains
   Assess whether the remains are subject to NAGPRA (US federal law), UNDRIP-implementing legislation (Australia, New Zealand, Canada), the Human Tissue Act 2004 and associated museum policies (UK), or BSA 2023 (India). Seek institutional legal advice if the status is unclear.
2. Identify community affiliation
   If the remains are from an archaeological, historical colonial, or mass-grave context, use available documentary and contextual evidence to identify any tribal, ethnic, or national community with a plausible affiliation. Begin consultation with that community before proceeding to laboratory analysis.
3. Select population-appropriate methods
   Choose biological-profile estimation methods validated on a reference sample with appropriate population affinity. Document the reference collection, its consent status, and its known demographic limitations.
4. Conduct laboratory analysis
   Perform analysis under a documented protocol (ISO 17025 or equivalent). Ensure all findings are recorded contemporaneously. Apply the qualified-opinion framework for any identification conclusion.
5. Peer review with ethics check
   The peer reviewer assesses not only the scientific conclusions but also whether the population-affinity method selection is appropriate, whether the provenance of the reference collection is adequately documented, and whether the report language is consistent with population-affinity rather than racial-typology framing.
6. Report with provenance transparency
   Include in the report: the methods used, the reference collection for each method, the consent status and demographic limitations of that collection, the error rates on the reference sample, and the appropriate uncertainty range for each metric estimate.
7. Repatriation facilitation (where required)
   Where remains are determined to be subject to NAGPRA or equivalent, facilitate the repatriation process in cooperation with the institution and the relevant tribal or community authority. Document the process and outcome in the case file.