You are on page 1of 5

Sejrsen, Lynnerup, Hejmadi 40

B. Sejrsen, N. Lynnerup, M. Hejmadi
Forensic Odontology-Anthropology Unit, Institute of Forensic Medicine, Copenhagen, Denmark

description, leading to identification of an unknown
ABSTRACT body. The forensic team has various methods to
The Institute of Forensic Medicine, Copenhagen, obtain information about the deceased, such as
houses a collection of historical skulls of unclear evaluation of sex, age and ethnic origin.1, 2
origin, marked with a general geographic or “racial
descriptor”. Would these historical skulls be of any
value for the forensic odontologist and anthropolo-
The various methods employed for determination of
gist concerned with teaching and casework? We tried ethnic origin include both metric and non-metric
to clarify this question by recording non-metric den- assessments. For example, computer programmes
tal traits and by performing craniometric analyses. have been developed which match certain cranio-
facial measurements with a database containing
A morphological and morphometric investigation of craniofacial measurements from osteological mate-
anatomical/dental traits in 80 adult skulls was per- rial of known ethnicity.3 However, the assessment of
formed. For each skull four non-metric dental traits
using the ASU-System and three non-metric cranial
non-metric traits may be more subjective, as they
traits were recorded. Nineteen cranial measures were cannot be discretely quantified, but rather reflect the
also taken following the FORDISC programme training and experience of the investigators in identi-
manual. The non-metrical data were tabulated as fre- fying and recording these traits. Furthermore, when
quencies, and the metric data were entered in the evaluating ethnicity it may be those self-same non-
FORDISC programme. Observed non-metric trait metric traits, least commonly seen by the investiga-
frequencies were compared with published data. The tor, which prove the most crucial. In Denmark, for
FORDISC programme computed a discriminatory
analysis for each skull and thereby assigned the skull
example, shovel shaped incisors are rarely seen in
to the most probable ethnic category. the general Danish population, but forensic identifi-
cation cases may involve the remains of purported
The results for the non-metric traits showed that the Greenlanders (Inuit/“Mongoloid”), who often have
traits generally followed the expected frequencies in these traits. Thus a Danish forensic odontologist, with
80% of the cases. The FORDISC programme Danish odontological training, may simply not have
correctly assigned ethnicity based on skull measure- much experience in discerning these traits. This may
ments in overall 70% of the cases.
be remedied by studying skeletal collections, which
It was found that this historical collection does show include material of varied ethnicity. While such
expected dental non-metric and craniometric traits collections do exist, especially in North America,4
and the collection may be of value in forensic case- most European collections tend to be of an historical
work in terms of comparison and for teaching and archaeological nature.
(J Forensic Odontostomatol 2005;23:40-4) The purpose of this investigation was to apply some
of the methods used for evaluation of ethnicity to a
large collection of historical and archaeological skulls,
Keywords: non-metric dental traits, FORDISC programme, in order to evaluate the usefulness of an older an-
ethnic group evaluation, historical skulls thropological skull collection concerning casework,
in terms of training and teaching. Non-metric dental
traits were recorded and craniometric analyses per-
INTRODUCTION formed. Results were compared with the known
Forensic odontologists and anthropologists are ethnic provenance.
regularly asked by the police to describe the physi-
cal traits of unknown dead bodies. The purpose of MATERIAL
these descriptions is to assist the police in their The material belongs to different skull collections at
search for a missing person with a matching the University of Copenhagen. Three groups of

The Journal of Forensic Odonto-Stomatology, Vol.23 No.2, December 2005

3 and the metric data Fig. although some of the skulls had their ethnicity FORDISC programme computed a canonical or provenance written on the parietal surface The Journal of Forensic Odonto-Stomatology. and 25 Asians/“Chinese or Malay”. Palatal shape was recorded to be either horseshoe- shape.2. as these represent skulls retrieved from inhumation breakpoint 5-7. 29 Africans/ “Blacks”.5 or provenance written on them in the parietal region We therefore specifically chose the more recent 17th (Fig. The programme allows scoring in various eth- Eskimo material was excavated in Greenland during nic categories and it was decided to include scoring 17th-18th century. Most only had a general category. “Black”. as several of the individu- Inuit (i. ric traits. All skulls with intact crania were selected. The standards of the Arizona State University Dental Anthropology System (ASU) were used for registration. In some individuals the lower RESULTS jaw was missing and in several cases part of the den. “China”. USA 4) enamel extensions on upper first molars. All observations and measurements were made in lation better as it is known that there has been sig.1: All observations and measurements were made in was entered in the FORDISC* programme. although some of the skulls had ethnicity nificant commingling between Europeans and Inuit. blind trials. Table 1 shows the distribution of the dental non-met- tition was lost postmortem.6-11 Nineteen cranial measures were taken following the FORDISC programme manual. Four non-metric dental traits were recorded for each while Shovel-shape could be scored in only 27/80 skull: 1) shovel-shape on upper central incisors. This was due to postmortem loss of the inci- breakpoint 3-6 (c-f). 2) Carabelli trait on upper first molars. Ten skulls were some skulls did not have a precise archaeological re-evaluated a month later with no significant description. This was especially evident for the Inuit skulls. 3) mandibular torus. Cohen’s Kappa was used for evaluating and 18th century material for this study.9 To keep the recording as simple as possible description of palatal shape was based only on morphological evaluation and not on measure- ments.41 Use of an historical skull collection different ethnicity labelled as. Inuit or Asian. The lated. * University of Tennessee. The non-metric data were tabulated as frequencies and compared with published data.7 Along with the recording of the non-metric traits mentioned above. (postdating AD 1721) material. Vol. Greenland Inuit/“Eskimo”.7 Three non-metric cranial traits were recorded in the jaws. sors. but for this study only ethnicity was tabu- eral racial descriptor as “Black” or “Chinese”. or a gen.8 Mandibular and palatine tori were recorded by the ASU system. December 2005 . These skulls were collected assigned the skull to the most probable ethnic in the 17th and 18th century. prior to AD 1721) and post-colonial Inuit als may represent mixed ancestry.6. Three of the traits (Carabelli cusps. 1) palatal shape. Cusp number and Enamel Extensions) could be scored in METHODS most cases (50/80. cases.23 No. 80 adult skulls were analysed. “Chinese” and discriminatory analysis for each skull and thereby “Eskimo” were evaluated. The post-colonial population might reflect the present Greenland popu. 3) cusp number on lower second molars. and consists of both pre-colonial for the category “White”.e. In all. 2) palatine torus. Knoxville. The programme assigns both sex and geographic descriptor as “Africa”. 46/80 and 50/80. v-shape or parallel-sided as described by Byers et al. 26 difference between the results of the two evaluations. respectively). even though agreement of the predicted ethnicity. ethnicity. The blind trials.1). the investigators assigned whether the skull was African.

Inuit 0. only in two cases was it Dental Trait “Black” “Chinese” “Eskimo” Total possible to determine shovel shape Shovel shape present 0 4 2 6 among the 26 skulls of Inuit provenance. Man- “Chinese” 3 16 3 22 dibular and palatine tori were rare in the “Eskimo” 0 5 21 26 “Black” and “Chinese” groups while both Total 23 24 25 72 kinds of tori were frequently seen in the Inuit group (Table 5b). 5 cusps 8 7 10 25 Total 15 15 16 46 Enamel present 19 3 4 26 The dental non-metric traits provided a Extensions absent 1 7 16 24 good basis for deciding ethnicity or prov- Total 20 10 20 50 enance (Table 3).09 Asians/“Chinese” have the lowest agree- “Chinese”-Asian 0. 8-11 “Black” . The overall agreement was 0.26 0. and low Kappa values. December 2005 . Lynnerup.00 Cusp number “Black”-African 0. frequencies The agreement rates and Cohen´s Kappa values for each ethnic category Dental Trait Published data Recorded data is given in Table 7.79 Table 2: Distribution of jaw traits and Cohen´s Kappa was 0.47 0.11 ment.47 (4 cusps) “Chinese”-Asian 0. Due to the quality The Journal of Forensic Odonto-Stomatology. Overall Table 4: Non-metric agreement rates and Cohen´s Kappa values for the above mentioned characteristics of each ethnic category this material were also described in Ethnic category Overall agreement Cohen´s Kappa former publications. “Chinese” had v-shape.37 0.00 the non-metric traits: in both instances Carabelli´s cusp “Black”-African 0.20 0.Sejrsen.55 0.2.90 0.79 0. Jaw traits were compared with published data. It is seen that Africans/ Total 24 22 25 71 Palatal tori present 1 1 6 8 “Blacks” had the best agreement.73 1.50 0. These values are Shovel shape “Black”-African 0.39 generally comparable to the values for “Eskimo”-Inuit 0.00 “Chinese”-Asian 0. “Eskimo”-Inuit 0.23 No. It was possible to assign ethnic provenance to 72 of the skulls.78 “Chinese” . Indeed.73 All 0. (Table 5a).38 The purpose of this investigation was to Enamel “Black”-African 0. and Ethnic category African Asian Inuit Total “Black” 20 3 1 24 “Blacks” had parallel sided palates.55 DISCUSSION “Eskimo”-Inuit 0.52 morphology studies. For the non-metric jaw traits versus labels (rows) “Inuit” in general had horseshoe shaped palates. Ten skulls Table 5a: Dental non-metric frequencies compared with published were assigned to the “White” category. Mandibular tori present 1 0 8 9 absent 21 19 15 55 Total 22 19 23 64 The results for the dental non-metric traits showed that the traits in general followed the expected frequencies Table 3: Agreement between ethnicity as by dental traits (columns).Asian 0. and Mandibular Cusp number 4 cusps 7 8 6 21 Tori in 64/80) (Table 2).11 0.African 0.8 0.69.69 assigned ethnicity based on skull meas- urements in 71 % of the cases (Table 6).88 0.00 evaluate whether an old collection of extensions “Chinese”-Asian 0.55 skulls was of any value with regard to “Eskimo”-Inuit 0.0-0.81 0.59. while absent 24 21 19 64 the Asian/“Chinese” category had the Total 25 22 25 72 worst. Vol. Cohen´s Kappa was 0. Total 20 15 15 50 Palatine Tori in 72/80.14 0.17 0. absent 14 7 0 21 Conversely.55 The FORDISC programme correctly “Eskimo” . the three non-metric traits Total 14 11 2 27 Carabelli´s cusp present 2 2 0 4 of the jaws could generally be scored in absent 18 13 15 46 most cases (Palatal Shape in 71/80. Hejmadi 42 Table 1: Distribution of dental non-metric traits burials. The Jaw Trait “Black” “Chinese” “Eskimo” Total Palatal Shape horseshoe-shape 1 7 20 28 agreement rates and Cohen´s Kappa v-shape 10 12 4 26 values for each ethnic category is given parallel-sided 13 3 1 27 in Table 4.43 0.

We chose to record the non-met. although it was not clear what frequencies of tori.African 0.79 0.00 . Except for the ric traits. is the question evaluation.12 such labels are meaningless as the variation of The Journal of Forensic Odonto-Stomatology.47 “Eskimo” . but these data were collated from the pre-colonial part of the collections. “Eskimo”-Inuit 0.55 . Even though our study was conducted in blind trials.59 could have included skulls of European (Danish) extraction but chose not to.2. When the collections were established. It should be “Black” 21 3 1 3 28 “Chinese” 2 13 1 6 22 added that the FORDISC programme “Eskimo” 1 4 19 1 25 has data on Inuit skulls from our collec- Total 24 20 21 10 75 tions. indicating fair agree- quencies of the dental non-metric traits recorded in ment. nor how many races there these traits more often than the two other groups. However. Vol.55 . the Kappa statistics compared with published data and overall the fre.71 0. which in our experience are the most used morphometric analyses of the Asian/“Chinese” skulls. cies and skull morphometrics. but rather how these all traits in all individuals. It was thought that even just a few skulls follow the pattern described.54 0.43 Use of an historical skull collection Table 5b: Frequencies of recorded non-metric jaw traits As seen in Tables 4 and 7. Indeed. December 2005 .71 study was to evaluate ethnicity. of race. there was a Jaw Trait Recorded data Published data slightly better overall agreement when Palatal Shape “Black”-African 0. Table 7: Morphometric agreement rates and Cohen´s Kappa values for Even though the FORDISC programme each ethnic category has scored sex of the individuals. The lower agreement is to Horseshoe shaped some extent due to the fact that the Palatine Torus “Black”-African 0. this study were in accordance with data published by Scott and Turner.23 No.9 and seemed to institutions. When judging the non-metric traits. both in terms of the labelling of the cases.47.05 0. although for the case of “Chinese”-Asian 0. direct com. Concerning the different races. in order to see if the predicted catego- resulting in the fact that the majority of these were ries did reflect real ethnic differences in trait frequen- lost postmortem. These teeth have extremely short roots was included.0.35 0. traits were represented in the collection. (rows) were belonging to the categories “Black”. skulls ate this trait. Hrdlicka. as well as the tabulation of Palatal shape was based only on a morphological traits and measures by ethnic group.6 Underlying the whole project. There were some difficul. ties especially concerning shovel-shape on Inuit Cohen´s Kappa statistic for evaluating the agreement incisors.Asian 0. African Asian Inuit European Total “Chinese” or “Eskimo”. with a number of missing teeth was because the aim of the study was and fragmented jaws.47 assignment was made to only the three categories of provenance. were in the range 0.25 0. could clearly “define” the special characteristics of parison was not possible in all cases.75 As a test for non-metric traits per se.08 none of the cases was labelled as such. The results were compared with data were traded and swapped between anthropological published by Gill8 and Byers et al. we All 0. the Inuit group tended to have was meant by the term. even though “Eskimo”-Inuit 0. were. it is not published here since the aim of this Ethnic category Overall agreement Cohen´s Kappa “Black” . “Chinese”-Asian 0. (columns) the investigators knew that all cases versus labels. Inuit/“Eskimo” category the opposite was V-shaped “Eskimo”-Inuit 0. No measurements were taken to evalu.78. it was not possible to record to evaluate not the single traits.Inuit 0. in forensic odontological casework. in the collections there is a skull This is in concordance with data published by labelled as “Gypsy”.10 Petersen11 and Hauser et al.05 FORDISC programme was allowed to “Chinese”-Asian 0.87 0. Clearly. Table 6: Agreement between ethnicity as by morphometry.25 Mandibular tori “Black”-African 0.80 0. “Chinese” .46 using non-metric traits than Parallel sided morphometrics.50 the case.05 . score a case as “White”. However. and another as “Swede”. This of the skull material.04 0.89 0. Results were with a Kappa statistic of 0.

Jobling ences. Spencer Larsen C. Byers SN. it was found that traits and MA. USA. The acquisition of this collection. December 2005 . Email: bs@forensic. geographical differences in dental and cranial traits Publishers. In the case of the Greenland samples. The State of the Science. Wiley-Liss. No 3. Oregon State University Press. Inc. International Institute of Human Evo- be of value in forensic casework in terms of lutionary Research. 9. Hejmadi 44 population. 1997.Sejrsen. sometimes by indiscriminate exhumation of graves. 241-50. New York. eds. Forensic analysis of the Tel: +45 35326160 skull: Craniofacial analysis. as it must be foreseen that previous strict geo. the dead. Hauser et al. Advances in Dental Anthropology.0 . formed following permission from the authorities.142. and research may only be per. and sizes do exist. Meddelelser om Grønland. Forensic Anthropology Center. Scott GR. How Caucasoids Got Such Big Cra- we would argue that this may indeed be an advan. eds.13. for example. FL: CRC Press 2003:61-111. even inside smaller population groups. Am J Phys Anthrop 1940. as recorded for modern Danish population samples. but rather as “clinal” differ. an agreement has been made with the Greenland National Museum so that it now owns the 10. Ist. Rosser ZH. equate: what exactly is meant by the term “Black” or 4. Collections and The Journal of Forensic Odonto-Stomatology. Wolfe. Tenessee. Forensic Science Inter- population groups. Afro-Americans. Ousley SD. University of skulls. national. Nordby JJ. prejudices of the past. Curran B. and Amerindians repatriated. sometimes 8. nia and Why they Shrank . Although not all skulls have definite archaeological descriptions. Kelly MA. eds. North Carolina.23 No. discriminate groups. Institute of Forensic Medicine Boca Raton. The labelling of the skulls is clearly inad. Copenhagen 1949. Iscan MY. Turner II CG. Gill GE.ku. 42 (1):3- provenance. 2003. tage. 13. are reflected in the collections. Jantz RL. Greenland material. It may ultimately be fitting that these skulls may assist in better and faster identification of 12. Lynnerup. In: Biological Anthro- This historical collection does show expected dental pology. DENMARK 2. may Linda D. 1991. Forensic Science: An Forensic Odontology-Anthropology Unit. Boaz & non-metric and craniometric traits and. McFarland and Co. 1995. Bosch E.132:228-32.2. and not only be an example of racial Schweizerbart. Address for correspondence: REFERENCES: Dr B Sejrsen 1. graphical or ethnic categories will not be applicable in the future. 1993. cannot be considered ethical by today’s standards. 7. but it cannot be seen as definite. Epigenic variants of the Human Skull. cal material. Liebermann L. Quigley C. Germany 1989. 1993: 71-96. 11.From Morton to Rushton. Pedersen PO. Forensic Sci 1995. J Forensic Sci 1997. This means that the skull collection can be used as 6. Hrdlika A. odontologist or anthropologist to see these traits and perform morphometrical analyses. 9. Churchill SE. American Indians from Whites osteologically. Nørby S. as such. this requires a specific knowledge of from palatal dimensions. Human Bone “Chinese. USA. is much bigger than can be accounted for by a few Ed. as it does allow the forensic human teeth. Y-chromosomal STR haplotypes in Inuit and cranial measures. gen. Wiley-Liss. 27:1-55. as it is considered a scientific and archaeologi. The Problematics of “Race” in Contem- CONCLUSION porary Biological Anthropology. The material still resides at the University of Copenha. Fordisc 1. Vol. comparison and for teaching purposes. 40 (5):783-8. Current Anthropology 2001:42(1). Frederik V’s Vej 11 DK-2100 Copenhagen. reconstruction and Fax: +45 35326150 identification. The East Greenland Eskimo Dentition. Helmer RP. Eds: Noel T. J tries (at the time run by European doctors and clerks).14 However.Users guide. Goodman A. Identification of On the other hand. 5. while some material has been Euro-Americans. James SH. The Anthropology of modern training material. New York.69. introduction to scientific and investigative techniques. Mandibular and Maxillary Hyperostoses. 3. Challenge of the frontier: Discerning from local anatomical departments in far away coun. Lynnerup N. from Penang”? On the other hand. 2001:263. 14. Cambridge University Press. Skulls and Skeletons. Stuttgart.