U.S. Department of Justice
    Federal Bureau of Investigation
 FBI Seal graphic

Layout graphic: Topline




Graphic link to the Current Issue's table of contents

Graphic link to links to Back Issues

Graphic link to a search engine for all issues

Graphic link to the Journal Description page

Graphic link to the Editorial Board for this issue

Graphic link to the Instructions for Authors

Graphic link to the Handbook of Forensic Services

Graphic link to the FBI Lab Home Page

Graphic link to FBI Publications page (publications listing)

Sidebar graphic

Masthead - Forensic Science Communications
October 2000   Volume 2   Number 4

A History of Smithsonian–FBI
Collaboration in Forensic
Anthropology, Especially
in Regard to Facial Imagery

Douglas H. Ubelaker
Curator, Department of Anthropology
National Museum of Natural History
Smithsonian Institution
Washington, DC

Paper presented at the 9th Biennial Meeting of the International Association
for Craniofacial Identification, FBI, Washington, DC
July 24, 2000

Read about...
Photographic Superimposition
Facial Reproduction


For more than 60 years, physical anthropologists at the Smithsonian Institution have provided continuous consultation in forensic anthropology to their neighbors in Washington, DC—the Federal Bureau of Investigation. This work has involved recovery of evidence at the scene; determination of human or nonhuman status; estimation of sex, age at death, ancestry, living stature, and time since death; evaluation of postmortem change and trauma; and other factors that contribute to identification, including facial reproduction and photographic superimposition. Although the nature and intensity of this consultation have evolved over the decades, its roots extend to the very foundation of American physical anthropology.


Aleš Hrdlika (1869-1943) is widely regarded as a key figure in the early history of American physical anthropology. Hrdlika immigrated to the United States from Bohemia in 1881 and later received medical training in New York. As his interests shifted to anthropology, he was hired in 1903 as the first curator of physical anthropology at the Smithsonian. He remained there for the rest of his extremely productive career (retired in 1942 and died in 1943) and amassed large collections of human remains in support of his research.

Although Hrdlika was best known for his contributions to other aspects of physical anthropology (Stewart 1940), he also played an important role in the early development of forensic anthropology (Ubelaker 1999a). Early in his career, he had been active professionally on various medico-legal issues, such as the biological basis for abnormal behavior and aspects of epilepsy and insanity as they relate to criminal behavior. As his interests shifted toward comparative human osteology and anthropology in general, he gradually became involved in problems within the area of physical anthropology that is recognized today as forensic anthropology.

As early as June 13, 1936, Hrdlika's scholarship attracted the attention of the FBI. Referencing his testimony before a committee of the House of Representatives, FBI officials noted his impressive credentials and described him as "the best informed man in the United States on anthropology" (Ubelaker 1999a:728). A month later, the FBI advised its Jacksonville office to contact Hrdlika on a forensic matter.

More formal consultation between Hrdlika and the FBI followed. On February 11, 1938, the Director of the FBI, J. Edgar Hoover (1895-1972), wrote to the Secretary of the Smithsonian, C. G. Abbot, requesting Hrdlika's assistance in evaluating specimens thought to be human. He wrote, "In view of the fact that the Bureau's Technical laboratory does not have facilities to conduct such examination it will be appreciated if the Anthropological Laboratories will conduct this examination and report upon the results of their examination. In accordance with previous arrangements the specimens will be delivered by a representative of the Bureau's Technical laboratory." This brief summary of the arrangement describes the essence of the process that has continued uninterrupted to the present.

Records are not entirely clear, but those available suggest Hrdlika reported on at least 37 cases for the FBI and corresponded with FBI Director Hoover on related matters (Ubelaker 1999b). Hrdlika did not publish directly on this work, and even his long-term assistant, T. D. Stewart (1901-1997), had minimal knowledge of this area of his activity (Ubelaker 2000). However, on at least these 37 occasions, Hrdlika offered opinions as to whether remains were of human origin, and if so, the age at death, antiquity, sex, stature, ancestry, and evidence of foul play.

There is no evidence that Hrdlika attempted or was involved directly with facial reproduction from crania. In fact, in his report of a 1940 case that was apparently studied at the request of the FBI, Hrdlika declined a request of the originating office for a facial reproduction, arguing that the process is "while practicable to some extent, unsafe and could readily prove fallacious" (Ubelaker 1999a:727). Although Hrdlika obviously did not have a high opinion of facial reproduction, he apparently was aware of its applications. His personal offprint files include a December 10, 1916, newspaper clipping from the Sun in Baltimore, Maryland, describing a clay facial reproduction that led to an identification in a New York case (Ubelaker 1999a:727).

Working on an Arizona case in 1932, Hrdlika did attempt his own comparison of a recovered skull with stereoscopic photographs of the person likely represented. Hrdlika viewed the photographs through a stereoscope and then compared the three-dimensional head shape with the submitted skull. In his carefully worded conclusion, Hrdlika wrote, "My examination discloses that all features of this skull closely correspond with the aforementioned photographs and information, and not a single feature fails to correspond. Furthermore, the mere fact that the skulls of different individuals greatly vary, in view of the aforementioned close correspondence, indicates that this skull is in all probability that of [the missing person]" (Ubelaker 1999a:727-728). Note that although details of exactly how the comparison was conducted are not available, the effort preceded the well-known work of Glaister and Brash in the Scotland case (Ubelaker 1994) by three years.

When Hrdlika retired in 1942 and died in 1943, he was succeeded as Smithsonian Curator of physical anthropology by his student and associate of many years, T. D. Stewart. Stewart also assumed the role of primary consultant in forensic anthropology for the FBI and personally reported on at least 167 cases between 1946 and 1969, mostly of skeletonized human remains submitted to the FBI from law enforcement agencies. When Stewart was not available, others at the Smithsonian, especially Marshall T. Newman (1911-1996), examined the FBI cases, but Stewart remained the primary contact from about 1943 to 1962.

Back to the top

Like Hrdlika, Stewart adopted a cautious approach to facial reproduction and did not use the technique in his own casework. In an early review of facial reproduction (1954) Stewart was critical. In his 1979 discussion of the technique, however, he was more positive, apparently influenced by the experiments reported by Snow et al. (1970). In 1983, Stewart himself contributed to the developing science of facial reproduction by dissecting cadavers to define the relationship between Whitnall tubercles and the location of the points of attachment of the palpebral ligaments (Stewart 1983).

In 1962, Stewart accepted a position as Director of the Smithsonian's National Museum of Natural History and no longer had time for major casework. Responsibility for the FBI casework shifted to J. Lawrence Angel (1915-1986), a very prominent but newly hired physical anthropologist and anatomist. Although Angel had little prior experience with forensic work, he adjusted quickly (Stewart 1979), and by the end of his career he was widely recognized as a leader in the field. He remained the primary FBI consultant for the following 15 years and reported on 646 cases for the FBI and other agencies between 1962 and 1986. These cases included facial reproductions that on at least one occasion led to a positive identification (Cherry and Angel 1977). In 1986, Angel summarized many of his attitudes and anatomical approaches to facial reproduction.

In 1977, Angel pursued a much-deserved sabbatical year focusing on research writing, and I accepted his invitation and that of the FBI to assume responsibility for the FBI casework. This consultation is ongoing, and to date I have reported on 667 cases beginning in 1975, 404 of which originated from the FBI. Of these 404 FBI cases, 12 (3 percent) involved techniques of photographic superimposition, and 35 (9 percent) involved facial reproduction.

Figure 1 plots the percentage of FBI cases I studied over a 22-year period between 1978 and 2000 that included facial reproduction or photographic superimposition. Information after 1980 is grouped into five-year periods to examine long-term trends. This figure reveals a steady increase in requests for facial reproduction, likely reflecting increasing awareness of the value of this technique.

The figure also reveals that attempts at photographic superimposition began in 1990. Requests for this technique peaked in the 1990-1994 period and subsequently declined, with last use of the technique occurring in 1996. These frequencies appear to reflect the availability of the necessary equipment and expertise in 1990, coupled with awareness of the value of this approach in the forensic science and law enforcement communities. The decline in use likely reflects both the increased awareness of the limitations of this technique and the greater availability of more precise methods of identification, especially the molecular approaches.

Figure 2 presents the percentage of FBI cases studied by me that involved these two techniques, plotted for each year. Although the chart displays some temporal variability, it provides general documentation of the increasing temporal importance of these two techniques in the FBI experience. In 1995, these two techniques combined accounted for more than 60 percent of FBI forensic anthropology cases.

Bar graph showing the percentage of facial reproduction and photographic superimposition cases examined by Ubelaker for the FBI between 1978 and 1999

Figure 1

Percentage of FBI cases examined by Ubelaker between 1978 and 1999 that involved facial reproduction and photographic superimposition.

Click on image for an
enlarged view.

Bar graph showing the percentage of facial identification cases examined by Ubelaker for the FBI between 1978 and 1999, with data organized by year

Figure 2

Percentage of FBI cases examined
by Ubelaker each year between 1978 and 1999 that involved either facial reproduction or photographic superimposition.

Click on image for an
enlarged view.

Anthropological involvement in photographic superimposition and facial reproduction varies greatly depending on the remains and associated information presented. Submission of a full skeleton that has not been previously studied calls for a full anthropological analysis. This involves determining as many characteristics of the individual as possible and calling attention to those features most useful in facial comparisons and/or reproduction. Many of these analytical foci (human or not, sex, age at death, ancestry, time since death, living stature, evidence of disease or treatment of disease, taphonomical assessment, and evidence of foul play) date back to the days of Hrdlika. However, the specific techniques used and the accuracy of the methodology have changed dramatically (Ubelaker 1999c). Full anthropological analysis likely will reveal characteristics that were previously unknown and critical to the facial work (e.g., sex, age at death, ancestry, and general build). This information is needed not only to emphasize appropriate features in reproductions, but also to utilize appropriate soft tissue-depth data.

In addition, anthropological analysis serves to call attention to specific cranial-facial details that need to be incorporated into the reproduction or especially considered in photographic superimposition. Such features routinely consist of the size and shape of the supraorbital ridges; shape characteristics of the nasal area, as suggested by the morphology of the nasal spine and nasal bones; evidence of healed trauma; and unusual dental features, especially prognathism and unusual dental occlusion. These features can be highly individualistic and very useful for identification purposes. Experience is required to properly interpret such features and judge their uniqueness and likely expression in the living.

Some cases are submitted with aspects of this forensic information provided from other sources or from previous analysis by other specialists. In such cases, the remains are examined generally to ensure that observations are consistent with the information provided, and then the features specifically required for the technique requested are assessed. For example, such a case might consist of a relatively complete, slightly decomposed body recovered by a medical examiner for which standard autopsy procedures and soft tissue observations suggested sex, age at death, and ancestry. If no contradictory observations were made, anthropological analysis would focus on those features of the face most useful for the procedure requested.

Back to the top

Photographic Superimposition

The technique of photographic superimposition is employed in situations in which investigation has suggested the likelihood that a set of remains relates to a particular missing person for whom photographs are available, but positive identification has not yet been established. In such cases, images of the skull can be compared with the submitted photographs to establish the probability of identification. This procedure is most useful for exclusion, but positive identification is possible depending upon the uniqueness of the features observed. Comparisons are conducted through the use of digitized images on a computer in the manner described by Ubelaker et al. (1992). Figure 3 presents an example dating from 1991. As noted above, requests for this technique have diminished as molecular approaches have increasingly become available.

Graphic image showing the photographic superimposition of an unidentified skull and a photograph from a 1991 case

Figure 3
A 1991 example of the
comparison involved in
photographic superimposition.

Click on image for an
enlarged view.

Facial Reproduction

Techniques of facial reproduction have evolved over time on FBI cases, although the nature of anthropological input has largely remained constant. Throughout my years of involvement, this process has begun with the necessary anthropological analysis, placement of appropriate depth markers on the skull, and my selection of likely corresponding facial components from a large selection of facial photographs. The work then shifts to the FBI artist, who must marry the skull image and all information provided in an appropriate way. The manner in which individual artists choose to work varies with individual preferences and training. Most of the reproductions done during my early years of consultation involved individual sketches, sometimes using overlays of skull photographs.

In the late 1980s and early 1990s, computer approaches became available to facilitate the process of facial reproduction. One such approach, in which the skull image and facial components are digitized through the use of a video camera and then adjusted as needed using computer software, is summarized by Ubelaker and O'Donnell (1992). Variations of this approach continue to be used today depending on the preferences of the artist involved. Having the images available on computer provides a key advantage in facilitating the final comparison of the reproduction with the underlying skeletal tissue to make the reproduction as accurate as possible.

In a typical case, after I discuss the biological features and select the proper facial components, the forensic artist will complete the reproduction using techniques of individual choice. We then have a final review session, usually involving a computerized comparison of the suggested reproduction with the submitted remains, and agree on any additional changes that are needed. This collaborative team approach works well in my experience and brings to the case both an anthropological and an artistic perspective.

Figures 4 through 15 present 12 facial reproductions generated from FBI cases dating from 1978 through 1998. Collectively, these figures document a diversity of presentation that largely reflects variability in artistic approach.

Back to the top


Techniques of photographic superimposition and facial reproduction continue to represent a fascinating blend of art and science. As we learn more about the correlations of facial hard tissue anatomy and soft tissue facial features, and as artistic technique continues to become more powerful, we can only expect the impact of these procedures in forensic science to increase. As noted by Stewart in 1979, there is continued reason for optimism.


Angel, J. L. Graphic reproduction of head and face from the skull, Scientific Illustration1986 (1986):6–8.

Cherry, D. G. and Angel, J. L. Personality reconstruction from unidentified remains, FBI Law Enforcement Bulletin (1977) 48(8):12–15.

Snow, C. C., Gatliff, B. P., and McWilliams, K. R. Reconstruction of facial features from the skull: An evaluation of its usefulness in forensic anthropology, American Journal of Physical Anthropology (1970) 33:221–227.

Stewart, T. D. The life and writings of Dr. Aleš Hrdlika (1869–1939), American Journal of Physical Anthropology (1940) XXVI:3–40.

Stewart, T. D. Evaluation of evidence from the skeleton. In: Legal Medicine. Ed. R. B. H. Gradwohl. Mosby, St. Louis, 1954, pp. 407–450.

Stewart, T. D. Essentials of Forensic Anthropology. Charles C. Thomas, Springfield, 1979.

Stewart, T. D. J. Lawrence Angel, American Journal of Physical Anthropology (1979) 51:507–516.

Stewart, T. D. The points of attachment of the palpebral ligaments: Their use in facial reconstructions on the skull, Journal of Forensic Sciences (1983) 28:858–863.

Ubelaker, D. H. Cranial photographic superimposition. In: Forensic Sciences, Vol. 2. Ed. C. H. Wecht. Matthew Bender, New York, 1994, pp. 27C-1–27C-39.

Ubelaker, D. H. Aleš Hrdlika's role in the history of forensic anthropology, Journal of Forensic Sciences (1999a) 44:724–730.

Ubelaker, D. H. The legacy of Aleš Hrdlika in Smithsonian forensic anthropology, IVth International Congress of Aleš Hrdlika—Aleš Hrdlika—130th Anniversary of Birth. Ed. M. Prokopec. Czech Anthropological Society, Prague, 1999b, pp. 17–21.

Ubelaker, D. H. Human Skeletal Remains: Excavation, Analysis, Interpretation, 3rd ed. Taraxacum, Washington, DC, 1999c.

Ubelaker, D. H. The forensic anthropology legacy of T. Dale Stewart (1901-1997), Journal of Forensic Sciences (2000) 45:245–252.

Ubelaker, D. H., Bubniak, E., and O'Donnell, G. E. Computer-assisted photographic superimposition, Journal of Forensic Sciences (1992) 37:750–762.

Ubelaker, D. H. and O'Donnell, G. E. Computer-assisted facial reproduction, Journal of Forensic Sciences (1992) 37:155–162.

Back to the top