Introduction

The practice of tooth evulsion was widespread in the Maghreb (the northern parts of Morocco, Algeria and Tunisia) during the late Pleistocene and Holocene starting during the Iberomaurusian and continuing into the Neolithic (Briggs 1955; Camps 1974; Camps-Fabrer 1975; Chamla 1978). Tooth evulsion (also known as tooth ablation) involves the deliberate removal of healthy teeth during the lifetime of an individual, and has been recorded in a variety of ancient and modern societies around the world (Tayles 1996, Han and Nakahashi 1996). The visual impact of this type of dental modification is very striking and would have been immediately obvious to other individuals from the same or different communities. Numerous reasons have been proposed for the custom including group identification, ornamentation, and rites of passage such as coming of age, marriage and mourning. The specific social meaning of tooth evulsion is likely to remain obscure for ancient populations and may have diversified over time within those groups (Han and Nakahashi 1996).

Dental evulsion can have a significant effect on the emergence, occlusion and wear patterns of the remaining teeth. The removal of the upper central incisors leads to a particularly distinctive mandibular morphology (Fig. 1). Ferembach et al. (1962) noted that evulsion of the upper central incisors in individuals from Taforalt caused an elevation of the symphyseal region of the associated mandible. The absence of the occluding upper incisors causes the lower incisors to emerge beyond their normal occlusal plane and to exhibit very little wear in comparison to the rest of the dentition. The lateral incisors, canines and sometimes the anterior premolars show a sloped pattern of wear, increasing towards the back of the mouth, in a curved or occasionally stepwise fashion. This gradient is observed in the anterior mandibular dentition as a whole and on individual teeth. Typically the lateral incisors demonstrate the steepest gradient between the mesial and distal edges of the tooth. The alteration of the normal occlusal plane and the uneven wear patterns create the impression of an elevated symphyseal region that has been described by previous authors (Ferembach et al. 1962) and is sometimes referred to as “compensation” (e.g. Marchand 1936).This pattern is so distinctive that it can be regarded as evidence of evulsion of the maxillary incisors, even in the absence of an associated maxilla.

Fig. 1
figure 1

Characteristic changes in the emergence and wear of the lower dentition in a young adult male from Hattab II individual showing evulsion of the permanent upper central incisors. The lower incisors have emerged above their normal occlusal plane into the void created by the absence of the upper incisors, and exhibit very little wear relative to the rest of the dentition. The lower lateral incisors, canine and premolar display a sloping wear gradient, with wear increasing from the front to back of these teeth. In some individuals the visual impact of this uneven wear pattern is further emphasised as a result of premature loss of the posterior teeth caused by dental disease. Photo by Ian R. Cartwright

Full size image

Removal of the two lower central incisors at a sufficiently young age can cause the remaining lower teeth to migrate forward, causing a mesial inclination of the crown of each tooth relative to the root, and occlusal abnormalities (Briggs 1955). In the absence of any confirmed examples of removal of the mandibular incisors without evulsion of the associated maxillary incisors, it is not possible to demonstrate what effect this would have on the emergence and wear of the maxillary teeth. Camps (1974) and Camps-Fabrer (1975) both include records of individuals with evulsion of (just) the lower incisors, but we have not been able to confirm the existence of any individual showing this pattern of evulsion where both mandible and maxilla could be observed. Deliberate removal of anterior teeth can also have a marked influence on the morphology of the craniofacial skeleton. Alveolar bone loss associated with the removal of the upper incisors is associated with a noticeable reduction in facial height (nasion-prosthion) and other measurements that use prosthion as a landmark (Arambourg et al. 1934; Ferembach et al. 1962; Lubell et al. 1984). Hadjouis (2002) provides a detailed account of the effect of tooth evulsion during development on craniofacial architecture.

Over the years several authors have undertaken comprehensive or partial reviews of the pattern of tooth evulsion in the Maghreb during the Iberomaurusian to Neolithic periods (Briggs 1955; Camps 1974; Camps-Fabrer 1975; Chamla 1978). Previous reviews of tooth evulsion in the Maghreb have addressed three main factors that may contribute to variation in the expression and prevalence of this practice: (a) Differences between the Iberomaurusian and subsequent groups including Capsian and various different Neolithic groups; (b) differences between males and females and (c) variation in the age at which evulsion was practiced.

Camps (1974) noted that Iberomaurusian groups belonging to the Mechta Afalou type systematically practiced evulsion of the upper incisors, and that the practice continued into the Neolithic in related groups. The most common practice involved removal of the upper central incisors, and the minimum expression involved evulsion of one central incisor. Since tooth evulsion was almost universal during the Iberomaurusian, there is no clear difference between the sexes (Camps 1974). During the Capsian, tooth evulsion was not universal but when practiced, it typically involved both upper and lower incisors (Camps-Fabrer 1975; Chamla 1978). Briggs (1955) suggested that there was no regular sexual preference governing the practice of evulsion during the Capsian, pointing out that the presence and absence of evulsion could be observed in both males and females. More recently, Chamla (1978) observed that during the Capsian period tooth evulsion was more common in females (77.7% out of 18) than in males (46.6% out of 15). During the Neolithic, tooth evulsion could involve maxillary and mandibular teeth or just maxillary teeth and was not universally practiced. Several authors have suggested that distinct regional differences in the prevalence and pattern of evulsion had developed by the Neolithic. Camps (1974) noted that evulsion in the Mediterranean Neolithic commonly involved teeth from both jaws. Chamla (1978) reported that evulsion was rare in the east with only 1 out of 28 individuals affected, but more common in the west where 30 out of 42 individuals (71%) were affected. Males and females seem to have been equally affected during the Neolithic (Chamla 1978).

A number of different approaches have been used to ascertain the age at which evulsion was practiced (Briggs 1955). First, among large samples including children, adolescents and young adults, the age distribution of individuals with and without evulsion may reveal a consistent pattern. In practice it has only been possible to apply this approach to a small number of ossuary samples in which individuals of all age ranges are reasonably well represented (Arambourg et al. 1934; Chamla 1970). In the case of multiple tooth extractions, the possibility of an age related progression could also be investigated. Second the amount of wear on opposing teeth that were retained could reveal whether the occluding tooth was present for long enough to cause wear. Third, the extent of drift of teeth retained in the same jaw and the degree of inclination of their long axis could give an indication of when the gap appeared. Another potentially informative feature would be the degree of alveolar bone resorption and remodelling among young adults of different ages.

In this review we present an updated summary of the geographical and chronological distribution of different patterns of tooth evulsion in the Maghreb during the Late Upper Palaeolithic and early Holocene (Fig. 2), using stricter criteria than those employed by previous authors. We have excluded cases where there is uncertainty over the dating of a site or the relationship of a burial to the cultural horizon in which it was found. We include only those individuals represented by a sufficiently complete maxilla and mandible, and have excluded individuals where this is uncertain (for example at Aïn Keda in Algeria).

Fig. 2
figure 2

Distribution of archaeological sites with human remains from the Iberomaurusian (top) Capsian (middle) and Neolithic (bottom) periods in Morocco, Algeria and Tunisia

Full size image

Iberomaurusian Sites

The Iberomaurusian represents the earliest Upper Palaeolithic/Late Stone Age technology in the Maghreb and covers a time span of more than 10,000 years from >20,000 to at least 10,000 BP (Lubell 2001; Bouzouggar et al. 2008). The majority of Iberomaurusian human remains are from caves and rock shelters located fairly close to the present day coastline (Fig. 2). Data are available for eight sites (Appendix): Afalou-bou-Rhummel, Columnata, Hattab II, Ifri n’Ammar, Kef-oum-Touiza, Rachgoun, Taforalt, Taza.

Capsian Sites

The Capsian period dates from around 9,800 BP and can be divided into an earlier Typical Capsian and a later Upper Capsian (see Jackes and Lubell 2008). Most human remains assigned to the Capsian period were excavated from escargotières, small to large open-air mounds characterised by large quantities of edible land snails within a dark grey ashy deposit (Lubell 2001). Sites are inland and concentrated in Eastern Algeria and Tunisia (Fig. 2). Most known Capsian sites contain human remains and those that meet the criteria for inclusion here (Appendix) are: Aïoun Beriche (site 12), Aïn Bahir, Aïn Dokkara, Aïn Misteheyia, Bekkaria, Bir Oum Ali, Faïd Souar II, Grotte des Hyènes, Koudiat-el-Kherrouba, Khanguet el Mouhaâd, Mechta el Arbi, Medjez II, and Mesloug.

Neolithic Sites

Human remains assigned to the Neolithic belong to a number of different cultural traditions and are found in a variety of open air localities, caves and rock shelters, both coastal and inland (Fig. 2). Those included here (Appendix) are: Aïn Guettara, Aïn Meterchem, Columnata, El Cuartel, Damous-el-Ahmar, Grotte du Polygone, Izriten (Site 11), Medjez I and Mugharet al Aliya.

Discussion

The sites presented in this review represent only a small proportion of the localities which have yielded human remains from the Iberomaurusian, Capsian and Neolithic periods. Fragmentary remains from numerous other sites reveal evidence of tooth evulsion, either directly, where teeth are missing in one of the jaws, or indirectly, where an isolated mandible shows a pattern of tooth emergence and wear that is characteristic of individuals with evulsion of the maxillary incisors. Since it is not usually possible to establish the pattern of evulsion in the missing jaw, we have excluded material from these sites in this summary. Information from the other sites may be helpful in establishing the chronological and temporal boundaries of particular practices but does not allow patterns of evulsion to be reconstructed on an individual by individual basis. In other cases we have had to exclude material due to lack of information. In many site reports, particularly those relating to sites from the Neolithic, tooth evulsion is simply not mentioned. This could be for several reasons, including lack of a comprehensive anthropological report or poor preservation, and may not imply the absence of tooth evulsion at those sites. Poor excavation and curatorial standards have led to the loss of a great deal of material excavated in the early part of the last century or previously, and in several cases, skeletal remains were lost before they had been fully described (Balout 1954; Briggs 1955). An example is the site of La Mouillah in Algeria where fragmentary remains of at least 15–16 individuals were excavated from a series of rock shelters between 1908 and 1910 (Balout 1954). Early reports imply that evulsion of the upper incisors, typically the upper centrals, was common in this series (Marchand 1936), but no details concerning the age and sex of these individuals were published and by 1949 the series was considered to be too damaged for inclusion in a broader comparative study (Briggs 1955).

A further and perhaps ultimately more important problem concerns the absence of direct dating at many of the sites that have yielded human remains. Several potentially interesting samples were excluded from this study as a result of differences in opinion concerning their age and cultural affiliation, including Grotte de La Tranchée, Gambetta, Djebel Taya, Grotte d’Ali Bacha and Champlain (Arambourg et al. 1934; Balout 1954; Balout and Briggs 1949a, b; Chamla 1978; Camps 1974). In the absence of independent dating, the age of a skeleton has sometimes been inferred from the cultural context, burial rites and items associated with the burial, and even the state of preservation or pattern of tooth evulsion. It is also possible that important material from some early excavations has been overlooked. Camps-Fabrer (1975: 319) warns that archaeologists may have been too quick to assume burials that are shallow, buried in an extended position, or lacking evulsion, could not be Capsian even if found in an escargotière.

A summary of the prevalence of different patterns of tooth evulsion according to period and sex is presented in Table 1. Among the Iberomaurusian sample, 97% of individuals had undergone extraction of at least one incisor from only the upper jaw. The most common pattern of tooth evulsion involved removal of both upper central incisors. The only Iberomaurusian individual recorded as showing evulsion of both upper and lower incisors is an adolescent from Columnata (H1a) aged between 15 and 18 years, and in this case it may be reasonable to ask whether the burial has been correctly identified as belonging to the Iberomaurusian period. Unusually for this period, one adult male, from Kef-oum-Touiza, lacked tooth evulsion, but this individual had an abnormal fusion of the first and second upper right incisors and a small natural midline gap in the maxilla, so the absence of evulsion may be due to the unusual natural configuration of the anterior dentition. Evulsion of the upper two central incisors can be considered characteristic but not diagnostic of the Iberomaurusian period.

Table 1 Summary of tooth evulsion according to period and sex
Full size table

Within the scope of this survey, the Transitional Epipalaeolithic is represented at a single site, Columnata. The minimum expression of tooth evulsion among adults within this sample involved removal of the four upper and lower central incisors. Half of the adult sample had experienced removal of all eight permanent incisors, and a further quarter had lost at least one lateral incisor. Since all adults in the sample had experienced tooth evulsion, there were no differences in the prevalence of this practice between males and females. The average number of teeth lost to evulsion was marginally higher in males (6.6) than in females (6.3.)

Within our Capsian sample, five males, one female and two individuals of undetermined sex showed no evidence of tooth evulsion, one individual of undetermined sex showed evulsion in only the maxilla, and three males, seven females and two individuals of undetermined sex had lost teeth in both jaws. In total, 62% of Capsian individuals experienced tooth evulsion, but the practice was noticeably more common among females (88%) than among males (38%).

The practice of tooth evulsion remained widespread in the Neolithic but very little material was sufficiently complete or sufficiently well described for inclusion in this study. Within our small sample, only three out of 11 individuals (27%) showed evidence of tooth evulsion. Of these, two individuals showed evulsion in only the upper jaw and one individual had lost teeth from both jaws.

Inferences concerning the age at which evulsion was practiced are very preliminary since few cranial remains of juveniles and adolescents have been recovered and fully described. At Afalou-bou-Rhummel, the youngest individual in the sample showing tooth evulsion was aged between 14 and 16 years, but two other adolescents, aged 12–14 years and 16 years, still have their incisors intact. One child of approximately six years had lost an upper deciduous incisor some time prior to death, but this loss may have been coincidental since three other children aged between 3 and 6 years showed no evidence of tooth evulsion. At Columnata, the youngest individual in the sample showing tooth evulsion was aged at 12–15 years (H43) whereas several juveniles aged between 6 and 10 years year lacked any evidence of evulsion. On this basis Chamla (1970) tentatively suggested that evulsion may have been practised on individuals aged between 10 and 12 years, which may be slightly earlier than what is implied at Afalou-bou-Rhummel. One juvenile aged approximately 11 years from the Capsian site of Aïoun Beriche lacked tooth evulsion, whereas three adults from the site had undergone evulsion of all eight incisors (Mary Jackes and David Lubell, personal communication).

This survey reveals marked regional and cultural variation in the prevalence of tooth evulsion and the manner is which it was expressed in the Late Pleistocene and early Holocene populations of the Maghreb, but these variations are not sufficiently distinct to be considered culturally diagnostic at the individual level. During the Iberomaurusian, tooth evulsion was almost universal but typically restricted to the upper central incisors. Among the epipalaeolithic sample from Columnata, evulsion remains universal and was extended to include removal of at least four upper and lower incisors in both males and females. During the Capsian the practice of tooth evulsion became less prevalent and more variable suggesting a possible change or diversification of social meaning. This diversification is notable in view the relative spatial proximity of the Capsian sites included in this survey (Fig. 2) as it may point to localised variation in the underlying meaning or differential group identification within a relatively restricted region. The marked bias towards females among Capsian individuals who experienced tooth evulsion supports this interpretation (see also Rahmani 2003: 296–7). The prevalence of tooth evulsion was further reduced in the Neolithic, but the small sample discussed here probably fails to represent a diversity of traditions among different Neolithic groups. The underlying motivation for this type of dental modification may never be known but further work focussing on the age of occurrence and other subtle variations between cultural groups may contribute to a greater understanding of this practice, which showed remarkable persistence for a period of more than 10,000 years.