An evaluation of the miles method of ageing using the Tepe Hissar dental sample.код для вставкиСкачать
An Evaluation of the Miles Method of Ageing Using the Tepe Hissar Dental Sample GEORGE WESLEY NOWELL Department of Anthropology, University of Pennsylvania, Philadelphia, Pennsylvania 191 74 KEY WORDS Ageing methods Tepe Hissar - Miles method . Dental wear . ABSTRACT The Miles system of ageing, based upon analysis of the rate of molar wear, was evaluated using the available dental sample from Tepe Hissar, Iran. The independently estimated ages for the mandibles and maxillae of the same individuals were found to be highly correlated (r = 0.87, p < 0.001). Ages of a subsample of the dentitions were compared with skeletal ages for the same individuals estimated from pubic symphyseal faces and found t o be significantly correlated (r = 0.82, p < 0.005) with no significant differences in the mean ages. Although a complete evaluation of the Miles method would require its testing on a controlled population, the available dental sample from Tepe Hissar provided evidence of the reliability and validity of the Miles method of ageing archeological populations on the basis of dental wear. Due to selective retention of burial remains, the available skeletal sample from Tepe Hissar 1-111 is not representative of the human populations of the site and can not be used to generate meaningful demographic statistics. Since dentitions have more often been recovered and retained for study than non-dental materials, a valid method of determining ages based upon dental materials is of considerable importance in the study of archeological populations. This paper evaluates a method of ageing based on dental wear previously described by Miles ('63). THE SAMPLE The dental sample used in the present assessment comes from the archeological site of Tepe Hissar 1-111,Iran and dates to ca. 4000-? (1900-1300) B.C. (Dyson, '68a,b). The majority of the dental material dates t o between ca. 2400-1300 B.C. None of the few skeletal remains from Tepe Hissar dating from the Partho-Sasanian or later periods is considered in this report. The site was excavated by Erich F. Schmidt in 1931 and 1932 (Schmidt, '33, '37) and the skeletal material was studied in part by Krogman ('40a,b,c) and Kappers ('34). It was not possible t o derive reliable demographic data from this sample. The basic conditions necessary for the generation of demographic statistics from archeological skeletal remains are: AM. J. PHYS. ANTHROP. (1978)49: 271-276. (1) the sample must accurately represent the original populations of the cultural levels of the site, and (2) each skeleton must be attributable to its proper level in the site. The evaluation of the present skeletal sample in these terms indicates that neither condition is adequately met. Since all of the skeletal material was found on the elevated portions of the site, there is no assurance in the final site report (Schmidt, '37) that the skeletal population excavated was not biased with respect to age or sex by cultural factors regulating the place of interment. Schmidt ('37) also did not test extensively for alternate burial locations in the unexcavated portions of the site and in the surrounding plain. Both Krogman ('40a) and Schmidt ('37) note a marked predominance of males over females in each level and it can not at present be determined why this should be so. No attempt was made to reassess the sex of the individuals in the sample during this study. Only a small fraction of the individuals excavated were represented in the available skeletal sample. Although Schmidt ('37) re- 271 272 GEORGE WESLEY NOWELL corded the excavation of 1,637 burials, Krogman ('40a,b,c) was sent fewer than 250 individuals to study and the author found skeletal remains of only 479 individuals in the collections of the University Museum of the University of Pennsylvania (table 1).The generally excellent condition of the bones and their small number relative to the total excavated population suggests that Schmidt brought t o the United States only the best-preserved bones. A comparison of the age distribution of this study to Schmidt's ('37) figures shows a great disparity in the number of juveniles, aged zero t o six years. In the present sample there were 19, compared t o 151 excavated during the second working season alone. There may be a similar bias against the presence of older individuals (45 years of age a t death) in the sample available for analysis. The extent of these possible biases is unconfirmable because of the unreliability of the ages determined in the field. The excavator's field records and reports + TABLE I Inventories of the Tepe Hissar I-lllskeletal material Krogman ('40a) Crania Mandibles Post cranial Individuals represented Present study 193 138 58 ' 216 211 238 299 479 ' Parts of 58 individuals Parts of 299 indinduals often do not attribute individual burials and skeletal materials to specific cultural sublevels. Thus, sampling and recording problems involving the skeletal materials both as excavated and as retained for study preclude their use as indicators of the demography of the populations of Tepe Hissar 1-111. These sampling problems do not significantly affect the evaluation of the Miles system of ageing dental remains, since all dentitions came from one site and the individuals were exposed to the same general diet. All 268 individuals from Tepe Hissar 1-111represented by at least one dentition with molars ageable by the Miles method were aged. The sample used to evaluate the Miles method consists of all 139 individuals represented by both ageable upper and lower dentitions and includes 120 adults (aged 15.0 years and older) and 19 juveniles (table 2). METHODS The Miles method of ageing (Miles, '63) assumes general uniformity of dental wear rates and patterns throughout the population studied. Only molars are used in the method, since the molars demonstrate the greatest degree of consistency in wear patterns and rates. The Miles method can only be employed on a dental population which includes dentitions of a t least 20 immature individuals ageable t o between 6 and 19 years on the basis of dental development alone. The immature dentitions serve as the first baseline for the TABLE 2 Comparison of the age estimates assigned to the mandibles and maxillae of the same individuals Campanson of means of estimated ages Group Correlation of ages assigned to same individuals N Adult (15.0years and older) Aged by Miles method Aged by Miles method (but nooccluaal partners) Aged by dental development Juvenile (younger than 15 years) Aged by Miles method Aged by dental development Entire sample h t i r e sample (less those with no occlusal partners) ' Pearson correlation coefficient. 't-test for the significance of r. * p < 0.0005. All agea in years. r' Mean age t2 Mandible Maxilla t 19.14' 19.58' 26.85 27.17 26.87 27.44 0.00 0.09 0.46 0.09 39.50 17.50 32.75 17.44 0.72 0.03 10.05 12.00 9.94 0.55 120 107 0.87 0.89 4 9 -0.31 0.03 19 1 18 0.95 12.04* 0.95 11.67* 9.58 11.00 9.50 139 0.93 29.90' 24.49 24.57 0.02 135 0.95 34.86' 24.05 24.33 0.10 - - - 0.51 273 MILES METHOD OF AGEING TEPE HISSAR DENTITIONS TABLE 3 Comparison of the age estimates assigned to individual mandibles and maxillae on two occasions (age-reagereliability su bsamplesl Comparison of means of estimated ages Correlation of age8 assigned to same individuals Dentition Mandible Maxilla Mean age ' " r2 t 3 Ageing Reageing 20 0.99 0.99 30* 30' 28.8 28.8 29.65 20 29.8 t 0 0.03 The mandibular and maxillary subsamples used were not identical. Pearson correlatian coefficient. 3t-test for the significance of r. * p <0.0005. All ages in years. I establishment of the basic rates and patterns of dental wear used to accomplish the rest of the ageing. The development of the baseline is dependent upon the comparison of the wear of the second and third molars of slightly older dentitions with similar stages of wear of first and second molars representing functional ages in the existing baseline. The functional age of a tooth is the number of calendar years a tooth is in occlusion with its partner in the opposing dentition. The older dentitions are assigned estimated chronological ages as described above and are integrated into the baseline t o extend the baseline t o the chronological limits of the population. Adding the functional age t o the eruption age of a particular molar yields an estimate of the chronological age of the individual a t death. Conversely, subtraction of the eruption age from the chronological age a t death yields the functional age of the stage of wear of the tooth. It is through the latter process that the baseline is established and through the former that the majority of the individuals are aged. The extended baseline is expanded during the study by incorporating into it those dentitions best illustrating each different pattern or degree of wear assigned to the same chronological age. In this study, first the mandibles were assigned ages in accordance with the Miles method outlined above. After the maxillae had been aged by the Miles method independently of the mandibles, each mandible was articulated with its maxilla if available and the age estimates were compared. After the entire dental sample had been initially aged, subsamples of 20 maxillae and 20 mandibles were selected randomly from the non-baseline materials. The subsamples were re-aged independently of the initial ageing against the appropriate baselines in the identical manner outlined above to assess the reliability of the author in the application of the Miles method of ageing. Also, all available pubic bones belonging to the individuals aged by the Miles method were aged according to the standards of McKern and Stewart ('57). RESULTS The insignificant differences in the mean ages and the correlation coefficients of the first and second ageings of the subsamples of maxillae and mandibles re-aged by the Miles method indicate the high degree of reliability of the author in his use of the Miles method (table 3). All individuals in the skeletal population for whom both the maxilla and mandible were present were divided into five subgroups on the basis of age, manner of ageing and the presence or absence of occlusal partners (table 2). There were no significant differences (p > 0.05) between the mean ages of the mandibles and maxillae in any of the subgroups, as indicated by the t-test for dependent means. The correlation coefficients between the estimated ages of the mandible and maxilla for the same person are r = 0.87 (table 2). For the entire sample and for each group with a size greater than ten, the correlation was significant (p < 0.005). These data suggest the reliability of the Miles method as used in this study. The absolute differences between the ages assigned to the mandible and maxilla of the same individuals were tabulated. Sixty-six percent of all the adults have ages assigned to both the mandible and maxilla of the same person within two years of each other and 93% have both ages assigned within five years of each other. In the absence of those cases in 274 GEORGE WESLEY NOWELL TABLE 4 Comparison of the age estimates assigned t o the dentitions andpubic bones of the same individuals Comparison of means of eatimated ages Subsample Maxilla-pubic Mandible-pubic Correlation of pubicdentition ages Mean ages N' r2 Dental Pubic Difference in mean age t 12 17 0.82* 0.88* 25.83 28.41 27.67 28.41 1.84years 0 1.46 0 ' Eight individuals were common t o both subsamples Pearson correlation coefficient. * p < 0.005 which all the occlusal partners of the ageable molars were lost ante-mortem, the correspondence of the estimated ages is notably closer. The mean estimated ages of the pubic bones and dentitions subsamples were not significantly different (p > 0.05) for either the mandible or the maxilla subsamples (table 4). The correlation coefficients between the estimated ages of each dentition and the pubic symphyses for the same persons were significant (p < 0.005) (table 4). These data support the validity of the Miles method. all three molars appeared and the final baseline well represented the consistent rates and patterns of wear. The sequence for the third molars was not as consistent as those of the first and second molars on a year-to-year basis, probably because the third molars did not erupt or become functional as regularly as did the first and second molars. Although Miles ('63) employed a wear rate ratio of 6 : 6.5 : 7 (M, : M, : MJ for the Anglo-Saxon population he described, there appeared t o be no consistent difference in the rate of wear between the first and second or first and third molars in the present study. DISCUSSION The observed differences in wear pattern To assess the validity of the statistics pre- could not be ascribed to any testable variable sented above, it is necessary to review the (for instance cusp pattern or sex) and are hypotheses upon which the Miles method is assumed to be due to differences in (1) individbased, the sources of error inherent in the ual occupations andlor habits, (2) chewing system, and the results of previous efforts to patterns, (3) diet or manner of food preparaage this population using different ageing tion, (4) tooth composition either due to variations in the genetics or the environment of the methods. It is hypothesized in the Miles method (1) developing individuals (or both), or most likethat dental wear is continuous during the ly to some combinations of several of the functional life of the tooth, (2) that wear on above. The differences in dental wear patterns all three molars proceeds a t comparable rates are most apparent in several individuals with through similar states and patterns and (3) dentitions worn relatively flat, but with that wear and diet are effectively uniform minimal dentine pits. The dental wear patwithin the population. Unfortunately, (1) terns from Tepe Hissar vary substantially quantification of the effect of tooth loss or de- from those of the Australian population discay on the rate and patterns of wear of the re- cussed by Murphy ('59) and had worn more maining teeth is a moot point; (2) that wear rapidly than those of the Anglo-Saxons illuson all three molars proceeds a t comparable trated by Miles ('63). Although the diet of the people of Tepe rates even in the absence of external changes, such as loss or decay of teeth, is unproven; and Hissar has been hypothesized t o have been (3) different patterns of dental wear are ap- uniform and based on the products of domestiparent in this population. There remains no cated plants and animals, the individual diets conclusive evidence that cultural factors af- undoubtedly varied due to economic exigenfecting dental wear rates and patterns did not cies and social or personal preferences. All vary significantly through the cultural levels variation in culture, genetics and environment would also adversely affect the validity of Tepe Hissar 1-111. However,in both the maxillary and mandib- of the dental ages assigned to the degree that ular baselines, a generally consistent, clear it affected the rates of wear of the permanent progression of degrees and patterns of wear in molars. However, one of the strengths of the MILES METHOD OF AGEING TEPE HISSAR DENTITIONS Miles method is its population specificity which enables a properly constructed baseline to reflect the normal range of dental wear rates and patterns resulting from such variations. By establishing the baseline on the basis of dental development through 19 years, the circumstances reflected are those of adults as well as juveniles. A baseline specific to one population can be used with another population only if variations in the factors affecting the rates and patterns of dental wear are nonsignificant. An important source of error in the Miles method is that any inaccuracies in the earlier parts of the baseline based on dental development would be retained throughout the extensions of the baseline. In fact, Garn et al. (’59) report that Kronfeld’s (‘54)chart, relied on by Miles (‘63) and in this study, is too narrow in its range of variation attributed to the stages of deciduous and permanent dentition formation. However, (1) elimination from the baseline of those dentitions with a large variation in the assigned molar ages reduces the potential error due to using one or more of the divergent molars; (2) the incorporation into the baseline of dentitions with different wear patterns expressly recognizes variation in the progression and patterns of dental wear; and (3) the averaging of the ages arrived at for several molars, each compared with different parts of the earlier baseline, helps to reduce the effects of any single part of the baseline and yields a more reliable age for the individual. It is unlikely that the baselines developed in this study were markedly skewed from their true chronological analogs, since the differences between the means of the ages of the maxillae, mandibles and pubic symphyses of the same individuals aged according t o independent baselines and methods are not significant (tables 2,4).However, since it is assumed that the first, second and third molars become functional during the sixth, twelfth and nineteenth years respectively, an individual whose teeth erupt or become functional irregularly or uniformly late/early would be aged more inaccurately than other individuals. The more variable the age of eruption or commencement of function of the third molars, the more difficult it is to establish a valid baseline and accurately age the adult population. Although table 2 indicates the high reliability of the Miles method, there were 17 cases where the difference between upper and lower dentition is five or more years. Esti- 275 mated age differences between the mandible and maxilla of the same individual of up to five years may be accounted for completely by the differences in the wear patterns of the two baselines in all cases but one. Differences of five years and greater in the age determinations were due to: (1) the ante-mortem loss of the occlusal partners of all ageable molars (4 cases) ; (2) ante-mortem loss and/or presence of caries or alveolar abscesses in the vicinity of the occlusal partners of one or more molars probably contributing to peculiar wear patterns and rates of parts of each dentition (4 cases); (3) extremely different, but common wear patterns or rates of the mandibular versus maxillary molars for no discernable reason yet so different as to suggest that the dentitions represent two different persons ( 3 cases); (4) unique wear pattern of one or both dentitions (3 cases); and (5) the small number of molars available (3 cases). These data indicate that although a careful ageing of both the mandible and maxilla of one individual will usually yield age estimates within five and even two years of each other differences of nine years can be due solely to different wear rates or patterns of the maxillary and mandibular molars. By articulating and evaluating the mandibles and maxillae before estimating the age of the individual, peculiar circumstances affecting the wear of the molars can be noted and a more valid age estimate developed. In this study, mandibles and maxillae of the same individuals were not articulated prior to ageing only in order to evaluate more objectively the reliability of the Miles method by comparing independently aged maxillae and mandibles. The validity of the Miles method can be assessed by comparison of its age estimates with those based on endocranial suture closure or on the pubic symphysis. In this case, comparison of the results of the Miles method with the results of each alternate method is not completely satisfactory since the samples are not the same. In the case of endocranial suture closure, Krogman (‘40a) aged 154 adult crania and reported a mean age of 29.1 years. The present study aged maxillary or mandibular dentitions of 133 of the Krogman sample with a mean age of 28.1 years. Mandibles or maxillae with no ageable molars were available for 11 more members of the original sample with no dental evidence apparent for the remaining 10 members. Since the samples were not identical, the 11 unageable individ- 276 GEORGE WESLEY NOWELL uals would probably be aged older than 30 years if molars were present and there is usually some difference between dental and skeletal ages, the observed differences in t h e estimated mean ages are not important. Comparison of dental and pubic symphysis age estimates (table 4) is conservative because of t h e difficulty of positively attributing the pubic bones and dentitions to the same individuals and t h e normal variation between the dental and skeletal ages of the same individual. Since McKern and Steward validated their procedures while working with a contemporary all-male sample, the application of their method to a 4,000- to 6,000-year-old population and especially to female pelves would introduce an additional source of error (Gilbert and McKern, '73). The favorable comparison of the dental ages with skeletal ages for the same individual further substantiates the equation of one dental functional year with one chronological year. It is apparent from these results that the Miles method can be a valid method of ageing skeletal populations. ACKNOWLEDGMENTS Special thanks is owed Erich Schmidt for his preserving so much skeletal material from Tepe Hissar at a time when excavators usually discarded bones. Doctor W. M.Krogman has been very helpful throughout this study both because of his past meticulous study of the collection, and his continuing interest in t h e relationship of physical anthropology to archaeology and forensic medicine. I thank Doctor Robert H. Dyson, Jr., Professor of Anthropology and Curator of the Near Eastern Section of the University Museum a t the University of Pennsylvania, for his helpful criticism of the manuscript in its early stages. I am grateful to Doctor Alan E. Mann, Associate Professor of Anthropology at the University of Pennsylvania, who as my advisor gave encouragement, constructive criticism, time and help throughout this project. I thank Mr. Robert White and the members of the Department of Radiology a t the Hospital of the University of Pennsylvania for their generosity in the preparation of the X-rays used in this study. And, I am especially indebted to my wife, Carla Senders Nowell, for her critical reading of the manuscript and her advice on the statistical content of this paper. LITERATURE CITED Dyson, R. H., Jr. 1968a Annotations and corrections of the relative chronology of Iran. Am. J. of Archaeology, 72: 308-313. 1968b The archaeological evidence of the second millennium B. C. on the Persian plateau. Cambridge Ancient History (Rev. Ed.) fascicle 66. Cambridge University Press, Cambridge. Garn, S. M., A. B. Lewis and D. L. Polacheck 1959 Variability of tooth formation. J. of Dental Research, 37: 135-148. Gilbert, B. M., and T. W. McKern 1973 A method for aging the female 0s pubis. Am. J. Phys. Anthrop., 38: 31-38. Kappers, C. U. A. 1934 An Introduction to the Anthropology of the Near East in Ancient and Recent Times. N. v. Noord-Hollandsche uitgevers maatschappij, Amsterdam. Krogman, W. M. 1940a Racial types from Tepe Hissar, Iran from the late fifth to the early second millennium, B. C. Verhandelingen der Koninklijke Nederlandsche Academie van Wetenschappen, Afdeeling Natuurkunde (Section 21, 39(2):1-87. - 1940b The peoples of early Iran and their ethnic affiliations. Am. J. Phys. Anthrop., 26: 269-308. 194Oc The skeletal and dental pathology of an early Iranian site. Bulletin of the History of Medicine, 8: 28-48. Kronfeld, R. 1954 Development and calcification of t h e human deciduous and permanent dentition. In: Basic Readings on the Identification of Human Skeletons. T. D. Stewart and M. Trotter, eds. Wenner-Gren, New York, pp. 3-10. McKern, T. W., and T. D. Stewart 1957 Skeletal changes in young American males. Quartemaster Research and Development Command Technical report EP-45, Natick. Miles, A. E. W. 1963 The dentition in assessment of individual age in skeletal material. In: Dental Anthropology. D. R. Brothwell, ed. Pergamon, New York, pp. 191-210. Murphy, T. 1959 The changing pattern of dentine exposure in tooth attrition. Am. J. Phys. Anthrop., 17: 167-178. Schmidt, E. F. 1933 Tepe Hissar Encavations 1931. The Museum Journal (University Museum, University of Pennsylvania), 23: 323-487. 1937 Excavations a t Tepe Hiasar Damghan. University of Pennsylvania Press, Philadelphia.