SEARCH

SEARCH BY CITATION

Keywords:

  • fetal osteology;
  • infant burial;
  • Ochtendung;
  • twins

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Materials
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

Compared with other mammals, multiple births are rare in humans. The attitude towards multiple births varies widely among cultures. Although expected to be present, evidence for twin burials from prehistoric times is scarce. However, knowledge about the attitude of ancient societies towards twins, as expressed for instance by funeral practices, would provide interesting insights into the spiritual world of these people. In 2001, an Iron Age settlement site was excavated near Ochtendung, Germany. The skeletons of two perinatal individuals were found in situ in shallow depressions close to post-holes of a house pit in the western part of the settlement. The individuals appear to have been deposited in the depressions without special care, as one of them was face down. No archaeological findings were associated with the skeletons. Age at death of the two individuals was estimated to be between 36 and 40 weeks' gestation. The cusp patterns of the lower left first molars exhibit a similar configuration of the disto-buccal cusps in both individuals, probably a division of the hypoconid. The investigation of the bones and the teeth revealed a similar stage of development of the individuals and supports the hypothesis of closely related individuals, probably twins. An interpretation as ‘building sacrifices’ seems unwarranted, considering the careless deposition of the bodies and the absence of any grave goods. Copyright © 2012 John Wiley & Sons, Ltd.


Introduction

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Materials
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

Compared with other mammals, multiple births are relatively rare in humans, with frequencies varying between eras and populations (Ball & Hill, 1996; Hoekstra et al., 2008). Reported frequencies range between about 1.3% (Hoekstra et al., 2008) and 3–3.5% (Shinwell et al., 2009; Chauhan et al., 2010). In many countries worldwide, the frequency of multiple births has increased with the onset of artificial reproductive techniques and a rise in maternal age (Tough et al., 2000; Chauhan et al., 2010). This can be attributed to an increase in the frequency of dizygotic twinning, whereas the frequency of monozygotic twinning remained relatively stable over time (Hoekstra et al., 2008).

The attitude towards multiple births varies widely among cultures. In some, multiples are viewed as a special fortune to the families, whereas in other cultures, the opposite is the case and multiple births bring misfortune and shame (Pector, 2002). As a consequence, twin infanticide is thought to have been practised in many cultures (Pector, 2002); however, this view is not shared by others (Ball & Hill, 1996).

Although expected to be present, evidence for twin burials from prehistoric times is very rare (Crespo et al., 2011). One reason for the rarity of evidence might be that the suspicion that two or more buried individuals are twins is commonly not aroused, and hence, a systematic search for traits that could support this assumption is not conducted. Suspicion of a twin burial may arise, however, when individuals are buried close to each other (double or multiple burials), in peculiar positions (e.g. children or newborns buried on top of each other), are of the same age at death, and share skeletal and/or dental traits. For example, the skeletons of two infants from the Upper Palaeolithic site of Krems-Wachtberg, Austria, were placed close to each other in flexed body position, both with the head oriented north and their faces east. The obviously contemporaneous burial of the individuals and the similar stage of development of their skeletons led the investigators to conclude that the individuals were probably twins (Einwögerer et al., 2006). A similar case is reported from Olèrdola, Spain, a site dated to the fourth century BC, where two perinatal individuals were buried on top of the each other in the same grave (Crespo et al., 2011). Again, the investigators concluded that the individuals were most likely twins.

The present paper describes a possible twin burial from the local Iron Age Hunsrück-Eifel-Culture, spanning the period from the seventh century to about 250 BC.

Materials

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Materials
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

In 2001, an excavation was conducted at a surface mining area near Ochtendung (Emminger Höfe), state of Rhineland-Palatinate, Germany, where volcanic pumice from the Laacher See volcano eruption is industrially mined. During these mining activities, an Iron Age settlement from the late Hallstatt period or the early La Tène period (represented by the local Hunsrück-Eifel-Culture) was discovered and a systematic excavation of the site was conducted. Although the findings were predominately of an archaeological nature, osseous remains of three individuals were excavated as well. One individual, an adult man, was found in a pit in the north of the settlement (this individual is not considered in the present study). According to the interpretation of the archaeologists, the pit originated when the soil for house building was removed (Materialentnahmegrube; Axel von Berg, personal communication). The other two individuals (162/331 and 162/338) were discovered in the western part of the settlement in a pit, which was interpreted as a remnant of a so-called pit-house (site 162; Figure 1). The pit measured approximately 3.60 m in length and 2.60 m in width. A number of post-holes were evident on the floor of the pit. Directly connected to two of these post-holes were shallow depressions, one on each of the smaller sides of the pit-house (post-holes numbers 1 and 11). In each of these depressions, the skeleton of a perinatal individual was found in situ, individual 162/338 in hole 1 (Figure 2) and individual 162/331 in hole 11. The individuals appeared to have been deposited in the depressions without special care, as one of them was face down. No archaeological findings were associated with the skeletons.

image

Figure 1. House pit of site 162 during excavation (left). The sketch (right) shows the position of the post-holes (numbers) and the depressions where the skeletons were found (asterisks). This figure is available in colour online at wileyonlinelibrary.com/journal/oa.

Download figure to PowerPoint

image

Figure 2. Individual 162/338 in situ. This figure is available in colour online at wileyonlinelibrary.com/journal/oa.

Download figure to PowerPoint

Methods

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Materials
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

Measurements of the bones of the skull base, the shoulder and the pelvis, and the upper and lower extremities were taken according to Kósa (1978). Measurements of the bones of the neurocranium were not taken because the borders of these bones are usually irregular and the distinction between regular and eroded borders is sometimes not clear. All measurements were recorded with a digital sliding calliper to the nearest 0.01 mm. Because bone size and age at death are strongly correlated in fetal and infant individuals, the measurements were used to estimate age at death by (i) comparing the values with data provided by Kósa (1978) and (ii) using regression equations for long bones given by Scheuer et al. (1980). Because a number of tooth crowns are also preserved, age at death was also estimated by the dentition (Ubelaker, 1989; Hillson, 1996). Additionally, the body length of the individuals was calculated using linear regression equations by Kósa (1978).

Determination of sex from the bones of fetuses, infants and children is difficult, although some morphological differences between the sexes do exist in these groups (Weaver, 1980; Schutkowski, 1987; Scheuer & Black, 2000). However, all sexing methods published thus far are either not fully validated or have low accuracy (Scheuer & Black, 2000; Baker et al., 2005; Lewis, 2007). Therefore, no attempt was made to estimate the sex of the individuals in the present study.

Results

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Materials
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

The skeletons of both individuals are preserved to a large extent (Figure 3), slightly better in individual 162/338 than in individual 162/331, including a left malleus in individual 162/331 and a left incus in individual 162/338. The surfaces of the post-cranial bones exhibit only minor erosion, and although a number of the long bones are fragmented, their state of preservation allows a reliable reconstruction of their actual size.

image

Figure 3. Skeletal remains of the two individuals (left: 162/331; right: 162/338). Vertebrae are largely preserved but not shown. The bar indicates centimetres. This figure is available in colour online at wileyonlinelibrary.com/journal/oa.

Download figure to PowerPoint

The state of preservation of individual 162/331 allows for seven measurements of the skull base, the shoulder and pelvic girdles, and the upper and lower extremities; in individual 162/338, 31 measurements could be taken (Table 1). Except for scapula length, all values are slightly larger in individual 162/331 than in individual 162/338. With the use of these metrical data, most measurements point to an age at death of between 36 and 40 weeks' gestation in both individuals (Table 2). Applying the regression equations by Scheuer et al. (1980) led to more homogenous values ranging from 36.2 to 37.5 weeks' gestation for individual 162/338 and to 39.3 weeks' gestation (only one value available for the set of equations) for individual 162/331.

Table 1. Measurements obtained on the individuals
Skeletal element(s)Individual 162/331Individual 162/338
RightLeftRightLeft
  1. Values are given in millimetres.

Occipital
Width of the basilar part15.0115.0113.9013.90
Length of the basilar part12.2812.2811.5911.59
Length of the lateral part26.02 23.1323.34
Shoulder girdle
Clavicle length  42.1142.88
Scapula length 29.79 30.91
Scapula width   28.10
Spina scapula length   31.06
Pelvic girdle
Ilium length   32.74
Ilium width   29.52
Ischium length  18.0418.33
Ischium width  11.9711.64
Pubis length  16.1215.33
Upper extremities
Humerus length  60.7160.63
Humerus width 17.1216.1216.20
Radius length54.08  50.98
Ulna length   57.63
Lower extremities
Femur length  68.6069.18
Femur width  20.3119.55
Tibia length  60.6061.05
Fibula length 61.1457.9757.93
Table 2. Results of the age at death estimation (gestational age in weeks) based on the osteometric data
Skeletal element(s)Individual 162/331Individual 162/338
RightLeftRightLeft
  1. Values without brackets are readings of the corresponding values adapted from Kósa (1978). Values in brackets are calculations based on the linear regression equations by Scheuer et al. (1980).

Occipital
Width of the basilar part40403838
Length of the basilar part38383737
Length of the lateral part40 3838
Shoulder girdle
Clavicle length  3838
Scapula length 36 36
Scapula width   39
Length of spina scapula   40
Pelvic girdle
Ilium length   38
Ilium width   39
Ischium length  3940
Ischium width  3938
Pubis length  4038
Upper extremities
Humerus length  38 [36.5]38 [36.5]
Humerus width >403939
Radius length>40 [39.3]  40 [37.5]
Ulna length   39 [37.1]
Lower extremities
Femur length  38 [36.2]38 [36.4]
Femur width  >4040
Tibia length  38 [37.0]38 [37.2]
Fibula length 403838

On average, the age at death estimations indicate a slightly higher age at death in individual 162/331 than in individual 162/338. Interestingly, the values for the widths of humerus and femur (in individual 162/338) gave older ages at death than the length of the respective bones, indicating a higher robusticity (lower length to width ratio) of the described individuals compared with the sample investigated by Kósa (1978).

Calculated body length (Table 3) was slightly larger in individual 162/331 (mean of all values: 491 mm) than in individual 162/338 (mean of all values: 448 mm). However, the range of variation of the results was high: depending on the measurement, reconstructed body length varied between 347 and 539 mm in the smaller individual (162/338) and between 428 and 572 mm in the longer one (162/331). In both individuals, reconstructed body length on the basis of radius length (572 mm in individual 162/331 and 539 mm in individual 162/338) markedly exceeds the results obtained from the other bones and therefore probably does not represent the ‘true’ value. On the other hand, body length reconstruction based on measurements of the bones of the pelvic girdle led to markedly lower values. As these elements are missing in individual 162/331, the recorded differences in the estimated age at death and body length are probably overestimated. Most values for body length in both individuals range between 430 and 490 mm, and this range seems to be the most likely one for them. Hence, the results of the body length reconstructions are in concordance with the age at death estimations by the bone measurements, indicating an age at death in the late fetal period for both individuals.

Table 3. Estimations of the body length of the individuals based on the regression formulas of Kósa (1978)
Skeletal element(s)Individual 162/331Individual 162/338
RightLeftRightLeft
  1. Values are given in millimetres.

Occipital
Width of the basilar part471471436436
Length of the basilar part522522492492
Length of the lateral part491 437441
Shoulder girdle
Clavicle length  502511
Scapula length 428 444
Scapula width   471
Length of spina scapula   491
Pelvic girdle
Ilium length   470
Ilium width   451
Ischium length  390396
Ischium width  414403
Pubis length  364347
Upper extremities
Humerus length  459458
Humerus width 488460462
Radius length572  539
Ulna length   474
Lower extremities
Femur length  446450
Femur width  467450
Tibia length  443446
Fibula length 469445444

Eight tooth crowns are preserved in individual 162/331 and five in individual 162/338. In contrast to the bones, the stage of dental development is almost identical in both individuals: the crowns of the first deciduous incisors are approximately two-thirds complete, the crowns of the canines are less than half, and the occlusal surfaces of the first molars are almost complete. These findings are in concordance with a perinatal age at death of the individuals (Ubelaker, 1989; Hillson, 1996) and support the results of the estimations based on skeletal elements.

In both individuals, the lower left first molar exhibits a similar configuration of its disto-buccal cusps (Figure 4), probably a division of the hypoconid. Although the phenomenon of division of the hypoconid has been described repeatedly (Jørgensen, 1956; Keene, 1994), it is an uncommon (epigenetic) trait, which supports the suspicion of closely related individuals.

image

Figure 4. Occlusal views of the crowns of the lower left first molar of individuals 162/331 (left) and 162/338 (right). Note the divided hypoconid (white oval). This figure is available in colour online at wileyonlinelibrary.com/journal/oa.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Materials
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

The recognition of a possible twin burial from an archaeological site is probably highly dependent on whether or not the archaeological context elicits such suspicion. The archaeological context of the two individuals described in the present study, namely the almost identical mode of deposition within the same house pit, raises the suspicion of at least closely related individuals. Investigation of the bones and the teeth of the individuals revealed a similar stage of development and shared morphological traits, thus supporting the suspicion of closely related individuals, probably twins. Further, ‘epigenetic’ traits that are commonly used to reconstruct relationships in adult human skeletal remains (e.g. Buikstra & Ubelaker, 1994) and could be used to confirm the suspicion are either not preserved in the present case or not yet developed in fetal bones. Genetic investigations on such small bones are not promising (Crespo et al., 2011) and were not performed in the present cases.

The actual difference in body size of the two studied individuals is difficult to estimate because measurements on different bones reveal highly variable results and only a small number of these measurements could be taken on both individuals. Body length estimations from the few corresponding skeletal elements differ between approximately 5% and 10%. These differences do neither contradict nor support the diagnosis of a twin burial, because discordances in weight and size at birth are common in twins (Blickstein & Kalish, 2003). Whereas minor discordances represent normal variation without any further relevance, major ones are considered a risk factor for perinatal mortality (Blickstein & Kalish, 2003; Bagchi & Salihu, 2006). In clinical studies, discordances between twins are usually expressed as differences in weight. For instance, a 6.5-fold higher intrauterine mortality rate has been reported in twin pregnancies with weight differences of more than 25% between the two fetuses compared with a group with lower weight differences (Erkkola et al., 1985). However, thus far, birth weight cannot be reconstructed from skeletal material, and therefore, the estimated size difference between the two burials cannot be extended to weight differences.

In the absence of meaningful traces on the bones, the cause of death of the two perinatal individuals remains unknown, as does the attitudes of their relatives and the wider community towards them. It appears, however, that the individuals were buried in an unusual manner. Because a cemetery related to the settlement has so far not been found, no further information is available concerning the ‘usual’ manner of treating deceased neonates in this culture. A deviant treatment of dead (singleton) newborns and early infants has been reported from a number of other Iron Age sites in Germany and neighbouring countries (Jarecki et al., 1999). In many of these cases, the newborns were deposited in pits within the settlement, some of them inside the house (Jarecki et al., 1999). A deposition of deceased neonates inside houses was apparently not frequently practised at the site Emminger Höfe as no other skeletons were found in the numerous excavated house pits. The present case of double burial therefore appears exceptional; however, the reason for this remains unknown. Interpretation of the two individuals as ‘building sacrifices’ seems unwarranted, considering the careless deposition of the bodies and the absence of grave goods.

The burial situation of the individuals described here differs markedly from the situation reported for the Upper Palaeolithic twins from Krems-Wachtberg. These individuals were covered with red ochre and decorated with ornaments, and their grave was covered with a mammoth scapula, which is indicative of a special importance of the individuals to the community (Einwögerer et al., 2006). Although the investigators of the twins from Olèrdola did not discuss the burial context, it seems that they were not treated differently compared with the other neonatal individuals found at the site. They were all buried in ‘handmade installations’ in the settlement (Crespo et al., 2011).

The three cases of presumed twin burials (Krems-Wachtberg, Olèrdola and Ochtendung) thus vary in the way the individuals were buried, which may at least partly reflect differences in the cultural attitude towards twins.

Acknowledgements

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Materials
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

The author gratefully acknowledges Axel von Berg, Head of the Generaldirektion Kulturelles Erbe, Rheinland-Pfalz, Amt Koblenz, for the important information about the archaeology of the site and for the permission to publish Figures 1 and 2. The author also acknowledges Uwe Kierdorf, Carsten Witzel, Maciej Henneberg and the anonymous reviewers for their enriching discussions and comments on the manuscript.

References

  1. Top of page
  2. ABSTRACT
  3. Introduction
  4. Materials
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References