Body Size, Weapon Use, and Natural Selection ... - Wiley Online Library

19 downloads 87 Views 1015KB Size Report
knees of male children to prepare them better for the rigors of kayak hunting ( ...... Ober die Verhaltnisse der Warmeokonomie der Thiere zu ihrer Griisse.
Body Size, Weapon Use, and Natural Selection in the European Upper Paleolithic and Mesolithic DAVID W . FRAYER University of Kansas

Evidence f o r a relationship between hunting strategies and body size is examined f o r human skeletons dating t o the European Upper Paleolithic and Mesolithic. Trends for reduced limb size and statute seem to be correlated with improvements in the types of weapons utilized and a shgt f r o m aggressive t o more docile game. Although some of these obsenxations fit the predictions of Brues concerning the spearman-archer model, it is suggested that selection f o r reduced metabolic demands u a more plausible explanation f o r decrease in body size f r o m the Upper Paleolithic t o the Mesolithic. [Upper Paleolithic, Mesolithic, hunting, body size]

IN A PAPER APPEARING I N THIS JOURNAL over 20 years ago, Brues (1959)presented a model describing the relationship between body build and offensive weapons. Through time, Brues’s model of the spearman and the archer has become a commonly accepted axiom in the anthropological literature, if one can label an axiom by the frequency and regularity of its citation in introductory textbooks. Yet, the spearman-archer model has never really been tested against paleontological or osteological data. There has been sporadic reference to “atlatl elbow” (Angel 1966;Ortner 1968) and to other morphological correlates of specific hunting patterns (see Brace and Montagu 1977:362),but the major features of Brues’s hypothesis have never been subjected to data drawn from populations which are lineally related to each other, yet which follow different technological means in hunting pursuits. In this paper, I will present evidence drawn from European Upper Paleolithic and Mesolithic groups in an attempt to test the basic contentions of the Brues spearman and archer model.

DAVID W. FRAYER. m i a t e profnmr at ihe University of Kansas. received hh Ph D. from the Univenity of Michigan in 1976. His main area of research U human palcontology especially metric and morphological trenda in the later aspecu of human evolution and their possible rdatioruhip to techno.cultural evolution. Other papen on evolutionary trends in the Upper Paleolithic and Mesolithic have appeared in the Amenron J o u m d of Physical Anfhropology. Journal of Human Evolution. and J o u m l of Denfol Research Frayer is rho the author of The Ewlution offhc Denfition in Upper Poleolirhtr ond Mrro lifhir Europe (1978) published by the University of Kansas.

Copwight G ’ 1981 by the American Anthropolog~cal Association 0002-7294/81/0100571792 2011

57

58

AMERICAN A N THR OPOL OCIST

[8S, 1981

THE MODEL

The central thesis of the spearman and archer model revolves around the morphological requisites of certain types of offensive weapons. Brues suggests that there is a “relation of body build to the efficiency with which different activities can be carried out” (Brues 1959:457), so that physique may be correlated with the better ability to utilize specific types of hunting armaments. Brues argues that “laterality” and “linearity” in physical build are important variables which are linked to specific types of weapons used in the hunt. For example: . . . the two principal weapons used by hunting peoples for thousands of years, the spear and the bow, demand different types of arm leverage. The range and effectiveness of a spear depend on the speed with which it is moving when it leaves the hand, and therefore on the maximum velocity the hand itself has attained at that moment. The range and effectivenessof the bow depend on the stored energy of the drawn bow,which is proportional to the pounds of pull the arm can exert on the bowstring. Obviously, people of linear build and long limbs will operate more efficiently with a spear, and people with lateral build and short limbs. with a bow [Brues 1977:171].

Following Brues’s model, then, one would predict that hunters relying upon the spear and spearthrower as the primary hunting weapons would have relatively longer arms and an elevated stature compared to groups using the bow exclusively in hunting pursuits. Presumably, linear hunters could generate more leverage and speed in throwing a spear because of proportionally longer arms, allowing for greater velocity. On the other hand, the most effective use of the bow involves maximum pulling force on the bowstring. Since the possible length of draw is limited by the individual’s arm length, best results are obtained if the stiffness of the bow is adapted to the individual archer’s pulling ability. . . The archer requires a power leverage in the arm, which is favored by short limb segments and relatively short and thick muscles [Brues 1959:465].

.

Besides leverage and other mechanical advantages of particular physiques in specific weapon-using capacities, it is necessary to consider other factors. In particular, size and aggressivenessof the prey and the effective killing distance between the prey and predator would seem to be important variables connected to overall body size in hunter-gatherer groups. It is reasonable to suggest that the farther a weapon allows a hunter to be from his quarry, the less danger involved. Also, the more accurate the missile, the better the chances the hunter has in strategically wounding the animal, thereby escaping direct contact with the prey. Hunting with a spear demands rather close proximity to the animal. With spear hunting, separation of the prey and predator is a direct function of the effective distance that the spear can be propelled, providing maximum accuracy and sufficient penetration to kill or immobilize the animal. Tipping spears with stone points contributes to the penetrating ability of the projectile (Semenov 1964) but probably has little effect on the necessity of close approach by spear hunters. Given the necessity of close approach by these hunters, the improbability the quarry would be killed instantly, and the general tendency of animals to retaliate when wounded, spear hunting throughout the Paleolithic was undoubtedly a very dangerous business. With the development of the spearthrower, however, an important change occurred in the effective killing distance between the hunter and the hunted. Experimental results indicate that the effective distance of the atlatl and dart is about 18-27 m. (Spencer 1974); accuracy and penetration are rather poor at greater distances. It is apparent that the extension of the arm provided by the spearthrower does confer a stronger propulsive force, and its use by a hunter affords a greater killing force and an increase in the distance between himself and the prey. Even though the spearthrower is a significant addition to the hunting arsenal, it is still likely that hunters using spearthrowers shot from

Fra yer]

BODY SIZE A N D WEAPONS

59

ambush or at close range into herds of animals (Coles 1973:127) and that hunters were still subject to considerable danger. Utilization of the bow as a primary killing tool signals a substantial improvement over the spearthrower. The greatest advantage relates to the increase in critical distance between prey and predator. Studies by Pope in the early part of this century on aboriginal and historic bows show that bows are more accurate than spearthrowers, can be used at a greater distance (effective at 91 m.), and have greater depth of penetration (Pope 1962). Pope reports killing a running buck deer at a distance of 75 m. with a single arrow, and a charging bear (actually killed with a rifle) which “received and resisted 5 arrows, one of which passed through her abdomen and flew out 10 metres” behind her (Coles 1973:126). Although there is a great variation in effectiveness of the different types of aboriginal bows tested by Pope, in most cases they were effective at distances greater than 47 m. Bow hunters, then, have a significant advantage over spear hunters in that they are more removed from the immediate proximity of the prey. Just as the rifle is a more effective device than the bow (particularly when killing a charging she-bear), the bow provides an advantage over the spearthrower and dart, and the spearthrower over the hand-held spear. From these considerations, one might predict the following relationships. Groups dependent upon the hand-held spear in hunting would show a large stature, proportionally long arms, and, depending on the type of the quany, a substantial amount of skeletal robusticity. The greater the body size and the aggressiveness of the game being hunted, the greater the skeletal robusticity in the hunters. Holding the type of game constant, utilization of the atlatl might be associated with a decrease in the amount of skeletal robusticity since the effective killing distance is increased and the relative chances of direct contact with wounded game are decreased. Hunting with bow should produce further reductions in overall body size as a response to the morphological requisites of effective weapon use. Bow hunters should be considerably less robust than spear hunters and, following Brues’s model, would have a lower stature and proportionally shorter arms. Morphological attributes, then, may have some intimate connection with the level of weapon sophistication. This is not to say that behavioral patterns are less important, for as Laughlin (1968) so effectively argues, the competitive edge of hunters is probably more dependent on hunting lore and ethological knowledge than on the specific weapons utilized. Nevertheless, it is important to note that Aleut sea hunters consciously made through stretching exercises morphological modifications in the shoulder, low back, and knees of male children to prepare them better for the rigors of kayak hunting (Laughlin 1968:306-307). If human beings purposefully modify morphology to produce more effective hunting behavior, it is not unlikely that natural selection would also have input concerning the most fit phenotype given specific hunting tactics and techniques.

THE TEST CASE Upper Paleolithic and Mesolithic populations of western, central, and eastern Europe provide an adequate test of this model. Human populations from these periods seem to be lineally related to each other through most, if not all, of the sequence, so that variation in morphology and physique is probably best regarded as the result of a combination of developmental and selective forces operating on successive populations through time. Although there are some archaeological discontinuities within the Upper Paleolithic, for the bulk of the period, and particularly for the transition between Upper Paleolithic and Mesolithic traditions, a considerable body of evidence exists for phyletic evolution of genetically and culturally linked forms (Clarke 1976; Cohen 1977; Czarnik 1976; Ferem-

60

AMERICAN ANTHROPOLOGIST

[8Y. 1981

bach 1978; Frayer 1978). Overlying evidence for continuous evolution from the Upper Paleolithic to the Mesolithic is a well-established dichotomy in hunting patterns. Hunting in the Upper Paleolithic was accomplished primarily through the spear and spearthrower (Bordes 1968; Bouchud 1976) and was concentrated on megafauna, primarily reindeer, mammoth, rhinoceros, horse, bison, cave bear, elk, and auroch Uellnek 1972). In the early part of the Upper Paleolithic (until the Solutrean) hunting was done primarily with hand-held spears. In Europe the spearthrower has an antiquity dating at least to the Solutrean (Mellars 1973:259), if not slightly earlier. Judging from the lithic and bone artifacts found after the Solutrean, a considerable part of the hunt in the later Upper Paleolithic was done through the use of atlatl and dart. Development of the harpoon in late Magdalenian groups provides further evidence for improvements of spear-hunting technology in the late Pleistocene (Collins 1976:114-116). Earliest use of the bow and arrow comes from either the terminal phase of the Magdalenian or what more appropriately may be labeled early Mesolithic (Clark 1967:93). Discovery of more than 100 pine shaft arrows at Stellmoor in northern Cermany (Rust 1943) provides conclusive evidence of bow and arrow utilization by around 10.000 B.P. Microlithic points found in late Magdalenian sites in western Europe (Sonneville-Bordes 1960) suggest these groups may have also had the bow as part of their offensive arsenal. Even though the bow probably developed in late Upper Paleolithic contexts, it appears that it was not until the Mesolithic that the bow became the predominant weapon used in the hunt (Ode11 1978). The high frequency of snapped blade tools in virtually all Mesolithic assemblages indicates that most projectiles were tipped with small stone points and propelled from bows. Along with the innovation of the bow among Mesolithic groups, there is a shift in the types of game hunted with the onset of the Holocene. With the decline of tundra conditions and replacement of deciduous forests across most of Europe, animal populations changed accordingly (Meiklejohn 1978). Gregarious megafauna became extinct or were replaced by smaller, more solitary, and probably less aggressive game. Larger mammals most frequently hunted in the early Holocene include red and roe deer, boar, elk, and auroch, each substantially smaller than the prey of late Pleistocene hunters. Furthermore, Mesolithic hunting was accomplished by more “distant” techniques. From evidence compiled by Noe-Nygaard (1974) concerning wound patterns on fauna from Danish Mesolithic sites, most of the quarry in her sample were initially brought down with the bow and arrow, and then spear-killed after the animal was immobilized. Although it is risky to generalize from her data to all European Mesolithic sites, given the common tool types and overall similarity in production of lithic forms found in the European Mesolithic, the bow seems to have been the primary hunting weapon in early Holocene groups. Based on the information from the archaeological record, then, it is apparent that hunting patterns differ substantially between the Upper Paleolithic and Mesolithic. Upper Paleolithic hunting was concentrated on megafauna which was no doubt killed at fairly close quarters by hunters armed with the spear and spearthrower. Although i t is impossible to rule out the importance of jumpsites, traps, poisons, and other more indirect hunting methods, the bulk of hunting during this period was most likely accomplished through rather close contact between the prey and human predator. The apparent absence of the spearthrower in Aurignacian and Perigordian groups suggests that these hunters killed most of their game at short distances using hand-held spears, while from the Solutrean to the Magdalenian, utilization of the spearthrower increased to some extent the distance between the hunter and the hunted. By the late Paleolithic, but particularly within the Mesolithic, the effective killing distance was increased substantially with the adoption of full-time bow hunting. If one measure of weaponry sophistication is reduction in the amount of peril associated with the hunt, it can be concluded that there is a general trend in the late

Frayer]

BODY SIZE A N D WEAPONS

61

Pleistocene and early Holocene for increased hunting efficiency. Successive innovations of the atlatl and bow provided these hunters a greater effective killing distance and less chance of direct contact with prey. Since it is generally assumed the Upper Paleolithic groups are broadly ancestral to subsequent Mesolithic populations and since these changes in the arsenals of human groups and the types of animals hunted are well documented, comparisons of body size changes between Upper Paleolithic and Mesolithic males should provide an appropriate test of a relationship between hunting and human physique. MATERIALS AND METHODS

Data for postcranial dimensions derive primarily from the literature, although a few of the specimens used in the analysis were measured by the author (see Table I). Comparisons between my own measurements and those from the literature indicate the existence of only minor discrepancies attributable to inter-observer error. However, a number of data come from reports written before the standardization of body metrics by Martin (1928). so that it is impossible, given the present evidence, to determine how closely these conform to contemporary measurement techniques. This consideration applies particularly to Verneau’s (1906) measurements on the Grimaldi series. Furthermore, materials from some of the sites are no longer available for study, so that original measurements must suffice. For example, since all the Piedmost material was destroyed at the end of World War 11. data given by Matiegka (1938) represent our only source on the postcranial remains of this extremely important Upper Paleolithic sample. Data for the Mesolithic sample are based on the recent study by Vallois (1977). on other measurements culled from literature, or those taken by myself. Due to the recency of most of these reports, it is reasonable to assume that these data were taken according to standardized procedures. Whether the Upper Paleolithic or Mesolithic sample is considered, it is important to note that the data are undoubtedly subject to inter-observer error, but since there is no adequate means of determining the level of error and since in reality the measurements correspond to rather well-established and easily identifiable landmarks, it is likely that the measurements used here are good estimates of the true morphological state. Determination of sex and the accuracy of the sex assessment for each specimen are critical factors in this analysis. For the majority of the Upper Paleolithic and Mesolithic specimens, sex was determined on the basis of postcranial morphology and cross-checked wherever possible with associated cranial remains. Since in most cases, long bones are found with pelvic remains, the assigned sexes used in this analysis are probably very accurate, In some cases, where the skeleton was not studied by the author, the sex assessment given in the literature was used, but only after it was verified through analysis of the textual descriptions and photographs provided in the original report. The fact that there are more males than females in both Upper Paleolithic and Mesolithic samples is probably not related to skeletal sexing bias discussed by Weiss (1972), but is a consequence of preferential burial patterns of males in both groups. Throughout the analysis percent dqference is used as an index of the amount of change cccurring in postcranial dimensions and stature. Since the means of males and females differ between the Upper Paleolithic and Mesolithic, this ind-x allows for valid comparisons as to the relative amount of change between the samples. The index is calculated by the following formula:

- MX 100 UPX where UPX = the Upper Paleolithic male (or female) mean and M x = the Mesolithic male (or female) mean. percent difference = upX

62

AMERICAN ANTHROPOLOGIST

[83, 1981

TABLE 1. POSTCRANIAL DATASET FOR UPPER PALEOLITHIC AND MESOLITHIC SPECIMENS a

Upper Paleolithic Males Baousso da Torre 1 (F) Verneau 1906 Cap Blanc Baousso da Torre 2 (F) Verneau 1906 Cro-Magnon 2 Barma Grande 1 Barma Grande 2 Barma Grande 5 Caviglione Chancelade Cheddar Combe-Capelle Cro-Magnon 1 Cro-Magnon 3 Grotte des Enfants 4 Oberkassel 1 Paviland Pavlov 1 PFedmost 1 Piedmost 3 Piedmost 9 Piedmost 14 Veyrier 1

(F) (F) (F) (F) (F) (F) (F)

Verneau 1906 Verneau 1906 Massari 1958 Verneau 1906 Vallois 1941-46 DWF Klaatsch and Hauser 1909 (Fi) Vallois and Billy 1965 (T) Vallois and Billy 1965 (F) Verneau 1906 (F) DWF (F) Wells 1963 (H) Vlcek 1961 (F) Matiegka 1938 (F) Matiegka 1938 (F) Matiegka 1938 (F) Matiegka 1938 (F) DWF

Grotte des Enfants 5 Oberkassel 2 Abri Pataud 22 Piedmost 4 Piedmost 5 Piedmost 10

Females (F) Bonin 1935 (H) Vallois and Billy 1965 (F) Verneau 1906 (H) DWF (H)Billy 1975 (F) Matiegka 1938 (F) Matiegka 1938 (F) Matiegka 1938

St. Germain-laRiviere (F) Vallois 1972

Mesolithic Culoz 1 Culoz 2 Fat’ma Koba 1 Gramat 1 Hiiedic 2 Hiiedic 5 Hijedic 6 Hbedic 9 Maritza 1 McKay Cave 217 Montardit Muge Murzak-Koba 2 Parabita 10 Parabita 20 Rastel Rochereil St. Rabier San Teodoro 1 San Teodoro 4 TCviec 2 Ttviec 4 Tbviec 7

Males (S) Vallois 1977 (S) Vallois 1977 (F) Debetz 1936 (F) Vallois 1944 (F) Vallois 1977 (F) Vallois 1977 (F) Vallois 1977 (F) Vallois 1977 (U) DWF (T) DWF (F) Sawtell 1931 (+) Ferembach 1974b (F) Zirov 1940 (F) DWF (F) DWF (F) Barral and Primard 1962 (F) Ferembach 1974a (S) Vallois 1977 (F) Wells 1963 (F) Wells 1963 (F) Vallois 1977 (F) Vallois 1977 (F) Vallois 1977

Birsmatten 1 Cheix Hoedic 4 Hoedic 7 Hbedic 8 Hiiedic 10 Muge Murzak-Koba 1 TCviec 1 Teviec 3 TCviec 6 TCviec 9 TCviec TCviec TCviec TCviec

10 14 15 18

Females (F) Bay 1964 (S) Vallois 1977 (F) Vallois 1977 (F) Vallois 1977 (F) Vallois 1977 (F) Vallois 1977 (+) Ferembach 1974b (F) hrov 1940 (F) Vallois 1977 (F) Vallois 1977 (F) Vallois 1977 (F) Vallois 1977 (F) (F) (F) (F)

Vallois Vallois Vallois Vallois

1977 1977 1977 1977

Fra yer]

BODY SIZE AND WEAPONS

63

Table I continued Mesolithic Males Teviec Teviec Teviec Teviec

8 11 IS 16

(F) (F) (F) (F)

Females Vallois 1977 Vallois 1977 Vallois 1977 Vallois 1977

a Data from source listed; DWF from unpublished measurements by the author; bone used to estimate stature is also given. (F-maximum femur length; H-maximum humerus length; T-maximum tibia length; U-maximum ulna length; Fi-maximum fibula length; S-stature given by author; * site averages only).

MORPHOLOGICAL TRENDS Following t h e predictions of the Brues hypothesis (1959) and the corollary additions concerning the importance of the types of game hunted and the techniques for killing them, it is reasonable t o predict that males i n the Upper Paleolithic would have longer arms and legs and a greater stature in comparison to males from the Mesolithic. Since t h e

TABLE 11. UPPER PALEOLITHIC A N D MESOLITHIC BODY MEASUREMENTS A N D STATURE ESTIMATES FOR COMPLETE SAMPLE."

Upper Paleolithic n s.d.

x Maximum humerus length males females Maximum radius length males females Maximum ulna length males females Maximum femur length males females Maximum tibia length males females Staturee males females

-

X

Mesolithic n s.d.

Percent difference

342 308

14 8

24 14

312 285

11 9

30 17

8.8C 7.5c

265 237

10 5

18 12

24 1 218

10

22 20

9.lb 8.0

282 255

13 6

18 14

263 242

7

19 17

6.7b 5.1

47 1 422

17 5

40 10

435 404

13

28 20

7.6' 4.3

406 354

13 5

23 13

362 334

16 11

28 15

10.8d 5.6b

1743 1593

20 9

94 41

1648 1539

26 15

66 48

5.5d

7 9

16

3.4'

All measurements in millimeters. Significant at .05 level with two-tailed student's 1 . Significant at .01 level with two.tailed student's t . Significant at .001 level with two-tailed student's t . Stature is calculated from data presented in Trotter and Gleser (1952) using white male and fcrnale formulas. a

64

AMERICAN ANTHROPOLOGIST

[83, 1981

TABLE 111. UPPER PALEOLITHIC AND MFSOLITHIC BODY MEASUREMENTS AND STATURE FSTIMATES FOR GROUPED DATA.^

Upper Paleolithic n s.d.

x Maximum humerus length males females Maximum radius length males females Maximum ulna length males females Maximum femur length males females Maximum tibia length males females Staturee males females

x

Mesolithic n s.d.

Percent difference 7.6b

341 305

6

26 14

315 284

9 5

27 28

6.9

265 234

8 4

19 11

246 225

8 4

18 14

7.2 3.8

282 250

0 4

20 11

266 245

8 5

12 13

5.7 2.0

466 422

11 4

38 11

446 413

11 5

29 24

4.3 2.1

406 352

9 4

22 14

369 335

9 4

22 17

9. l d 4.8

1737 1591

13

85 20

1678 1553

15 6

63 57

3.4 2.4

1

7

a All measurements in millimeters. Significant at .05 level with two-tailed student’s t . Significant at .01 level with two-tailed student’s t . Significant at .001 level with two-tailed student’s t . Stature is calculated from data presented in Trotter and Gleser (1952) using white male and female formulas.

technological factors related to these hypothesized changes primarily involve male economic pursuits, females over the same two periods would be expected to show fewer changes in overall body size measures. Data to support this position are given in Tables I1 and 111. Table I1 presents data for the complete sample drawn from the Upper Paleolithic and Mesolithic. In this sample, relevant measurements are included for all individuals whether they derive from the same site (or burial within the same site) or from different sites. Because this sampling procedure may bias the resulting statistics toward sites with a greater number of individuals contributing to the overall mean (e.g., Piedmost in the Upper Paleolithic and Hikdic and TCviec in the Mesolithic), the data were regrouped into another sample (Table 111) where only site averages were used.’ Here, all sites contribute equally to the overall mean for the specific skeletal dimension. Inspection of Tables I1 and 111 indicates that there is very little difference between the two sampling techniques with respect to long-term trends for reduction in body size. Since Table I1 contains the larger sample sizes, statistical results reported in it will be referred to exclusively in the following analysis. Trends for reduction in body size among Upper Paleolithic males and females are remarkably consistent with the predictions concerning the relationship between hunting techniques, the animals hunted, and human physique. Both sexes show substantial reduction in all limb segment lengths, but males consistently exhibit a greater percent reduction for each length dimension. In the upper limb, the humerus reduces 8.8 percent in males and 7.5 percent in females; the radius 9.1 percent and 8.0 percent in males and females respectively; and the ulna 6.7 percent and 5.1 percent in the two sexes. For the lower limb, the femur reduces 7.6 percent in males and 4.3 percent in females and the

BODY SIZE A N D WEAPONS

Frayer]

65

tibia 10.8 percent and 5.6 percent in males and females respectively. In addition, reduction in the mean dimensions is more often of statistical significance in males than in females. For all five maximum length measures, male means show a statistically significant reduction (p < .05) using a two-tailed t test. Trends in overall body size produce a similar picture. Male stature reduces 5.5 percent, while females show a 3.4 percent reduction in stature. Just as in the lengths of the individual postcranial bones, then, males over the Upper Paleolithic and Mesolithic demonstrate a greater (absolute and percent) reduction in body height, following the predictions outlined above for an interrelationship between hunting modes and body size. Females, whose economic roles and subsistence equipment appear to be more stable through time, are characterized by more conservative evolutionary change, yet still in the same direction as males. These reduction trends in body size and arm and leg segments suggest that different intensities of selection and/or developmental factors were operating on males and fernales in these late Pleistocene/early Holocene groups. Figures for body size trends within the Upper Paleolithic are presented in Table IV. Because sample sizes for the individual arm and leg bones are too small to break the group into separate periods, only stature estimates are given. From the table i t is apparent that Upper Paleolithic males are remarkably stable over the two periods.* The difference in average stature between the early and late Upper Paleolithic males is less than 3 mm. Females on the other hand show a trend for slight reduction in overall size, although when the female Upper Paleolithic sample is broken into two divisions, sample size is extremely small. These data may indicate that females during the Upper Paleolithic are gradually decreasing in stature, but a larger sample is needed to be certain. Data for males, however, clearly demonstrate that they are maintaining a high stature throughout the time span and that only with the onset of Mesolithic times does stature reduce significantly. Changes in arm segment proportions, arm size relative to stature, and body robusticity are also relevant here, especially concerning the merits of the Brues hypothesis. Unfortunately, I have only been able to collect specific data on the first two considerations. There are virtually no conclusive data concerning changes in robusticity, although several authors (Weidenreich 1945; Vallois 1977) have discussed trends for “gracilization” in Holocene human populations. In this area considerable research is needed. Data reviewing changes in arm segments relative to each other and to stature are presented in Table V. The first two indices relate the bones of the forearm to the upper arm. In neither the radius nor the ulna is there a marked change in their proportions to the humerus over time. For the radius the change between Upper Paleolithic and Mesolithic males and females is less than 1 percent. Relative ulna size is also rather stable through time, showing less than a 2 percent change in males and about a 3.5 percent change in females. In either case the trend is for relatively longer forearms (or conversely, shorter upper arms), but the differences are not statistically significant, nor particularly TABLE IV. STATURE ESTIMATES FOR EARLY UPPER PALEOLITHIC (EUP). LATE UPPER PALEOLITHIC (LUP). AND MESOLITHIC (MESO) MALES AND F E M A L E S . ~

EUP Stature males females a All

LUP

MESO

X

n

s.d.

X

n

S.d.

x

n

s.d.

1742 1613

10 5

101 31

1744 1567

10 4

92 41

1648 1539

26 15

66 48

measurements in millimeters.

,769

,764

,763 .766

5 7 .65%

10 12 .39%

,023 ,025

,018 ,025

,822 ,851

,837 ,852

a Significant at .01 level with two-tailed student's t .

Upper Paleolithic females Mesolithic females Change in means

Upper Paleolithic males Mesolithic males Change in means

6 7 3.5370

13 9 1.79%

,035 ,109

,027 ,025

max. ulna I t . max. humerus It: X n s.d.

max. radius It.

max. humerus It. X n s.d.

.149 .141

.150 ,145

X s.d.

It,

4 8 5.3770

,026 ,008

11 .012 12 ,005 3.33%a

stature n

max, radius

,160 ,156

,163 ,159

X

-

6 8 2.50%

14 10 2.45y0

stature n

.009 ,006

,124 ,005

s.d.

max. ulna I t .

,193 ,184

,196 ,187

x

s.d.

8 ,005 10 ,006 4.66%a

15 .029 13 ,009 4.59%a

stature n

rnax. humerus It.

TABLE V . RELATIVE ARM MWSUREMENTS FOR UPPER PALEOLITHIC AND MESOLITHIC MALES A N D FEMALES.

m

W

Y

2

0

20

0 t l

b

52

6

Frayer]

BODY SIZE A N D WEAPONS

67

striking. These data only weakly support the Brues hypothesis (1959) in that the absolute shortening of the arm is accompanied by a decreased humerus length relative to the ulna or radius, but the fact that the changes are so small and that a greater change occurs in the females lends little support to a relationship between specific weapon type and arm size. When the maximum lengths of the bones of the arm are compared to stature, it is apparent that the size of the radius, ulna, and humerus are relatively smaller in the male and female Mesolithic samples. However, females show a greater percent reduction in relative lengths of all three arm bones which is the opposite one would expect if changes in arm dimensions are conforming to Brues’s predictions. Consequently, from the data on relative arm size, there is little to support a relationship between arm proportions and the use of either spears or bows in the Upper Paleolithic and Mesolithic. DISCUSSION

The close correlation between body size and technological/hunting modes fits remarkably well some of the predictions outlined in the first section. It is possible to interpret these results in an evolutionary model where body size may confer an adaptive advantage with respect to particular weapon types and kinds of game exploited. Just as certain occupations in modem society have characteristic morphotypes (Harrison, Weiner, Tanner, and Barnicot 1977:381-384), specific behavior related to the efficiency of weapons or hunting patterns may select for the most proficient skeletal and muscular arrangements. To some extent these data can be corroborated with changes occurring in physique in other areas of the world. For example, Brace (1980:147) has noted that prehistoric Australian aborigine samples show a heightened degree of skeletal robusticity, greater sexual dimorphism, and increased tooth size when compared to modem native inhabitants of Australia. He suggests that prehistoric Australian Aborigines were hunting mega-marsupials and argues that “the physical demands placed upon the hunters of rhinoceros-sized Diptotodon and others of the now-extinct megafauna were greater than those required of hunters of the emu and the grey kangaroo” (p. 147). Since little climatic or technological change has occurred between the late Pleistocene and modem human cultures of Australia, this hypothesized connection between size of prey and size of predator seems a likely explanation from the changes observed in Australia. However, a more detailed study of the Australian sample is needed to verify this conclusion. Other factors accounting for the reduction of body size cannot be ruled out. For example, climatic factors may have had an important role in affecting the morphological trends outlined for the Upper Paleolithic and Mesolithic. Despite the fact that there were periodic cold fluctuations in the early part of the Mesolithic, for the bulk of this period temperatures were much warmer than found throughout most of the Upper Paleolithic. As a consequence of this climatic difference, it is likely that physical morphology during the Upper Paleolithic was more subject to cold stress than during the Mesolithic. In general, mammals inhabiting colder climates have more compact bodies than those inhabiting warmer climates (Bergmann 1847). In human populations, Roberts’s work (1952) has demonstrated an association between colder temperatures and increased weight, but a weak correlation with stature. For the limbs, following Allen’s law, most mammals inhabiting polar areas have shorter appendages - a conclusion also verified by Roberts (1973) for humans. Applying these results to the Upper Paleolithic-Mesolithic data leads one to the conclusion that climatic adaptation is of minimal importance for explaining trends in body size through time. Upper Paleolithic body dimensions reduce to the Mesolithic, indicating that cold adaptation is not primarily involved. If this were

68

AMERICAN .4NTHROPOLOGIST

[83, 1981

the case, arm and leg lengths, as well as stature, might be expected to increase as the climate shifted from cold to more temperate conditions. Nutritional differences and diet are also known to affect body size in extant populations and likely had an effect on the hunters of the Upper Paleolithic and Mesolithic. A plethora of studies have shown a relationship between adequacy of food intake and ultimate adult body size (Harrison et al. 1977; Malina 1979), and it is generally assumed, although still not conclusively demonstrated (Meredith 1976). that better nutritional standards are the cause for the present secular trend in stature. That some doubts still remain concerning the mechanism behind the increase in stature is well indicated by a recent study which concludes that “we still do not know why children grow quicker and reach a final bigger size than in the past” (Olivier, Chamla. DeVigne, Jacquard, and Iagolnitzer 1977:343). Whatever the mechanism accounting for the environmental effect on stature, enough studies have been done to support the general assertion that males tend to show more plasticity in certain morphological traits than females. For example, Stini (1979) has demonstrated the pronounced upper arm muscle reduction in males suffering protein-calorie malnutrition, compared to a greater stability in females under the same nutritional stress. In a later article, he stressed the selective importance of stability in females as a consequence of the physiological demands associated with childbearing and nurturance (Stini 1975). Under conditions of nutritional deprivation it is apparently advantageous to reduce body size in males thereby decreasing the amount of energy intake necessary for survival, whereas much less of an advantage is afforded females under the same conditions. Irrespective of poor or adequate nutrition, the cost of large body size in either sex is of major consequence. According to Stini (1976): . . . a 70 kg. man walking moderately fast for two hours will require 1 1 more grams of protein or carbohydrate than will a 60 kg. man doing the same thing. Even while sleeping, the heavier man will bum off the equivalent of 18 grams more protein or carbohydrate during the night. Calculated over an entire day’s round of activities, the difference is in excess of 100 grams, and, over a year, the difference is about 37 kg. This may not seem like much, but when we consider the aggregate effect over a population, the effect is substantial.

Clearly, then, an important benefit for the possession of large body size must be operating in Upper Paleolithic populations to compensate for the maintenance of a large body size in the face of heightened nutritional costs. As discussed above, hunting modes may provide the answer. A further concern involving nutritional factors relates to differential dietary stress in Upper Paleolithic and Mesolithic groups. If, in fact, the morphological trends observed are the result of nutritional factors, it is necessary to demonstrate that Mesolithic groups were subject to more stringent or more unpredictable protein-calorie resources than found in the Upper Paleolithic. Although Upper Paleolithic groups of Europe have been traditionally characterized as inhabiting a veritable paradise loaded with game, a number of recent analyses have conjectured otherwise. Work by Sturdy (1972, 1975) and Burch (1972) indicate that in all likelihood reindeer were not the stable resource they are often assumed to be. Upper Paleolithic hunters could not have moved with the herds but must have intercepted reindeer during the herd’s yearly migrations. Hence, as Bahn (1977) suggests, Upper Paleolithic groups were probably not sedentary or even semisedentary, but must have followed seasonal rounds, covering different geographic parts of Europe. Moreover, reindeer probably did not follow the same migration routes every year, and consequently were probably more undependable than dependable. Also they were not found in the immense herds so often mentioned in the literature (Sturdy 1972). Like the Nunamiut caribou hunters described by Binford and Chasko (1976). Upper Paleolithic populations very likely suffered periods of starvation. Specialization on a

Fra yer]

BODY SIZE A N D WEAPONS

69

narrow-based food source in the Upper Paleolithic does not continue into the Mesolithic. These groups tended toward a more sedentary, broad-spectrum exploitation where a wide variety of different faunal and floral species were utilized (Clark 1952; Clarke 1976; Cohen 1977; Koztowski 1973; Meiklejohn 1978). Given their lack of specialization on an unpredictable food resource and a broad-based exploitation of a geographically limited niche, Mesolithic groups were probably much less subject to nutritional stress. Consequently, trends reported here for reduction in body size from the Upper Paleolithic to Mesolithic would not seem to be influenced significantly by nutritional stress. The group most likely under the greater nutritional stress has the larger body size. Nevertheless, nutritional factors may be the important pivotal effect of the reduction of body size in the Mesolithic. I suspect that large body size in the Upper Paleolithic was maintained as a function of the types of hunting weapons and size and aggressiveness of the quarry. Spear hunting demanded strength, power, and overall robusticity; each correlated with the hunters' ability to withstand the rigors of close-quarter killing. Large body size in Upper Paleolithic males, then, directly related to their survivorship, and, indirectly, to their potential fertility. With the onset of the post-Pleistocene climate, hunting technology and the types of fauna exploited show a significant change, so that large body size was of less selective importance in the hunt. Strictly adhering to the Brues hypothesis, one would have to argue that body size reduced as a function of more lateral males' greater effectiveness in utilizing the bow. However, given the data presented earlier, this seems an unlikely selective factor for producing such a change. Rather, I suggest that since a smaller body size has a lower metabolic demand and since personal danger was reduced with the adoption of the bow and arrow, natural selection operated to reduce body size in Mesolithic populations as a form of nutritional conservation. The additional factor of more effective bow use in hunters of a lateral physique, if indeed this is true, may have been more of a consequence than a cause of the reduction in body size. In summary, the basic predictions of Brues concerning the relationship between body build and weapon type are not supported by data drawn from European Upper Paleolithic and Mesolithic populations. Body size reduces over these periods, with a greater absolute decrease characterizing males. Yet, proportions of the limb segments show no significant modification. Consequently, it is unlikely that body size changes are related only t o the adoption and use of weapon types. Whereas climatic adaptation and malnutrition may be ruled out as important factors, the reduced metabolic demand of smaller individuals in the Mesolithic, coupled with no selective pressure for maintaining robust body size in these bow hunters, may be the critical factor accounting for decrease in overall body size from the Upper Paleolithic to the Mesolithic.

NOTES Acknowledgments. Some of the research on which this paper is based was supported by NSF Grant # GS-38067and University of Kansas General Research Grant # 3234. I would like to thank Dr. George J . Armelagos (University of Massachusetts, Amherst) for some initial prodding to do the study and Drs. Anta Montet.White and David E. Willer (University of Kansas) for reviewing the manuscript before its submission. The numerous reviewers of the original and revised versions of this paper also made substantial improvements. Site averages were calculated in the following manner. When more than one individual of the same sex derived from the same site, the specific long bone measurements were summed and divid. ed by the number of individuals. For example, there are three humerus measurements for females at Piedmost. These three dimensions (for Predmost 4. 5 , and 10) were summed and divided by three. Since all other Upper Paleolithic females are from different sites, sample size in Table I11 in. cludes a total of six specimens, all from different locations. The manner in which the data were

'

AMERICAN ANTHROPOLOGIST

70

[83, 1981

assembled in Table 111, then, eliminates any bias resulting from the overrepresentation of sites containing more than one individual. * Early and late Upper Paleolithic samples are divided according to techno-cultural assemblages. Early Upper Paleolithic (EUP) includes material dating to Chatelperronian, Aurignacian, or Early Gravettian periods. Late Upper Paleolithic (LUP) is confined to specimens associated with Upper Perigordian, Solutrean, Magdalenian, or late Gravettian tool assemblages. See Frayer (1978)for further discussion of the rationale of this procedure. REFERENCES CITED Angel, J. L. 1966 Early Skeletons from Tranquility, California. Smithsonian Contributions to Anthropology

2(1):1-19. Bahn, P. G . 1977 Seasonal Migration in South-west France During the Late Glacial Period. Journal of Archaeological Science 4:245-277. Barral, L., and S. Primard 1962 L'Homme du Rastel. Bulletin du Mude d'Anthropologie PrChistorique de Monaco 9:

I71-190. Bay, R.

1964 Das Menschliche Skelett. I n Birsematten-Basisgrotte. H. G. Bandi, ed. pp. 101-121. Bern: Stampfli and Cie. Bergmann, C. 1847 Ober die Verhaltnisse der Warmeokonomie der Thiere zu ihrer Griisse. Giittinger Studien

3:595-708. Billy, C.

1975 fitude Anthropologique des Restes Humains de 1'Abn Pataud. In The Excavations of the Abri Pataud. H.L. Movius, ed. pp. 901-261.Cambridge, Mass.: Peabody Museum. Binford, L. R., and W. J. Chasko 1976 Nunamiut Demographic History: A Provocative Case. I n Demographic Anthropology. E. B. W. Zubrow, ed. pp. 64-144.Albuquerque: University of New Mexico Press. Bonin, G . von 1935 European Races of the Upper Paleolithic. Human Biology 7:196-221. Bordes, F. 1968 The Old Stone Age. J. E. Anderson, trans. New York: McCraw-Hill. Bouchud, J . 1976 La Chase. I n La PrChistoire Francaise. H. delumley, ed. pp. 589-596.Pans: Centre Nationale Recherche Scientifique. Brace, C. L. 1980 Australian Tooth-Size Clines and the Death of a Stereotype. Current Anthropology 21:

141- 164. Brace, C. L., and M. F. A. Montagu 1977 Human Evolution. 2nd ed. New York: Macmillan. Brues, A. M. 1959 The Spearman and the Archer. American Anthropologist 61:457-469. 1977 People and Races. New York: Macmillan. Burch, E. S. 1972 The Caribou/Wild Reindeer as a Human Resource. American Antiquity 37:339-368. Clark, G. 1967 The Stone Age Hunters. London: Thames and Hudson. Clark, J. G . D. 1952 Prehistoric Europe. Stanford: Stanford University Press. Clarke, D. L. 1976 Mesolithic Europe: The Economic Basis. I n Problems in Economic and Social Archaeology. G. Sieveking, J. H.Longworth, and K. E. Wilson, eds. pp. 444-481.London: Duckworth Press.

Fra yer]

BODY SIZE A N D WEAPONS

71

Cohen, M. N. 1977 The Food Crisis in Prehistory. New Haven: Yale University Press. Coles, J. 1973 Archaeology by Experiment. New York: Scribner's. Collins, D. 1976 Early Man. In The Origins of Europe. D. Collins. ed. pp. 19-126. New York: Thomas Y. Crowell. Czamik, S . A. 1976 The Theory of the Mesolithic in European Archaeology. Proceedings of the American Philosophical Society 120:59-66. Debetz, G . F. 1936 The Tardenoisian Skeleton from the Shelter of Fat'ma Koba in Crimea. Antropologicheskii Zhumal 3:144-163. Ferembach, D. 1974a Le Squelette Humain Azilien de Rochereil (Dordogne). Bulletin et Memories de la SocietC d'Anthropologie de Paris 2:271-290. 1974b Le Gisement MCsolithique de Moita do SebastiPo. Lisboa: Instituto de Alta Cultura. 1978 L'Origine des MCsolithiques en France. Bulletin et Memoires de la Societe d'Anthropologie de Paris 5:231-237. Frayer, D. W. 1978 The Evolution of the Dentition in Upper Paleolithic and Mesolithic Europe. Monographs in Anthropology, Vol. 10. Lawrence: University of Kansas Press. Hamson, G . A.. J. S . Weiner, J. M. Tanner, and N. A. Barnicot 1977 Human Biology. 2nd ed. Oxford: Oxford University Press. Jellnek, J. 1972 Das Grosse Bilderlexikon des Menschen in der Voneit. Berlin: Bertelsmann LexikonVerlag. Klaatsch, H.. and 0. Hauser 1909 Homo aurignacensis hauseri. Prahistorisches Zeitschrift 1:273-338. Koztowski, S . K. 1973 The Mesolithic in Europe. Warsaw: Warsaw University Press. Laughlin, W. S . 1968 Hunting: An Integrated Biobehavioral System and Its Evolutionary Importance. In Man the Hunter. R. B. Lee and I. DeVore, eds. pp. 304-320. Chicago: Aldine. Malina, R. M. 1979 Secular Changes in Size and Maturity: Causes and Effects. In Secular Trends in Human Growth, Maturation, and Development. A. F. Roche, ed. Monographs of the Society for Re. search in Child Development 44(3-4):59- 102. Martin, R. 1928 Lehrbuch der Anthropologie. Volume 11, Kraniologie, Osteologie. Jena: Fischer. Massari, C. 1958 Alcuni Rileivi sul Quinto Scheletro della Barma Grande. Rivista di Scienze Preistoriche 13:47-53.

Matiegka, J. 1938 Homo piedmostensis. Fosilnl clovek z Piedmostf na Morave. 11. (The Fossil Site of Piedmost in Moravia, Part 11). Ostatnl &isti KostrovC. Prague: Nakladem Ceskt Akademie Vrd a Umcnl. Meiklejohn, C. 1978 Ecological Aspects of Population Size and Growth in Late-Glacial and Early Postglacial North-Western Europe. In The Early Postglacial Settlement of Northern Europe. P. Mellars, ed. pp. 65-79. London: Duckworth Press. Mellars, P. 1973 The Character of the Middle-Upper Paleolithic Transition in Southwest France. In The Explanation of Cultural Change. C. Renfrew, ed. pp. 253-276. Pittsburgh: University of Pittsburgh Press. Meredith, H. V. 1976 Findings from Asia, Australia, Europe, and North America on Secular Change in Mean

AMERICAN ANTHROPOLOGIST

72

[ 8 3 , 1981

Height of Children, Youths, and Young Adults. American Journal of Physical Anthropology 44rS15-326.

Noe-Nygaard, N. 1974 Mesolithic Hunting in Denmark Illustrated by Bone Injuries Caused by Human Weapons. Journal of Archaeological Science 1:217-248. Odell, G . H. 1978 Preliminaires d’une Analyse Fonctionnelle des Pointes Microlithiques de Bergumermeer (Pays.Bas). Bulletin de la Societe Prehistorique Francaise 75:37-49. Olivier, &orbs, Mine-Claude Chamla, Gorges DeVigne, Albert Jacquard, and Edmond I agolnitzer 1977 L’Accroissement de la Stature en France. Comptes Rendus, Academie de Science (Paris), SCrie D 285:343-345. Ortner, D. J. 1968 Description and Classification of Degenerative Bone Changes in the Distal Joint Surfaces of the Humerus. American Journal of Physical Anthropology 28:139- 156. Pope, S . T. 1962 Bows and Arrows. Berkeley: University of California Press. (1st ed. 1918). Roberts, D. F. 1952 Body Weight, Race and Climate. American Journal of Physical Anthropology 11 :533 558. 1973 Climate and Human Variability. Addison-Wesley Modules in Anthropology, Number 34.

Reading, Mass.: Addison-Wesley. Rust, R. 1943 Die Ah- und Mittelsteinzeitlischen von Stellmoor. Neumunster: K. Wachholtz Verlag. Sawtell, R. 0. 1951 Azilian Skeletal Remains from Montardit (Ariege), France. Papers of the Peabody M u seum (Harvard) 11:217-254. Semenov. S. A. 1964 Prehistoric Technology. M. W. Thompson, trans. New York: Barnes and Noble. Sonneville-Bordes, D.de 1960 Le Paleolithique SupCrieur en Perigord. Paris: Centre Recherche Nationale Scientifiqur. Spencer, L. 1974 Replicative Experiments in the Manufacture and Use of the Great Basin Atlatl. In Great Basin Atlatl Studies. T. R. Hester, M. P. Mildner, and L. Spencer, eds. pp. 37-60. Ramona, Calif.: Ballena Press. Stini, W.A. 1972 Reduced Sexual Dimorphism in Upper Arm Muscle Circumference Associated with Protein-Deficient Diet in a South American Population. American Journal of Physical AnthropolO ~ Y36:341-352.

Adaptive Strategies of Human Populations Under Nutritional Stress. In Biosocial Inter actions in Population Adaptation. E. S. Watts, F. E. Johnston, and G. W. Lasker, eds. pp. 19-42. The Hague: Mouton. 1976 The Limits of Human Adaptability and How to Exceed Them with the Best of Intentions. Ms. Paper read at the 1976 AAAS Meetings. Sturdy, D. A. 1972 The Exploitive Patterns of Modem Reindeer Economy in West Greenland. In Papers in Economic Prehistory. E. S. Higgs, ed. pp. 161-168. Cambridge: Cambridge University Press. 1975 Some Reindeer Economics in Prehistoric Europe. I n Paleoeconomy. E. S. Higgs. ed. pp. 55-96. Cambridge: Cambridge University Press. Trotter, M., and G. C. Gleser 1952 Estimation of Stature from Long Bones of American Whites and Negroes. American Journal of Physical Anthropology 10:463-514. Vallois, H. V . 1941-46 Nouvelles Recherches sur le Squelette de Chancelade. L’Anthropologie 50:65 202. 1944 Sur le Squelette de Gramat. I n Le Gisement Mksolithique de Cuzoul de Gramat. 1975

Fra yer]

BODY SIZE A N D WEAPONS

75

R. Lacam, A. Niederlander, and H. V. Vallois, eds. pp. 75-90. Archives de I'Institut de Paltontologie Humaine, Mtmoire 21. Paris: Masson. Anthropologie. In Le Gisement et le Squelette de Saint-Germain-la-RiviPre. R. Blanchard, D. Peyrony. and H. V. Vallois, eds. pp. 49-109. Archives de I'lnstitut de PalContologie Humaine, Memoire 34. Pans: Masson. 1977 Les MCsolithiques de France. Archives de I'Institut de Pakontologie Humaine, Memoire 37. Paris: Masson. Vallois, H. V., and G . Billy 1965 Nouvelles Recherches sur les Hommes Fossiles de I'Abri Cro-Magnon. L'Anthropologie 1972

69:47-74, 249-272.

Vemeau. R. !906 Les Grottes de Grimaldi (Anthropologie). Monaco: Imprimerie de Monaco. Vli-rk, E. 1961 Pnirstatkv Mladopleistocennfho ClovPka L Pavlova. Pamatky Archaeologicke 52:46-56. Weidenreich, F. 1945 The Brachycephalization of Recent Mankind. Southwestern Journal of Anthropology 3 7 . 239-251.

Weiss. K . 1972 On the Systematic Bias in Skeletal Sexing. American Journal of Physical Anthropology 37:239-251.

Wells, L. H. 1963 Stature in the Earlier Races of Mankind. In Science in Archaeology. D. Brothwell, E. Higgs, and G. Clark, eds. pp. 365-378. New York: Basic Books. kirov. E. 1940 The Skeletons from the Cave of Murzak-Koba. Sovetska;a Arkheolo6a 5:179-186. Submitted 6 July 1978 R e w e d manwcnpt submitted 6 June 1979 Second revision submifted 18 January 1980 Accepted ZJ March 1980