AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 63:39-56 (1984)

Patterns of Molar Wear in Hunter-Gatherers and Agriculturalists

B. HOLLY SMITH
Center for Human Growth and Development, The University of Michigan,
Ann Arbor, Michigan 48109

KEY WORDS Dental Attrition, Hunter-Gatherers,

Agriculturalists
ABSTRACT Tooth wear records valuable information on diet and methods
of food preparation in prehistoric populations or extinct species. In this study,
samples of modern and prehistoric hunter-gatherers and agriculturalists are
used to test the hypothesis that there are systematic differences in patterns of
tooth wear related to major differences in subsistence and food preparation.
Flatness of molar wear is compared for five groups of hunter-gatherers
(N = 298) and five groups of early agriculturalists (N = 365). Hunter-gather-
ers are predicted to develop flatter molar wear due to the mastication of tough
and fibrous foods, whereas agriculturalists should develop oblique molar wear
due to a n increase in the proportion of ground and prepared food in the diet.
A method is presented for the quantitative measurement and analysis of
flatness of molar wear. Comparisons of wear plane angle are made between
teeth matched for the same stage of occlusal surface wear, thus standardizing
all groups to the same rate of wear. Agriculturalists develop highly angled
occlusal wear planes on the entire molar dentition. Their wear plane angles
tend to exceed hunter-gatherers by about 10" in advanced wear. Wear plane
angles are similar within subsistence divisions despite regional differences in
particular foods. This approach can be used to provide supporting evidence of
change in human subsistence and to test dietary hypotheses in hominoid
evolution.

Tooth wear is one skeletal feature that pre- inids, the flatness of molar wear has im-
serves direct evidence of the masticatory be- pressed observers (see Keith and Knowles,
havior of mammals. Mastication is inti- 1911-12; Dart, 1929; Clark, 1967). Flat ver-
mately related to diet, and patterns of tooth sus oblique molar wear is one of the classic
wear can be used to make inferences about distinctions between hominid and pongid
diet in prehistoric populations and extinct dentitions (Clark, 1955). Although flat wear
species. From a n anthropological point of has been traditionally attributed to canine
view, tooth wear may record important reduction that allowed a rotary chewing mo-
stages in human biological and cultural evo- tion, the assumption that tough or fibrous
lution, including evidence of food resources foods are also responsible underlies discus-
utilized by ancestral hominids, development sions of hominid or protohominid tooth wear
of fire and cooking, invention of food process- (see Robinson, 1956; Jolly, 1970; Simons,
ing utilizing grinding tools, adoption of agri- 1976). Changes in food consistency may be
culture, invention of pottery, and other the most systematic trend in human food
refinements in the way food is processed. Mi- preparation:
croscopic features of wear (Walker et al., Among human beings, the extent
1978; Grine, 1981), rates of tooth wear (Ped- to which the process of mechani-
ersen, 1949), and the form and pattern of cal food disintegration is done in
tooth wear (Molnar, 1972) may all be inform- the mouth, depends inversely on
ative for studies that seek to reconstruct the the development of food prepara-
diet and methods of food preparation of ear- tion methods. From primitive
lier populations. One pattern that may give man's usage of stone knives,
evidence on these aspects of human adapta- pounding and grinding stones,
tion is the flatness of molar wear. Received July 5, 1983; revised September 6 , 1983; accepted
Since the earliest discoveries of fossil hom- September 8. 1983.

1984 ALAN R. LISS. INC

40 B.H. SMITH

down through the ages to modern tion? How sensitive is a given wear pattern
usage of knives, forks, spoons, fast as a n indicator of foods and food preparation?
working mincing machines, How important is a particular food, as com-
heavy rolling mills and highly re- pared with its mode of preparation?
fined foodstuffs, the mechanical
MECHANISMS
disintegration of food has gradu-
ally, but surely, been shifted from Explanation of the mechanism proposed to
inside to outside the mouth account for a correspondence between molar
(Campbell, 1939:53). wear planes and food consistency requires a
A substantial descriptive literature indi- brief review of molar wear.
cates that prehistoric human groups exhibit The development of human molar wear is
great variation in tooth wear (e.g., Nicholls, best understood a s a modification of a wear
1914; Campbell, 1925; Leigh, 1925; Rabkin, pattern common to primates and some other
1943; Pedersen, 1949). Some investigators omnivorous mammals. Hominoids, in partic-
have suggested how tooth wear might reflect ular, share a basic wear pattern due to broad
cultural changes in food preparation (Brace, similarities in crown morphology, occlusion,
1962; Greene et al., 1967; Molnar, 1972; Hin- and the dynamics of mastication (see Butler,
ton, 1981). 1973; Mills, 1955).Based on cineradiographic
The adoption of agriculture (or intensive studies of mastication, Hiiemae and associ-
collection and usage of wild grains) probably ates divided mastication into two cycles, each
represents a great change in human food and characterized by a different type of tooth
food preparation. Brace (1962)suggested that wear (Crompton and Hiiemae 1970; Hiiemae
the intensive use of grinding stones on grains and Kay, 1973; Hiiemae, 1976).In the initial
and the appearance of pottery in the Neo- cycle, termed “puncture-crushing,” teeth do
lithic are associated with a substantial re- not contact and repeatedly chop the food
duction in food toughness. The change from bolus, producing blunting wear over the tooth
a hunter-gatherer subsistence to a diet based surface. This is followed by a cycle of “chew-
on ground grains and food cooked in water ing” in which teeth shear and grind across
should produce a reduction in food tough- each other, producing the characteristic “lin-
ness, fibrousness, and resistence, and thus a gual and buccal phase” wear facets (see Fig.
reduction in the role of the teeth in break- 11, well known from the work of Butler (1952,
down of foods. The product of this change in 1973) and Mills (1955). It should be noted
food consistency, Brace argued, was a change that the mandibular buccal and maxillary
from flat molar wear to a more oblique wear lingual cusps receive wear on both faces of
pattern. cusp slopes whereas the remaining cusps
This study is an attempt to build a n inter- wear on one face (Fig. 1, left). As attrition
nally consistent model identifying the deter- advances, this differential wear tends to pro-
minants of one pattern of wear, specifically, duce a n oblique surface (Fig. 1, right). An
the flatness of the occlusal wear plane angle. oblique wear plane is the “pattern” expected
Samples of prehistoric and modern hunter- from near or actual tooth-to-tooth contact in
gatherers and agriculturalists for which “chewing.” “Puncture-crushing,” on the
there is a good archaeological or ethno- other hand, should contribute to wear on the
graphic context are used to test the hypothe- entire crown surface and lead to a flatter
sis that there are communalities in flatness wear plane.
of molar wear related to broad similarities in The relative amount of “puncture-crush-
major food sources and methods of food prep- ing” required for different foods could influ-
aration. Such a broad generalization calls for ence the obliquity of the wear plane angle.
supporting evidence from a wide-ranging This was first proposed by Taylor (1963),some
sample of human groups. In this study, the time before the current terminology
development of molar wear is compared for appeared:
ten major human groups. When tough fibrous foods are
With these broad samples and a quantita- prominent in the diet and teeth
tive approach to the study of tooth wear, it is do not so often make contact dur-
possible to investigate a number of long- ing its mastication, there will
standing questions: Are there comrnunalities tend to be less difference in the
in tooth wear patterns due to similar cul- wear of the buccal and lingual
turalltechnological features of food prepara- cusps (1963:99).
 MOLAR WEAR IN HUNTER-GATHERERS AND AGRICULTURALISTS 41

necessary for a hard or tough diet, although

M‘ M2 this is not always clearly stated (Dart, 1929;
Clark, 1955; Robinson, 1954, 1956; Simons,
1968, 1976; Jolly, 1970).
Scattered references to the effect of food
consistency on mandibular motion can be
found in the dental literature (Black, 1895;
Pickerill, 1923; Leigh, 1925) and this has
been the subject of experimental work. Stud-
ies that have traced human mandibular
movement typically find markedly different
responses to different test foods:

v Tough food usually is associated

MI M2 with grinding strokes with pro-
nounced lateral deflexion of the
Fig. 1. Development of primate molar wear pattern. mandible, whereas soft food pro-
At left are maxillary and mandibular first and second duces more vertical, chopping
molars of a gorilla drawn after Mills (1955). Buccal (B)
and lingual (L) directions are indicated. Shaded areas movements (Ahlgren 1976:127).
show location of buccal and lingual phase wear facets.
Mandibular buccal and maxillary lingual cusps (arrows)
receive wear on both faces of cusps whereas remaining
Nearly all treatments of the subject,
cusps wear on one face. As this unequal wear progresses
whether experimental or descriptive, simply
(at right), molar surfaces develop obliquely angled wear
planes. use the word “tough” as the quality of food
under consideration. Vague as it is, this rep-
resents the status of the model. Until there
are more efforts a t defining the properties of
foods and their effect on mastication, we are
Animal experiments indicate that the per- relegated to a comparison of “tough”, “resist-
iod of “puncture-crushing” can be signifi- ant,” or “fibrous” foods with “refined” or
cantly increased for “tough” foods (Hiiemae, “soft” foods. The direct “puncture-crushing”
1976). This explanation predicts that more hypothesis and the indirect “mandibular
highly oblique wear results from more me- movement” hypothesis both predict that
chanically refined foods. tough and fibrous foods produce flatter wear.
A second mechanism that could produce They are not exclusive mechanisms, and both
the same effect can be proposed. First, a wider may operate simultaneously.
mandibular movement creates more even
wear across buccal and lingual cusps. Sec-
REVIEW OF MATERIALS
ondly, mandibular movement responds to
food consistency: tougher or more fibrous A series of five, clear-cut agriculturalists
foods are chewed with a wider lateral excur- from a variety of environments were chosen
sion of the mandible. Thus tough foods also for study: Neolithic Europeans, prehistoric
contribute indirectly to flatter molar wear. and early historic Nubians, early historic
This hypothesis, in its entirety, is difficult British, Mississippian American Indians, and
to attribute to any one individual. The notion Puebloan American Indians. Hunter-gath-
that canine reduction allows a more “rotary” erers are represented by a Middle and Upper
chewing motion that creates flat wear is Paleolithic sample frbm Europe and the Near
found in early discussions of fossil hominids East, a Mesolithic sample from France, Ar-
(Keith and Knowles, 1911-1912). The fraud- chaic American Indians, recent Australian
ulent flat wear on the Piltdown molars placed aborigines, and precontact Canadian and
this feature in the limelight a t a n early date. Alaskan Eskimos. In most cases these hu-
Reduced canines and flat wear subsequently man “groups” were created by combining
became a n accepted cause and effect (Keith, material from several neighboring sites or
1913; Gregory, 1922; Schultz, 1925; Gregory subperiods. Groups and their sub-member
et al., 1938). Most discussions seem to as- sets are listed in Table 1, with brief refer-
sume that a “rotary” type of mastication is ences to background information. All mater-
42 B.H. SMITH

TABLE 1. Classification and summary of study sample, with periods or

approximate dates, and references
Group
sitelsub-group PeriodDate Number ~
Reference'
Hunter-gatherers
Paleolithic
Western Europe Middle paleolithic' 6 Oakley e t al., 1971
Levant Middle-transitional3 7
France Upper paleolithic4 14
Mesolithic (France)
Hoedic ca 6570 B P 7 Pequart and Pequart, 1934
Teviec ca 6570 BP 8 Pequart et al., 1937
Baume de Montclus ca 6000 BP 1 Ferembach, 1974a
Rochereil ca 7000 BP 1 Ferembach, 1974b
Australian
South Australia pre-1900 32 (University of Edinburgh)
Northern Territories 26
New South Wales 9
Victoria 6
West Australia 4
Eskimo
Thule (Canada) ca 1200 AD 58 Merbs, 1968
Point Hope (Alaska) pre-contact 20 (Smithsonian)
Archaic
LU 25 (Alabama) 2 4700 BP 46 Webb and DeJarnette,
1942
Subtotal 298
Agriculturalists
Neolithic Europe
British Neolithic, n.d. 15 Duckworth, 1898
Dolmen de Sauveterre ca 1400-1000 BC 15 Prunieres, 1873
(France)
Nubian
A-Group ca 3300-2800 BC 28 Nielsen, 1970
C-Group ca 2300-1800 BC 42
Pharaonic ca 1800-1200 BC 7
Meroitic ca AD 0-350 41
X-Group ca AD 350-550 33
Christian ca AD 550-1000 11
C-, X-, or Meroitic ca 2300 BC-AD 350 13
Historic British
Maiden Castle Iron Age, AD 42-47 26 Wheeler, 1940
Poundbury Iron Age 30 Green, I974
Burwell Anglo-Saxon 24 Brash et al., 1935
Comet Place Medieval, ca AD 25 (Cambridge University)
1300
Mississippian
LU 25 (Alabama) ca AD 1200-1500 23 Webb and DeJarnette,
MS 80 (Alabama) 3 AD 1000 32 1942
Webb and Wilder. 1951
Puebloan
Gran Quivira (New Mexico) AD 1400-1675 57 Hayes et al., 1981
Subtotal 365
Total 663

'Reference given for publications about particular site, dates may come from other sources (see Smith, 1983a). Location
of collection given if no publication is available.
:Sites: Montrnaurin, La Ferrasie, La Quina, Gibraltar.
.'Sites: Tabun, Skhul, Qafzeh.
*Sites: Cro-Magnon, St. Germaine-la-Riviere, Grotte dIsturitz, Abri Pataud, Abri Lachaud, Laugerie-Basse.
 MOLAR WEAR IN HUNTER-GATHERERS AND AGRICULTURALISTS 43
ial is in research/museum collections in the vanced caries, were eliminated from study).
United States and Europe.' By all evidence, both hunter-gatherers and
The hunter-gatherers probably span agriculturalists had a n abrasive diet. Rate of
greater diversity in food preparation and food wear alone (studied using chronological age
sources than do the agriculturalists. Austral- estimates) will not differentiate them. The
ian aborigines, for example, used grinding dry, sandy environment of the agricultural-
stones to produce a grass seed cake (these ist Nubians, for example, apparently led to a
cakes have been described as "hard as a rate of wear as high as that of Australian
brick" [Nicholls, 19141). This practice was aborigines and Archaic American Indians.
unavailable to Eskimos, who, however, did Patterns of tooth wear may be sensitive to
boil meat in stone bowls over lamp flames differences in food consistency that are less
(see Jenness, 1922). Raw, frozen, or dried affected by local environmental factors. Since
meat also formed a major part of Eskimo it is not possible to prove a difference in food
diets by most accounts (Thalbitzer, 1914; Jen- consistency between the two subsistence
ness, 1922; de Poncins, 1941). Although the groups (in the sense of measurements of the
five groups of hunter-gatherers subsisted on toughness of native foods), this study must
a variety of plant and animal foods, none had instead build a n internally consistent model
pottery and none are thought to have had a of the correspondence of tooth wear and
diet primarily based on ground grains; even subsistence.
grinding stones are absent in some groups. Samples were purposely chosen to cross-cut
In contrast, the five agriculturalist groups a variety of specific environments, particular
are known to have made extensive use of foods, and craniofacial forms to minimize the
grinding stones to grind some grain crop, and chance that observed differences in tooth
to have used water-tight, fire-resistant con- wear are merely reflections of idiosyncracies
tainers (pottery) to cook food for long periods of particular people, places, and times. Para-
in water. An overview of the diet and food doxically, tooth wear is of little use for a
preparation of the ten human groups ana- synthetic understanding of diet and masti-
lyzed here can be found in Smith (1983a). cation if it is either completely nonspecific or
There is some independent, skeletal evi- so specific that it is idiosyncratic. This study
dence of a difference in foods. Caries fre- was designed to determine if there are com-
quency is a good indicator of the approximate munalities in molar wear related to broad
amount of processed carbohydrates in the diet levels of subsistence or food preparation that
(see Hardwick, 1960; Moore and Corbett, overcome regional factors.
1978; Turner, 1979).Caries frequency is near
zero in the five groups of hunter-gatherers, METHODS
and substantial in the five groups of agricul- Differential buccal and lingual wear is
turalists, supporting the proposed differences viewed as a continuum that may be altered
in proportion of processed grains in the diets or modified in degree, but not categorically
of hunter-gatherers and agriculturalists (ex- changed by food consistency. Compilation of
act frequencies cannot be given since discrete data on the location of wear has not
extreme pathologies, including cases of ad- proven to be productive (e.g., Zuckerman,
1954; Gantt, 1979).In this study, the angle of
the occlusal wear plane surface relative to a
'Sources of material: Neolithic Europe, Duckworth Laboratory horizontal occlusal plane was measured with
of Physical Anthropology, Cambridge University, and the Musee
de 1'Homme; Nubians, Laboratory of Physical Anthropology, a modified protractor (see Fig. 21, similar to
The University of Copenhagen, and University of Chicago col- instruments used by Butler (1972) and Hall
lection in the care of Dr. Charles Merbs of Arizona State Univer- (1977). Angles were recorded to the nearest
sity: historic British, British Museum (Natural History), and
Duckworth Laboratory, Cambridge University: Mississippian 0.5". Slopes to the buccal (as in Fig. 2) are
American Indians, collections of the Alabama Museum of Natur- arbitrarily designated as positive and slopes
al History on loan to the University of Alabama; Puebloan
American Indians, Gran Quivira National Monument collection to the lingual are designated negative. Wear
on loan to Dr. Christy Turner a t Arizona State University; plane angles are measured on occluding sur-
Skhul V, Harvard Peabody Museum, Harvard University; Mid-
dle and Upper Paleolithic, British Museum (Natural History), faces on upper and lower teeth: on the tal-
Musee de I'Homme, and Institut de Paleontologie Humaine; onid basin of mandibular molars and across
Mesolithic, Institut de Paleontologie Humaine; Archaic Ameri- the trigone of maxillary molars. These are
can Indians, collections of the Alabama Museum of Natural
History on loan to The University of Alabama: Australian abo- the primary crushing cusps and basins,
rigines, University of Edinburgh Medical School, Department of which are also the areas of greatest wear.
Anatomy; Thule Eskimos, Canada National Museum of Man
collection on loan to Dr. Charles Merbs a t Arizona State Univer- Unworn teeth are measured from cusp to
sity; Alaskan Eskimos, Smithsonian Institution. cusp, not from cusp to fossa. Both teeth must
44 B.H. SMITH

c I:/ .........'..-...
9
x

Fig, 2. Instrument used to measure occlusal wear trigone) and the far arm is placed on the homologous
plane angles, constructed by mounting a Sears Roebuck point on the antimere. The bar is brought down across
and Co. metal protractor no. 9-4029 on a straight edge. the occlusal surface and the pointer indicates the angle
Metal pieces were added to the straight edge to make of deviation of the crown surface from a horizontal de-
the arms flush with the protractor face. The instrument fined by the base of the instrument. This mandible has
is centered on highest point on the molar talonid (or a wear plane angle of 17" to the buccal on the first molar.

be in place in bone, and teeth partly extruded applicable to a variety of human groups;
from their sockets during life, obviously dis- stages were altered in scope and more pat-
located, or virtually destroyed by dental car- tern variability was anticipated. A prelimi-
ies cannot be measured. nary version of this scale was developed in
Wear plane angles in the study sample cooperation with R. Hinton (see Hinton,
showed a total range of more than 80°, from 1981).The present version, fully described in
-30" to +34" in the mandible, and from Table 2 and illustrated in Figure 3, differs in
-43" to +41" in the maxilla. Remeasure- some details. Only the inolar scale is used
ment of wear plane angles for 247 teeth here. Replicability is about 90% for molars
showed good replicability. Approximately and declines to about 85% for teeth that are
70% of errors (deviation from original mea- simpler in form, based om a n error study in-
surement) were within *LO", and 80%were volving N = 1000 rescores. Errors of more
within & 1.5" of the original measurement. than one stage are rare or nonexistent after
The goal of the analysis was to compare some practice with the s8ystem.To date, the
development of wear plane angles indepen- molar wear scale seems to be applicable to a
dently of population differences in rates of wide variety of human groups, pongids, and
overall wear of the occlusal surface. There- even to omnivorous fossil mammals. Special
fore all populations were standardized to the scales must be designed for teeth quite differ-
same rate of wear by studying wear plane ent in form and functi'on (e.g., pongid ca-
angles by stage of wear rather than by chron- nines).
ological age. This analysis is concerned with This analysis presents data for the right
the angle of wear at some stage x, regardless side, using left side values only when a right
of whether this stage occurs at age 20 or 30 side tooth was damaged. Use of one side or
years. another is preferable to averaging sides when
Amount of molar wear has traditionally a n ordinal variable (wear stage) is involved.
been scored on an ordinal scale (Broca, 18791, Since there can be considerable asymmetry
and more than a dozen such scales appeared in tooth wear, a right MI wear stage was
in the literature from about 1930 to 1980. A never paired with a left MI angle, etc.
pilot study using different scales indicated Dental pathologies were noted and taken
that the best system for this study was the into account in all analyses. At least 16 scor-
eight-stage system of Murphy (195913, see also able teeth per dentition were required for
1959a). His system, however, was based on inclusion in the study (less than 16 were al-
his experience with Australian aborigines, lowed for Paleolithic specimens since mate-
and it created difficulties when used on rial is limited and pathologies are rare). This
American Indian test samples. Murphy's cut-off was based on the reasonable assump-
scale was modified to make it more widely tion that individuals who have lost about
 MOLAR WEAR I N HUNTER-GATHERERS AND AGRICULTURALISTS 45

TABLE 2. Descriptions of stages of occlusal surface wear

Molars Premolars Incisors and canines
0. Missing or cannot be coded 0. Missing or cannot be coded 0. Missing or cannot be coded
1.Unworn to polished or small 1. Unworn to polished or small 1. Unworn to polished or small
facets (no dentin exposure) facets (no dentin exposure) facets (no dentin exposure)
2. Moderate cusp removal (blunting). 2. Moderate cusp removal 2. Point or hairline of dentin
Thinly enamelled teeth (human decid- (blunting) exposure
uous molars, chimpanzee molars) may
show cusp tip dentin but human
permanent molars show no more than
one or two pinpoint exposures
3. Full cusp removal and/or some dentin 3. Full cusp removal and/or 3. Dentin line of distinct
exposure, pinpoint to moderate moderate dentin patches thickness
4.Several large dentin exposures, 4. At least one large dentin 4.Moderate dentin exposure
still discrete exposure on one cusp no longer resembling a line
5. Two dentinal areas coalesced 5. Two large dentin areas (may 5. Large dentin area with
be slight coalescence) enamel rim complete
6 . Three dentinal areas coalesced, or 6 . Dentinal areas coalesced, 6 . Large dentin area with
four coalesced with enamel island enamel rim still complete enamel rim lost on one side
or very thin enamel only
7. Dentin exposed on entire surface, 7. Full dentin exposure, loss 7. Enamel rim lost on two sides or
enamel rim largely intact of rim on a t least one side small remnants of enamel remain
8. Severe loss of crown height, break- 8. Severe loss of crown height; 8. Complete loss of crown, no
down of enamel rim; crown surface crown surface takes on shaDe enamel remaining: crown surface
takes on shaDe of roots of roots takes on shape ofyoots

half their teeth to painful, infectious disease ing wear is, logically, a n independent varia-
do not have normal patterns of mastication. ble, i.e., the cause of change in wear plane
Dropping out all pathologies, on the other angle. Using stage of wear as a categorical
hand, would limit inference to a highly re- variable ignores the inherent ordered rela-
stricted subset of the population, and elimi- tionship of one stage to the next and does not
nate most old individuals in some popula- make full use of the data. Although only a n
tions. Pathologies in remaining dentitions ordinal variable, eight stages of wear provide
were classified by level of severity, and tested some approximation to the underlying con-
for systematic effects. Dentitions with sev- tinuous phenomenon. Error in the x variate
eral major carious lesions and/or abscesses, does not demand a major axis fit if error is
severe and widespread periodontal destruc- uncorrelated with x (Sokal and Rohlf, 1969),
tion, or several molars lost antemortem were a condition that is fulfilled here. Least
deemed “pathological.” Separate analyses of squares regression assumes that there is a
wear plane angles were performed both with normal distribution in y for every x. This
and without these dentitions for each of the assumption is satisfied in these data, how-
ten human groups. Remarkably, removal of ever they do not make a good approximation
“pathological” dentitions amounting to as to a bivariate normal distribution as re-
much as 20% of the total sample has almost quired for a major axis fit. Finally, develop-
no effect on results, even though this also ment of wear plane angle with stage of wear
preferentially removes old individuals. Thus, appears to be approximately linear (see Fig.
there is no reason to believe that the more 4). Small deviations are not critical if the
disease- or caries-prone individuals in any same range in x is compared in each regres-
group have systematically (directionally) dif- sion. Even so, individual means and tests not
ferent wear plane angles, or that pathologies assuming linearity are also provided and ul-
per se produce observed between-group dif- timate conclusions do not depend on a strictly
ferences. It should be realized, however, that linear model.
as all dental tissues break down in terminal
wear stages, extreme variants can occur. RESULTS
Parametric statistics are used in the anal- The full data set for wear plane angle of
ysis of the relationship of wear plane angle the mandibular first molar organized by
to stage of wear. Stage of wear is the indepen- stage of wear appears in Figure 4,with agri-
dent variable (x) and angle of the occlusal culturalists a t left and hunter-gatherers a t
surface is the dependent variable inor) right in each paired column. Several features
regression analysis. This method of analysis of wear plane angle development are univer-
can be justified on several grounds. Increas- sal and do not depend on subsistence group.
46 B.H. SMITH

INCISORS &
MOLARS PREMOLARS CANINES
L U U U

I 0
2

[1z
3
a
w
3 4
LL
0
<n
W
(3
5 0-0-0
a
I-
cn 6 @-@a8 - 0
7 0-0-8@ - @
8 8-8-8 0 0 0
Fig. 3. Diagrams of crown surfaces used to score stages one arch to the other except in the case of premolars,
of tooth wear (matched to descriptions in Table 2). Bars where examples for each are given. Borderline judg-
between outlines connect common variants of pattern ments at stages 1-3 should be hased on cusp removal,
and degree of wear allowed in a stage. Abbreviations L and at stages 4-8, proportion of dentin versus enamel on
(lower) and U (upper) designate the arch used in the the crown.
prototype. There is no difficulty applying prototypes for

First molars erupt with a lingual orientation trated on the buccal half of mandibular
of the crown (negative occlusal surface an- crowns to some degree, tends to flatten the
gle). This is typical of mandibular molars, occlusal surface angle and finally to reverse
which show increasingly severe lingual incli- it. This is a good explanation for the confus-
nations from mesial to distal molars (Demp- ing reports of both flat arid oblique wear in a
ster et al., 1963). Wear, typically concen- single human group (see Rabkin, 1943;
 MOLAR WEAR IN HUNTER-GATHERERS AND AGRICULTURALISTS 47

Moses, 1946). Since an individual tooth may has been predicted). Agriculturalists develop
change in angle some 15-30" in a lifetime, higher angles than hunter-gatherers, ulti-
variance is quite large if age (or wear stage) mately reaching a 10" difference in high wear.
is ignored. At stage 8, the severely-angled remnant of
When occlusal plane angle is sorted by stage the crown tends to fracture in agricultural-
of wear, differences in central tendency of the ists, but not in hunter-gatherers. This is
two subsistence groups can be seen, even probably because agriculturalists often de-
though ranges and standard deviations are velop deeply cupped-out dentin that further
high when sorted by only eight stages. Table 3 weakens the tooth. Hunter-gatherers do not
summarizes mean wear plane angles for ag- show cupping to this degree, and molar wear
riculturalists and hunter-gatherers. At wear plane angle continues to increase at stage 8
stages 1and 2 the two groups are not signifi- wear. Recent experimental evidence sug-
cantly different, suggesting that angles of gests that cupping results from the presence
eruption of first molars are not widely differ- of fine particles in food (see Costa and
ent (notethat sample sizes become adequate at Greaves, 1981).Cupping itself may be a sign
stage 2). All subsequent comparisonsindicate of processing grains with grinding stones, in
that wear planes are significantly different at which grains are reduced to fine particles
comparable wear stages (the more conserva- and fine particles of stones are introduced
tive two-tailed p values are reported through- into food. This differential stage of fracture
out although one-sided tests would be does not affect the analysis to a great degree
appropriate since the direction of difference since few relatively healthy dentitions can

40"

RAW DATA
AGRICULTURALISTS IN LEFT COLUMN
30'- HUNTER-GATHERERS IN RIGHT COLUMN
i I

2 I
I
I 2 I1
I I 4 31
I 5 2

5 I 61 7 3
I I 41 51 I 1
I 2 3 12 5 1
I 72 52 63 65
21 8 4 72 3 35
4 62 53 3 3
I I x5 75 5 24
34 X6 7 -3 33 _ _ _ - _ 2 _ - . _ .
- 3 2- ---- 3 3 - - - - 25----.
13 85 14 I
13 24 12 I 1 I
31 14
23 I I
42 2 I
2 I
22 21 I
2 2
I
I
I

z
a

I 2 3 4 5 6 7 0
STAGE OF M, WEAR
Fig. 4. Raw data for human MI wear plane angle indicates a perfectly flat (0") surface. Note that wear
(ordinate) at each stage of wear (abscissa). Wear plane causes a change in angle from lingual to buccal, and that
angles for agriculturalists are at left and hunter-gath- the relationship between wear plane angle and stage of
erers at right in each paired column. Numbers indicate wear is approximately linear.
N at each point and X stands for N 10. The dotted line
48 B.H. SMITH

TABLE 3. Student's t-tests comparing M I wear plane angle at each stage of wear for agriculturalists
versus hunter-gatherers
M, Wear olane a n&
Wear Agriculturalists Hunter-gatherers Mean
stage Mean S.D. N Mean S.D. N difference' W
~ -.

1 - 10.3 5.3 17 - 6.5 10.2 5 -3.8 ,273

2 -2.8 7.1 28 -5.0 6.6 26 2.3 ,230
3 3.6 5.5 72 -0.9 5.6 39 4.5 ,001
4 5.7 4.7 46 1.5 4.8 25 4.2 ,001
5 11.1 6.8 40 4.1 4.1 21 6.9 ,001
6 11.9 5.9 40 7.4 6.5 31 4.5 ,003
7 18.7 10.6 21 8.9 7.3 27 9.8 .001
FI
" - - 3
- - - 5 - -

' A g k u l t u r a l i s t minus hunter-gatherer value, in degrees.

be found to measure a t this stage of wear in wear and wear plane angle are related.
any group. Dentin cupping itself does not Regression slopes are directly interpretable
alter the measurement of angle of wear be- as degrees of change in angle per stage of
fore enamel borders fracture. wear. Rates of change are generally higher
Sample sizes a t each wear stage are insuf- in agriculturalists (3.4-5.3 per wear stage)
O

ficient to compare means for each of the ten than in hunter-gatherers (1.9-3.5 per wear O

human groups. Comparisions of individual ~ t a g e ) Analysis

.~ of covariance was used to
groups are based on least squares regression determine if slopes and overall regression
lines projected through available data (see lines are significantly different. The five ag-
Fig. 5). Agriculturalists (solid lines) and riculturalist regression lines cannot be dis-
hunter-gatherers (broken lines) are inter- tinguished statistically and may therefore be
mixed a t stage 1, indicating no systematic described by a single line with a slope of 4.0
difference in orientation of unworn crowns. (intercept of 10.2"). Similarly, the hunter-
Lines begin t o separate about stage 3 and do gatherers overall have a single line with a
not overlap subsequently; by stages 5-8 cen- slope of 2.9 (intercept of 10"). The five
tral tendencies are quite distinct. The tightly hunter-gatherers seem to be less homogene-
clustered lines representing agriculturalists ous than the agriculturalists, and slightly
are above all hunter-gatherers, showing the larger samples might confirm this statisti-
development of oblique wear planes in the cally. Overall slopes of regressions (rates of
former. Hunter-gatherers, except the Paleo- change) for the two subsistence categories
lithic sample, also appear to be quite similar. are significantly different Srom each other at
The Paleolithic stands out from other groups, p < .001. Agriculturalists develop angles
appearing to maintain the flattest wear. This some 8-10' higher than hunter-gatherers in
is intriguing since one would expect the advanced wear stages. The difference in an-
heaviest demands to have been placed on the gle a t stage 7 predicted from least squares
dentition in the Paleolithic, but unfortu- lines (8") is quite close to the actual mean
nately, a sample size of ten is insufficient to difference (10"). Statisticail confirmation is
confrm this difference statistically.2 Hunter- not dependent on a linear model; trends are
gatherers have relatively flat, even wear clear enough that a simple sign test based on
across the crown surface that leads t o a low individual means of the ten human groups
rate of change in the occlusal surface angle. supports the hypothesis that agriculturalists
No group shows wear that is perfectly flat develop more oblique wear than hunter-
with respect to the tooth axis; this would lead gatherers(p < .05).
to a steadily maintained angle of about -5". At first molars, the overall difference in
Table 4 documents least squares lines and rate of change is about 1"per stage of wear.
statistical tests for regression lines shown in Although this may seem to be minor, the
Figure 5. Individual slopes are significantly magnitude of the difference can be consider-
different from zero, confirming that stage of
"The Eskimo regression line i s pulled up sharply because of a n
'Paleolithic specimens in M I analysis include: Montmaurin, extreme lingual orientation of unworn crowns; final wear planes
Tabun I, Skhul V, Abri Pataud, Abri Lachaud 3 and 5, Laugerie- a r e quite flat. Eskimo dental arch morphology is different enough
Basse, St. Germaine-La-Riviere B1, B3, and B4. to strain t h e limits of this simple linear model.
 MOLAR WEAR IN HUNTER-GATHERERS AND AGRICULTURALISTS 49

able in advanced wear. When teeth are Meroitic (combined since they each span a
matched for stage of wear, the differences in relatively short time). Pharaonic and Chris-
wear plane angle of molars of hunter-gath- tian are omitted because sample size is less
erers and agriculturalists are visible to the than eight individuals. Least squares lines
eye (see Fig. 6). for all groups are still above lines for hunter-
Analyzed in this fashion, wear plane angle gatherers (shaded region), although individ-
appears to be sensitive to major differences ual data points overlap. The early A-Group,
in food and food preparation. With large sam- thought to have a more mixed economy, is
ples it should be sensitive to finer gradations perfectly intermediate between other more
of change in foods and methods of prepara- clear-cut hunter-gatherers and agricultural-
tion. If so, a time trend might be visible in ists. There is a distinct trend toward more
the Nubian sample since it is of moderate oblique wear in more recent groups parallel-
size and spans several periods thought to ing decreases in craniofacial robusticity
show increases in the intensity of agriculture through time described by Carlson (1976)for
(see Carlson and Van Gerven, 1979).Figure 7 this same sample. Trends in tooth wear and
shows regression lines for MI wear plane an- craniofacial robusticity could both be re-
gle for A-Group, C-Group, and X-Group plus sponses to a n increasing reliance on ground

30'

LEAST SQUARES LINES M1 WEBLOAN N=27

f 25'
- AGRICULTURALISTS
2 -- HUNTER-GATHERERS MISSISS lPPlAN N= 36
0
w 20' BRITISH N.67
z
4
a 15
K
a
W AUSTRALIAN N.62
3 MESOLITHIC N.13
-I l o
3
3
0 5
0
0 / ,PALEOLITHIC N=IO
/
LL /

0 0
W
-I
c3
z
6 -5

-10

STAGE OF M, WEAR
Fig. 5. Least squares lines from regression of MI wear plane angles than hunter-gatherers (broken lines). The
plane angle on stage of wear for all ten human groups. Paleolithic sample stands apart from recent hunter-
All groups begin with a lingual orientation of the crown gatherers, showing the flattest molar wear found in this
surface that is gradually changed to the buccal with study.
wear. Agriculturalists (solid lines) develop higher wear
50 B.H. SMITH

TABLE 4. Least squares slope and intercept for M I angle regressed on stage of MI wear for
agriculturalists and hunter-gatherers
Group Slope Y-Intercept Number P'
Agriculturalists
Neolithic 4.3 -11.4 14 .001
Nubian 4.2 -11.3 123 ,001
British 3.4 -7.2 67 .001
Mississippian 3.5 -7.9 35 ,001
Puebloan 5.3 - 16.4 27 .001
Overall 4.0' - 10.0 266 ,001
Hunter-Gatherers
Paleolithic 1.9 -4.2 10 .038
Mesolithic 2.1 -6.7 13 ,090
Australian 2.6 -8.3 60 ,001
Archaic 2.9 -9.5 38 ,001
Eskimo 3.5 - 13.9 58 ,001
Overall 2.9' - 10.2 179 ,001
'Slope significantly different from zero.
"Agriculturalists a r e significantly different from hunter-gatherers a t p < .001.

Fig. 6. Lateral views of well-worn mandibles of an exposed on the entire occlusal surface. The agricultural-
agriculturalist (A) and hunter-gatherer (B) illustrating ist (Nubian A-Group) shows a moderately oblique wear
differences in wear plane angle in advanced wear. Both plane, whereas the hunter-gatherer dentition (Thule
human first molars are at stage 7 wear, with dentin Canadian Eskimo) is worn flat.

grains or increasing thoroughness of prepar- summarizes estimates of Ieast squares pa-

ation of all foods and lessening demands on rameters and results of statistical compari-
the masticatory apparatus. These sample son for all mandibular and maxillary molars.
sizes are not large enough to show that the Agriculturalists show higher rates of change
rates (regression slopes) are statistically dif- in angle at every molar tooth. Rates for M1
ferent (3.0",4.0°,and 4.7" from early to late). or M2 differ a t p < .001, but statistical sig-
A simple sign test suggests that the trend in nificance is not reached for third molars.
mean wear plane angle is significant a t Sample size for third molaas is greatly dimin-
p < .05, since 12 out of 15 pairwise compari- ished for highly worn teeth, and it is only in
sons of individual means are in the proper advanced wear that diet,ary differences are
direction. These results are encouraging for clearly seen.
the use of wear plane angles to help identify Figure 8 illustrates overall least squares
change in human subsistence. Obviously, the lines €or the entire mandibular arch. Molar
smaller the scale of the comparison, the wear stages for M1, Mz, and M3 are stag-
larger the sample size needed. gered on the abscissa in order of eruption,
This analysis can be applied to the entire although they are spread apart somewhat for
molar dentition with similar results. Table 5 clarity. Wear plane angle for each tooth is
 MOLAR WEAR IN HUNTER-GATHERERS AND AGRICULTURALISTS 51

YEARS BP
M+X 2000
MEROITIC + X-GROUP N.54
C GROUP N.37
A GROUP N=15
C 4000

A 5000

HUNTER-GATHERERS
a
0

-
I 2 3 4 5 6 7 0
STAGE OF M, WEAR
Fig. 7. Least squares lines from regression of MI wear over 3000 years. Trends in tooth wear and craniofacial
plane angle on stage of wear for three time periods in robusticity should be responses to an increasing reliance
Nubia. Range of least squares lines (not individual data on agriculture (see Carlson and Van Gerven, 1979). Al-
points) of hunter-gatherers is shaded for comparison. though the trend is significant by a sign test, larger
Although all three groups fall above hunter-gatherers samples are required for reliable determination of small-
as expected, molar wear tends to become more oblique scale differences in rates of change.

plotted on the same scale on the ordinate. cal age were used a s the independent var-
Lines are stopped a t stage 7 for Mz and stage iable. When each molar is sorted by its proper
6 at M3 to emphasize the lack of data beyond stage of wear, hunter-gatherers and agricul-
these points. Figure 8 shows that all molars turalists differ consistently in wear plane
are initially inclined lingually, increasingly angle.
from MI (ca. -5" to -10') to M2 (ca. -10" to Figure 8 is also informative on the effect of
-15") to M3 (ca. -15" to -20"). Since each dental arch morphology on wear plane angle.
tooth erupts a t a different time, begins with This picture demonstrates that the entire
a different orientation, and proceeds at its molar dentition develops in similar fashion
own rate of wear, surfaces of the three molars with increasing tooth wear, countering the
contrast in angle on a single dental arch. objection that minor morphological variants
Each molar in a dentition is in a different can greatly change wear plane angle devel-
phase of development of wear plane angle a t opment. The second molar tends to have a
one point in time. Given these sources of slightly higher rate than the first (ca. +0.4"),
variability, it should be evident that the pic- but this effect is not consistent across indi-
ture would be much less clear if chronologi- vidual human groups. On closer inspection,
52 B.H. SMITH

the only tooth to differ much from the overall

pattern is the third molar. This tooth has a
slowed rate of change compared to M1 (ca.
- 1.0') in eight of the ten human groups.
Human mandibles are markedly widened
a t third molars, and occlusion of these teeth
is slightly different (Smith, 1938b). The
change in arch width relationship probably
alters the path of excursion of lower cusps
across upper cusps. Tooth wear tends to be
more evenly distributed across the crown on
third molars than it is on first or second
molars.
If relative width of upper and lower arches
can influence wear plane angles, can it ac-
count for the differences between hunter-
gatherers and agriculturalists? An attempt
was made to take population differences in
*wm cr-e maxillary molar overjet into account in
Lo?? W??
I l l regression analysis. Overjet was calculated
for M1 and M3 as the maxillary arch width
minus mandibular width. Whether MI wear
plane angles were adjusted for overjet within
populations, within subsistence divisions, or
for all humans, results were virtually un-
changed. This factor seems to be capable of
explaining no more than 2-3% of the vari-
ance in wear plane angle at first molars. The
direction of differences in overjet between
hunter-gatherers and agriculturalists sug-
gests that, if anything, subsistence differ-
ences have been underestimated in the
simple analysis (as in Table 5). These find-
ings may help clarify some aspects of human
mastication, but they cannot easily explain
the systematic differences between hunter-
gatherers and agriculturalists. At third mo-
lars the overjet factor becomes more impor-
tant, and may explain 5-10% of the variance
in M3 wear plane angle. In general, the man-
dible is slightly better than the maxilla, and
anterior molars are better than posterior mo-
lars for use as dietlsubsistence discrimina-
tors. The mandibular first molar has the
lowest variance of angle of wear plane, and
the smallest species and population differ-
ences in structure of the dental arch. Fur-
thermore, first molars always include the
greatest number of teeth in the advanced
wear stages that are most distinctive. Al-
though third molars are the poorest teeth for
these purposes, even these highly variabIe
teeth show the same trend.
DISCUSSION
It was proposed that changing from a
hunter-gatherer subsistence to a diet based
 MOLAR WEAR IN HUNTER-GATHERERS AND AGRICULTURALISTS 53

more heavily on ground grains and food stage of tooth wear, but even they show a
cooked in water has the effect of reducing slight predominance of buccal wear that
food toughness, resistance, or fibrousness, steadily changes the occlusal surface angle
leading to a reduction in masticatory forces (see Fig. 9). What seems to be a slight modi-
and amount of food preparation done by the fication of a wear pattern adds up to a consid-
dentition. As tools increasingly take over the erable difference, on the order of 8-10' in
function of teeth, molar wear planes should advanced wear.
become increasingly oblique. This may be a n Portions of remaining within-group vari-
indirect effect due to a lessened lateral man- ance might be explained by differences in
dibular excursion responding to more refined axial inclinations of tooth roots, maxillary
foods, or a direct effect of less "puncture- arch overjet, particular dental pathologies,
crushing" mastication. individual food preferences, or sex differ-
When wear plane angles are ordered by ences. Future studies could attempt to esti-
comparable stages of occlusal surface wear, mate the contribution of some of these
consistent differences can be seen between factors, although large samples might be re-
hunter-gatherers and agriculturalists. Hunt- quired. For example, experience with these
er-gatherers are the only groups to maintain data suggests that sex differences in human
relatively flat, even wear a t a n advanced wear plane angles should not be addressed

25'

2d LEAST SQUARES LINES /

AGRICULTURALISTS

w
Z 15'
a
J
a
LT 10'
a
W
3 5(
a
u
u
I
_I

J 3
m
3f
3 0' FLAT-
0 _J

0 a
3
0 -5' c3
LI
z
J
0
w
J
c3
z
a
-Id

-16
I
-26

M,I) I 2 3 4 5 6 7 8
M29 1 2 3 4 5 6 7
M39 1 2 3 4 5 6
MOLAR WEAR STAGE
Fig. 8. Overall least squares lines from regression of molar teeth differ primarily in y-intercept. Note that
wear plane angle on degree of wear for mandibular den- agriculturalists (solid lines) always develop higher wear
titions of agriculturalists and hunter-gatherers. Stages plane angles than hunter-gatherers (broken lines) re-
of wear for MI, M2, and M3 are staggered on the abscissa gardless of tooth position.
in order of eruption, spread apart for clarity. The three
54 B.H. SMITH

Fig. 9. Occlusal views of moderately worn mandibles restricted pattern of wear on the first molar. The hunter-
showing pattern of dentin exposure on teeth of an agri- gatherer (Mesolithic site of Hoedic) shows wear more
culturalist (A) and hunter-gatherer (B), both with left evenly distributed across the crown surface. This seem-
first molars in stage 5 wear. Both mandibles show some ingly small difference in the relative restriction of wear
degree of predominance of wear on buccal cusps, Note ultimately produces a substantial difference in angle of
that the Nubian agriculturalist (X-Group) shows a very the occlusal wear plane.

without a t least 40 individuals of each sex of OH 5 (“Zinjanthropus’?gives a wear plane

and a prior hypothesis concerning the direc- angle for M1 that is far outside the range of
tion of difference. angles found in comparable human denti-
The systematic differences between hunt- tions. In this case, wear is so flat with respect
er-gatherers and agriculturalists cannot eas- to the tooth axis that it appears that the
ily be explained by morphological factors original buccal angle of eruption of the tooth
since each division covers a wide range of did not change. Hominids from Australopi-
tooth size, tooth morphology, dental arch size, thecus afarensis to Homo erectus could show
and craniofacial form. Molar wear is smilar intriguing differences in the development of
within subsistence divisions whether agri- angle of molar wear planes that might give
culture is based on wheat or corn, and information on changes in adaptation and
whether meat comes from sea mammals or behavior.
marsupials. However, larger samples could
CONCLUSIONS
clarify some smaller-scale differences.
This method would be of particular interest Angle of molar wear planes can be shown
applied to transitional stages in the origin of to have a n orderly development with increas-
agriculture. Prepottery intensive collectors ing tooth wear on all maxillary and mandi-
of wild grains or “incipient agriculturalists” bular molars. Molar wear is more evenly dis-
may be difficult to recognize archaeologi- tributed on molars of hunter-gatherers, re-
cally. In a number of cases the dates of inten- sulting in a relatively low wear plane angle
sive utilization of grain are being pushed in advanced wear. Agriculturalists show a
back far before confirmed agriculture, as in more restricted pattern of wear and tend to
the Middle East and Egypt (Wendorf, 1968; develop oblique wear planes. This difference
Reed, 1977; Kryzaniak, 1981; see also is attributed to a reduction in food toughness
Schoeninger, 1982). Analysis of tooth wear or fibrousness that is associated with the ap-
could provide supporting evidence for change pearance of intensive collection of grains and
in diet or food preparation. intensive use of grinding stones and pottery
A change in wear plane angles should be in food preparation. This model of the rela-
an early sign of a change in diet in basal tionship between food consistency and tooth
hominids that still show substantial morpho- wear may be tested or sharpened by applying
logical similarity to pongids. Early hominid the analysis to more human groups.
tooth wear may develop in the mode of hu- Flatness of molar wear appears to be a good
man hunter-gatherers or may show some indicator of change in food or food prepara-
more extreme pattern. Measurement of a cast tion. This pattern and other features of tooth
 MOLAR WEAR IN HUNTER-GATHERERSAND AGRICULTURALISTS 55

wear can be used to support dietary infer- man Evolution. Chicago: University of Chicago Press.
ences at important points in cultural and Clark, WE Le Gros (1967) Man-Apes or Ape-Men? The
technological evolution in humans, and to Story of Discoveries in Africa. New York: Holt, Rine-
hart and Winston.
illuminate changes in mastication and diet Costa, RL, and Greaves, WS (1981) Experimentally pro-
in earliest hominids. duced tooth wear facets and the direction of jaw mo-
tion. J. Paleontol. 55t635-638.
ACKNOWLEDGMENTS Crompton, AW, and Hiiemae, K (1970) Molar occlusion
I would like to thank Professors Peter An- and mandibular movements during occlusion in the
American opossum Didelphis marsupialis. Zool. J.
drews, Yves Coppens, James Garlick, Ole Linn. SOC.49:21-47.
Vagn Nielsen, Jean-Louis Heim, Charles Dart, RA (1929) A note on the Taungs skull. S. Afr. J.
Merbs, George Romanes, Erik Trinkaus, Sci. 26~648-658.
Christy Turner, Kenneth Turner, and Doug- Dempster, WT, Adams, WJ, and Duddles, RA (1963) Ar-
las Ubelaker for their kind permission to rangement in the jaws of the roots of the teeth. J. Am.
Dent. Assoc. 67:779-801.
study collections in their care. C. Loring Duckworth, LHD (1898-1899)Notes on the anthropolog-
Brace contributed both ideas and commen- ical collection in the Museum of Human Anatomy,
tary, and John Speth, Stanley Garn, and Al- Cambridge. J. Anat. Physiol. (Proc. Anat. SOC.G. Br.
phonse Burdi improved the argument. I and Ireland) 33:xxiv-x1.
thank Raymond Costa, Philip Gingerich, Ferembach, D (1974a) Les hommes de 1’Epipaleolithique
et du Mesolithique de la France et du nord-ouest du
Karen Rosenberg, and two anonymous re- bassin Mediterraneen. Bull. Mem. Soc. Anthropol.
viewers for comments on the manuscript. Paris, Ser. 13,2201-236.
Karen Klitz drew the illustrations. This proj- Ferembach, D (197413)Le squelette Epicastelnovien de la
ect was supported by grant BNS-7921733 Baume de Montclus (Gard).Bull. Mem. SOC.Anthropol.
from the National Science Foundation. Paris, Ser. 13, 1:109-127.
Gantt, DG (1979) Patterns of dental wear and the role of
LITERATURE CITED the canine in Cercopithecinae. Am. J. Phys. Anthropol.
51:353-360.
Ahlgren, J (1976) Masticatory movements in man. In DJ Green, CJS (1974) Interim report on excavations at
Anderson and B Matthews (eds): Mastication. Bristol: Poundbury, Dorchester, 1973. Proc. Dorset Nat. Hist.
John Wright and Sons, pp. 119-130. Archaeol. SOC.9597-100.
Black, GV (1895) An investigation of the physical char- Greene, DL, Ewing, GH, and Armelagos, GJ (1967) Den-
acters of the human teeth. Dent. Cosmos 37:469-484. tition of a Mesolithic population from Wadi Halfa, Su-
Brace, CL (1962) Cultural factors in the evolution of the dan. Am. J. Phys. Anthropol. 27r47-56.
human dentition. In MFA Montagu (ed): Culture and Gregory, WK (1922) The Origin and Evolution of the
the Evolution of Man. New York: Oxford University Human Dentition. Baltimore: Williams and Wilkins.
Press, pp. 343-354. Gregory, WK, Hellman, M, and Lewis, GE (1938) Fossil
Brash, JC, Layard, D, and Young, M (1935) The Anglo- anthropoids of the Yale-Cambridge India expedition
Saxon skulls from Bidford-on-Avon,Warwickshire and of 1935. Carnegie Inst. Wash. Publ. no. 495, pp. 1-27.
Burwell, Cambridgeshire, with a cornparision of their Grine, FE (1981) Trophic differences between ’gracile’
principal characters and those of the Anglo-Saxon and ’robust’ australopithecines: A scanning electron
skulls in London Museums. Biometrika 27r388-407. microscope analysis of occlusal events. S. Afr. J. Sci.
Broca, P (1879) Instructions relative a l’etude anthropo- 77:203-230.
logique du systkme dentaire. Bull. SOC.Anthropol. Hall, RL (1977) Functional relationships between dental
Paris 2: 128-152. attrition and the helicoidal plane. Am. J. Phys. An-
Butler, PM (1952) The milk-molars of Perissodactyla, thropol. 45t60-76.
with remarks on molar occlusion. Proc. Zool. SOC.Lond. Hardwick, JL (1960) The incidence and distribution of
121:777-817. caries throughout the ages in relation to the English-
Butler, PM (1973) Molar wear facets of early Tertiary man’s food. Br. Dent. J. 108r9-17.
North American primates. In MR Zingeser (ed): Sym- Hayes, AC, Young, JN, and Warren, AH (1981) Excava-
posia of the Fourth International Congress of Prima- tion of Mound 7, Gran Quivira National Monument,
tology, Vol. 111. Basel: Karger, pp. 1-27. New Mexico. Publications in Archaeology Volume 16.
Butler, RJ (1972) Age-related variability in occlusal wear Washington DC: US Department of the Interior, Na-
planes. Am. J. Phys. Anthropol. 36t381-390. tional Park Service.
Campbell, TD (1925) Dentition and Palate of the Austra- Hiiemae, KM (1976) Masticatory movements in primi-
lian Aboriginal. Adelaide: The Hassell Press. tive mammals. In DJ Anderson and B Matthews (eds):
Campbell, TD (1939) Food, food values and food habits of Mastication. Bristol: John Wright and Sons, pp.
the Australian aborigines in relation to their dental 105-118.
conditions (parts I-V). Aust. J. Dent. 43t2-15, 45-55, Hiiemae, KM, and Kay, RF (1973) Evolutionary trends
73-87,141-146,177-199. in the dynamics of primate mastication. In MR Zinge-
Carlson, DS (1976)Temporal variation in prehistoric Nu- ser (ed): Symposia of the Fourth International Con-
bian crania. Am. J. Phys. Anthropol. 45:467-484. gress of Primatology, Vol 111. Basel: Karger, pp. 28-64.
Carlson, DS, and Van Gerven, DP (1979) Diffusion, bio- Hinton, R J (1981) Form and patterning of anterior tooth
logical determinism and biocultural adaptation in the wear among aboriginal human groups. Am. J. Phys.
Nubian corridor. Am. Anthropol. 81r561-580. Anthropol. 54:555-564.
Clark, WE Le Gros (1955) The Fossil Evidence for Hu- Jenness, D (1922)The life of the Copper Eskimos. Report
56 B.H. SMITH

of the Canadian Arctic Expedition 1913-1918, Vol. 12. Hitchcock.

Ottawa: FA Acland, pp. 5a-269a. Prunieres, D (1873) Sm les objets d‘industrie contenus
Jolly, CJ (1970)The seed-eaters: A new model of hominid dans les megalithes de la Lozere. Mat. 1873r360.
differentiation based on a baboon analogy. Man 51- Rabkin, S (1943) Dental conditions among some prehis-
26. toric Indians of Kentucky. J. Dent. Res. 22r355-371.
Keith, A (1913) Problems relating to the teeth of the Reed, CA (1977) Origins of agriculture: Discussion and
earlier forms of prehistoric man. Proc. R. SOC.Med. some conclusions. In CA Reed (ed): Origins of Agricul-
(Odont. Sec.)6r103-124. ture. The Hague: Mouton, pp. 879-953.
Keith, A, and Knowles, FHS (1911-1912)A description of Robinson, JT (1954) Prehominid dentition and hominid
teeth of palaeolithic man from Jersey. J. Anat. Physloi. evolution. Evolution 8t324-334.
46r12-27. Robinson, JT (1956)Dentition of the Australopithecinae.
Kryzaniak, L (1981) Origin and early development of Transvaal Museum Memoir no. 9. Pretoria: Transvaal
food-producing cultures in northeastern Africa. Curr. Museum.
Anthropol. 22t693-694. Schoeninger,MJ (1982)Diet and the evolution of modern
Leigh, RW (1925) Dental pathology of Indian tribes of human form in the Middle East. Am. J. Phys. Anthro-
varied environmental and food conditions. Am. J. Phys. pol. 58:37-52.
Anthropol. 8: 179-199. Schultz, AH (1925) Studies on the evolution of human
Merbs, CF (1968)An archaeological study of the Eskimo teeth. Dent. Cosmos 67:935-947,1053-1063.
Thule culture in the northwest Hudson Bay area. Natl. Simons. EL (1968) A source for dental comuarison of
Geog. SOC.Res. Rep. 1968r247-254. Rumapithecus with Austrulopithecus and Homo. S. Afr.
Mills, JRE (1955)Ideal dental occlusion in the primates. J. Sci. 64t92-112.
Dent. Pract. 6t47-61. Simons, EL (1976) The nature of the transition in the
Molnar, S (1972) Tooth wear and culture: A survey of dental mechanism from pongids to hominids. J. Hum.
tooth functions among some prehistoric populations. Evol. 5.511-528.
Curr. Anthropol. 34:175-190. Smith, BH (1983a) Dental Attrition in Agriculturalists
Moore, WJ, and Corbett, ME (1978) Dental caries expe- and Hunter-Gatherers. Unpublished Ph.D. disserta-
rience in man: Historical, anthropological and cultural tion, The University of Michigan, Ann Arbor.
diet-caries relationships, the English experience. In Smith, BH (1983b) Hominid attrition and occlusion at
NH Rowe (ed):Diet, Nutrition, and Dental Caries. Ann third molars: Understanding the helicoidal plane. Am.
Arbor: The University of Michigan School of Dentistry J. Phys. Anthropol. 60r255 (abstract).
and The Dental Research Institute, pp. 3-19. Sokal, RR, and Rohlf, FJ (1969)Biometry. San Francisco:
Moses, CH (1946) Studies of wear, arrangement and oc- WH Freeman and Company.
clusion of the dentitions of humans and animals and Taylor, RMS (1963) Cause and effect of wear of teeth:
their relationship to orthodontia, periodontia, and Further nonmetrical studies of the teeth and palate in
prosthodontia. Dent. Items Interest 68:953-999. Moriori and Maori skulls. Acta Anat. 53t97-157.
Murphy, TR (1959a) The changing pattern of dentine Thalbitzer, W (1914) The Ammassalik Eskimo: Contri-
exposure in human tooth attrition. Am. J. Phys. An- butions to the Ethnology of the East Greenland Na-
thropol. 17: 167-178. tives. Copenhagen: Biancohuno.
Murphy, TR (1959b) Gradients of dentine exposure in Turner, CG (1979) Dental anthropological indications of
human tooth attrition. Am. J. Phys. Anthropol. 17r179- agriculture among the Jomon people of central Japan:
185. X. Peopling of the Pacific. Am. J. Phys. Anthropol.
Nicholls, B (1914)A contribution to the study of the teeth 51:619-636.
of the Australian and Tasmanian aboriginals. Trans- Walker, AC, Hoeck, H, and Perez, LM (1978) Microwear
actions of the Slxth International Dental Congress. of mammalian teeth as an indicator of diet. Science
London: The Committee of Organization, pp. 67-102. 20I:908-910.
Nielson, OV (1970) Human Remains: Metrical and Non- Webb, WS, and DeJarnette, DL (1942)An archaeological
metrical Variations. Scandinavian Joint Expedition to survey of Pickwick Basin in the adjacent portions of
Nubia, Vol. 9. Copenhagen: Andelsbogtrykkeriet. the states of Alabama, Mississippi and Tennessee. Bull.
Oakley, KP, Campbell, BG, and Molleson, TI (1971)Cat- U.S. Bur. Am. Ethnol. 129:l-335.
alogue of Fossil Hominids, Part 11: Europe. London: Webb, WS, and Wilder, CG (1951) An Archaeological
British Museum (National History). Survey of the Guntersville Basin on the Tennessee
Pedersen, PO (1949)The East Greenland Eskimo Denti- River in Northern Alabama. Lexington: University of
tion. Meddelelser om Gronland, Vol 142, no. 3. Copen- Kentucky Press.
hagen: CA Reitzel. Wendorf, F (1968)Late Paleolithic sites in Egyptian Nu-
Pequart, M, and Pequart, SJ (1934)La necropole mesolit- bia. In F Wendorf(ed): The Prehistory of Nubia, Vol. ll.
hique de l’ile d’Hoedic (Morbihan). L’Anthropologie Dallas: Southern Methodist University Press, pp. 791-
44:l-20. 953.
Pequart, M, Pequart, SJ, Boule, M, and Vallois, H (1937) Wheeler, REM (1940) Maiden Castle, Dorset. Reports of
Teviec: Station-Necropole Mesolithique du Morbihan. the Research Committee of the Society of Antiquaries
Paris: Institut de Paleontologie Humaine. of London no. 12. Oxford: Oxford University Press.
Pickerill, HP (1923)The Prevention of Dental Caries and Zuckerman, S (1954) Correlation of change in the evolu-
Oral Sepsis. 3rd edition. London: Balliere, Tindall and tion of higher primates. In J Huxley, AC Hardy, and
cox. EB Ford (eds): Evolution as a Process. London: George
Poncins, G de (1941) Kabloona. New York Reynal and Allen and Unwin, pp. 300-353.