CHAPTERS IN THE PREHISTORY
OF EASTERN ARIZONA, III

PAUL S. MARTIN

WILLIAM A. LONGACRE

JAMES N. HILL

FIELDIANA: ANTHROPOLOGY

VOLUME 57

Published by

FIELD MUSEUM OF NATURAL HISTORY

JULY 21, 1967

FIELDIANA: ANTHROPOLOGY

A Continuation of the
ANTHROPOLOGICAL SERIES

of

FIELD MI/SLUM OF NATURAL HISTORY

VOLUME 57

FIELD MUSEUM OF NATURAL HISTORY
CHICAGO, U.S.A.
1967

CHAPTERS IN THE PREHISTORY

OF EASTERN ARIZONA. Ill

UTAH

COLORADO

NEW
MEXICO

SILVER CITY

I MEXICO

Map showing eastern Arizona and western New Mexico.

CHAPTERS IN THE PREHISTORY
OF EASTERN ARIZONA. Ill

PAUL S. MARTIN

Chief Curator Emeritus, Anthropology, Field Museum of Natural History

WILLIAM A. LONGACRE

Assistant Professor of Anthropology, Uni Arizona

JAMES X. HILL

Assistant Professor of Anthropolo^ y, ( University of California, at Los Angeles

FIELDIANA : ANTHROPOLOGY

VOITMI 57

Publish. 7 I')

FIELD MUSEUM OF NATURAL HISTORY

JULY 21, 1907

Library of Congress Catalog Card Number: 62-21153

PRINTED IN THE UNITED STATES OF AMERICA
BY FIELD MUSEUM PRESS

Contents

PAGE

List of Illustrations

[.Introduction 11

/' V Martin

1 i>cation 11

Surroundings 12

Climate 12

Geology 12

Appearance of Ruin Before Excavation 13

Method of Excavation 13

Evolution of Pueblo Architecture 14

Problems 15

II. Description of Architectural Details 16

Paul S. Martin

Secular Rooms 16

Walls 16

Foundations 16

Construction 16

Types of Masonry 16

Unhanded Rubble 16

Banded 20

Dimensions 20

Materials Used 20

Surfaces 20

Spalls 20

Mortar 20

Plaster 20

Doorways 20

Ventilators in Rooms . . 23

Floors 23

Materials 23

Alterations 23

Bins (storage, mealing) 25

Firepits 25

Ceilings 26

Height ' 26

Construction 26

General 28

Dimensions of Rooms 28

Number of Stories 28

Deposits 28

3

CONTENTS

PAGE

Burned Rooms 30

Functions of Rooms 30

Nuclear or "Core" Units 32

Alterations 36

Plaza 39

Outlying or Detached Rooms 39

Kivas 39

General 39

Surface Kivas — Rooms 4-6 40

Surface Kivas — Room 29 46

Subterranean Kiva — Under Room 41 46

Shape 46

Fill 46

Size 46

Depth 46

Walls 46

Floor 46

Firepit 46

Deflector 49

Bench 49

Ventilator 49

Niches 49

Pits 50

Entrance 50

Postholes 50

Roof 50

Secondary Ventilator 50

General Remarks 52

Northwest Plaza Kiva 54

Location 54

Fill 54

Shape 54

Orientation 54

Size 54

Depth of Floor 54

Walls 54

Floor 54

Bench 54

Pilasters 54

Firepit 54

Ashpit 54

Deflector 54

Ventilator 54

Cupboard 54

Niches 55

Pits in Floor 55

Postholes 55

Roof 55

Entrance 55

General Remarks 55

CONTENTS 5

PAGE

III Artifacts 56

William A. Longacre

Ground .m<l Pecked Stone 58

Manos 58

Metates 61

Mortars l>-

Pestles 64

I'. nut Palettes 64

Polishing Stone 66

Rubbing Stone 68

Hammerstones 68

Worked Slabs 71

Axes 71

Mauls 72

Arrow-Shaft Tools ~2

Simple Grooved Abraders 75

Stone Pipes 75

Chipped Stone

Projectile Points

Drills 81

Saws 82

Gravers 84

I tilized Flakes 85

Flake Knives 85

Blades 86

Knives 89

Scrapers 90

Side Scrapers 92

End Scrapers 92

Ovate Scrapers 94

Shapes of Scrapers and Knives 94

Choppers 99

Cores 99

Miscellaneous Stone Artifacts 101

Bone, Antler and Shell Artifacts 104

Bone Awl 104

Bone Needles 108

Weaving Tools 108

Miscellaneous Bone Artifacts 108

Miscellaneous Shell Artifacts 108

Antler Flakes 108

Antler Wrench 108

Ceremonial Items, Ornaments, Pigments and Minerals Ill

Pendants Ill

Beads 112

Shell Bracelets 112

Pigments 115

Turquoise .... 115

Asbestos 115

6 CONTENTS

PAGE

Bone Rings 115

Bone Tinklers 115

Painted Bone Items 115

Miniature Vessels, Worked Sherds, and other Ceramic Artifacts . . 118

Miniature Pitchers 118

Miniature Jars 118

Miniature Ladles 118

Worked Sherds 118

Sherd Ladles 121

Miscellaneous Sherd Tool 121

Cordage 122

Lithic Waste 122

IV. Pottery .... 1 26

Paul S. Martin

V. Dating of Broken K Pueblo 139

Paul S, Martin

VI. The Problem of Sampling 145

James N. Hill

Introduction 145

Theory 145

Sampling at Broken K Pueblo 151

VII. Structure, Function and Change at Broken K. Pueblo ... 158
James JV. Hill

VIII. A brief Appraisal 168

Paul S. Martin

Bibliography 172

Index 177

List of Illustrations

Map showing location of Broken K Pueblo, Arizona . . . . Frontispiece

Text Figures

PAGE

1. Map of Broken K Pueblo, East Central Arizona 18 19

2. Panoramic view c > t Broken K Site 21

3. Masonry, South wall, room 11 21

4. Masonry, Wesl wall, room 11 22

5. Masonry, East wall of room 50 22

6. Masonry, South wall, room 39 23

7. Scaled doorway, room 11 24

8. Scaled doorway, room 54 25

9. Sealed doorway, room (>8 26

10. Mealing bins, room 92 27

1 1 . Firepit, round, room 60 28

12. Hexagonal firepit in coiner of room 5 29

13. Slab-lined fire pit, room 2 31

14. Slab-lined rectangular fire pit, room 7 31

15. Room 2^ 32

16. Room 24 33

17. Room 62 34

18. Room 79 35

19. Room 80 35

20. Room 53 ... . 36

21. Room 69 37

22. Room 61 37

23. Earlier rooms (about A.l). 1150), some of which were occupied throughout

the life of the Pueblo 38

24. Eater rooms, most of which were occupied until abandonment .... 38

25. Room 49 39

26. Room 64 40

27. Rooms 41 and 43 41

28. Room ^8 42

29. Plaza, fire pit 7 43

30. Plaza, fire pit 4 44

31. Plaza, lire pi t^ 1 and 2 45

32. Outlying room 2 47

33. Outlying mom 1 48

34. Surface kiva (room 6) 49

35. Surface kiva (room 29) 51

7

8 LIST OF ILLUSTRATIONS

PAGE

36. Kiva under room 41 52

37. Kiva under room 41 ; detail of East end 53

38. One hand mano, single grinding surface 57

39. One hand mano, two grinding surfaces 58

40. Two hand mano, single grinding surface 59

41. Two hand mano, two grinding surfaces with finger grooves 60

42. Beveled type 60

43. Two hand mano, beveled type with three grinding surfaces .... 61

44. Mortars 63

45. Paint palettes 65

46. Pestles 66

47. Hammerstones, chopper and block core 67

48. Stone axes 69

49. Mauls 70

50. Arrowshaft tools 73

51. Simple grooved abraders and polished and incised arrowshaft tools . . 74

52. Miscellaneous stone artifacts 76

53. Projectile points 77

54. Projectile points 79

55. Projectile points 80

56. Drills and punches 82

57. Knives 83

58. Saws and notched side scrapers 84

59. Graving tools 85

60. Distribution of dimensions of utilized flakes 86

61. Flake knives made from prepared cores 87

62. Flake knives 88

63. Distribution of dimensions of random-flake-knives 89

64. Blades 90

65. Side scrapers 91

66. Distribution of dimensions of side-scrapers 92

67. End scrapers 93

68. Distribution of dimensions of end-scrapers 94

69. Ovate scrapers 95

70. Ovate scrapers 96

71. Distribution of dimensions of ovate scrapers 97

72. Cores 100

73. Miscellaneous stone artifacts 102

74. Miscellaneous stone, bone, and shell artifacts 103

75. Bone awls 105

76. Bone awls 106

77. Bone awls 107

78. Miscellaneous bone artifacts 109

79. Antler flakers and antler wrenches 110

80. Pendants 112

81. Ornaments of bone and shell 113

82. Pigment and turquoise 113

83. Bone ring material 114

84. Bone tinklers, notched sherd and sherd ladle 116

LIS I OF ILLUS1 RATIONS 9

PAGE

85. Painted bone 117

86. Miniature vessels II'-1

87. Worked sherds 120

88. Sherd-scrapers 121

89. Bowl. Cat. No. 282877. Snowflake black-on-white: Ha) Hollow variety;

associated with burial "1 12"7

90. Bowl. Cat. No. 284074, Springerville polychrome; room 92; floor » 1; bin 128

91. Bowl. Cat. No. 282971, St. Johns polychrome; room 48; level A ... 128
Bowl. Cat. No. 282974, St. Johns polychrome; room 53; floor # 1 . . 129

93. Bowl. Cat. No. 283476, St. Johns polychrome; room 69; level B . . . 130

94. Bowl. Cat. No. 283477, Pinto polychrome; room 69; level B . . . 130
Howl. Cat. No. 282891, Querino polychrome. Associated with burial # 1 . 131
"Canteen." ('at. No. 284075, Snowflake black-on-white; Carterville va-
riety; room 80, floor # 1 131

97. Jar. Cat. No. 282879, Snowflake black-on-white; Snowflake varietv; burial

= 1 132

98. Jar. Cat. No. 283475, Snowflake black-on-white; Carterville variety;

room 69; levels B, C, and floor § 1 ; parts of the rim from room 4, level B 133

99. Jar. Cat. No. 282970, Snowflake black-on-white; Hay Hollow variety;

room 69, East wall trench 1 34

100. Jar. Cat. No. 282868, Snowflake Black-on- white; Snowflake variety;

room 41; floor « 1; Pit A 134

101. Jar. Cat. No. 284076, Snowflake black-on-white; Snowflake variety; room

39; level B 135

102. Jar. Cat. No. 283245, Snowflake black-on-white; Snowflake variety; room

33, pit O 135

103. Jar. Cat. No. 283244, Snowflake black-on-white; Snowflake variety; room

33; floor § 1, pit O 136

104. "Seed" Jar. Cat. No. 282887, Snowflake black-on-white; Snowflake va-

riety; burial § 1 137

105. Chart showing series of radio-carbon dates 140

106. The locations of residence units and their sub-units 162

107. Distribution of Factor Number 1 of the factor analysis of pottery types;

from floors of rooms, only 163

List of Tables

PAGE

1. Shapes and significance of variation in five tool-types from Broken K. Pueblo 98

2. 1 requencies of lithic waste 123

3. Mean frequencies of lithic waste per room 124

4. Totals of sherds from all rooms (fill and floors), plaza, features outside the

pueblo proper, and from four outlying rooms 136

5. Table of Carbon 14 and Tree-Ring Data 141-142

I >istribution of stylistic catagories used in isolating the loci of residence units 165

I. Introduction

Location

Excavations in I'U,} were undertaken at Broken K Site, about 10 miles
east of Snowfiake, Arizona (Frontispiece). The ruin is on the ranch of
Mi. fames ( larter whose brand mark we adopted for the name of the site.
It is located near the banks of Hay Hollow Wash, a tributary of Silvei
Creek and of the Little Colorado River, in the southwest quarter of the
northeast quarter of Section 8, Township 13 North, Range 2} East, Gila
and Salt River Meridian; in longitude 109°33' X. The elevation is aboul
5760 feet above sea level (pocket altimeter).

The research reported herein was made possible by a grant from the
National Science Foundation (G22028) and by financial assistance from
the field Museum.

Statistical analyses, including cost of computer time, (Chapters III,
VII) were paid for by a grant (1524) that the Wenner-Gren Foundation
for Anthropological Research made to the Museum. Without this support,
the statistical handling of the data would not have been done. We are
deeply grateful for this assistance.

financial aid from Mr. C. E. Gurley, Gallup, New Mexico: Mr. Max-
well Hahn, Scarsdalc, New York; Mr. Fd Alexander and Mr. Judd
Sackheim, Chicago, is acknowledged with thanks.

The work reported here was conducted between 1 July and 2d Sep-
tember, 1963, although preliminary digging was done in September.
1962, under the supervision of Dr. John B. Rinaldo. Excavations were
supervised by John M. Fritz and James N. Hill, doctoral students. Uni-
versity of Chicago. They were assisted by student-trainees: Dennis DeWitt.
Tulane University; David Harper, Rocky Mountain School: Neb John-
sou. 1 enger High School, Chicago; and Paul Boyer, University of Illinois.
Students receiving compensation were: Tom Marks, Highland Park High
Si hool; Anna Rose and Mark Sconce. Antioch College, Yellow Springs,
Ohio.

In addition to the students, we employed several men who have dim
for us in other seasons: James BrinkerhofT, I ,ibor Garcia, Celedonio ( rriego,
Elias Padilla, Guillermo Padilla, Gilbert Padilla, Fee Padilla. Porfirio
Padilla, Ned Saiz, and Joe Velasquez.

ll

12 PREHISTORY OF EASTERN ARIZONA, III

All of these people, under the able direction of Messrs. Fritz and Hill,
dug with vigor, interest, and intelligent cooperation. They followed natural
levels, sifted, and trowelled with care. A tremendous amount of dirt was
moved. I am happy to thank those co-workers and to salute them for a
job well done.

Surroundings

The region about the pueblo of Broken K is a gently rolling plateau
broken just east of the site by Point of the Mountain, a tongue of lava
that rises about 500 feet above the plain. The White Mountains lie about
40 miles to the south.

The surface of the plateau-mesa near the site has been badly eroded
by overgrazing and high winds. Arroyos, dry most of the time, become
swift-running destructive streams after rains. Junipers occur but are
stunted; grass is sparse and chenopods and amaranths abound. In brief,
the area is bleak, dreary, and uninviting.

Climate

Annual precipitation is meager and is probably the same as that which
falls on Snowflake, Arizona, where weather records have been kept since
1898.

The mean annual precipitation for Snowflake is about 12 inches; the
mean annual temperature is 51.4° F.; with a mean annual maximum of
69.5° F. and a mean annual minimum of 33.2° F. Temperatures range
from 102° to a minus 24° F. The average number of frost-free days is 133,
eight years of which (between 1898 and 1930) produced a growing season
of less than 120 days. The heaviest rainfall (1898-1930) is in July (2.39
inches average) and in August (2.57 inches average).

To obtain a crop of corn, one must consider not only the frost-free
days but also such items as night temperatures and ground moisture in
May and June and the time of the onset of summer rains. I would judge
from the weather records at hand that the average conditions today are
unfavorable for growing corn.

If the same conditions existed during the occupancy of Broken K Puebo,
or during any part of its life span, the Indians may have been unable to
produce a crop every year; and if several consecutive dry years occurred,
their supply of viable seed corn may have diminished and eventually
have been reduced to the danger point.

Geology
The geological features of the area have not been intensively studied.
The sandstone underlying and near the ruin is of the Moenkopi formation,

I.\I KOIH CI IO\ 13

of the Triassic period. About a mile to the east lies .1 great tongue of lava
il>. 1^. ih) that juts northwestward from the main lava field. This tongue
overlies rocks belonging to the Chinle formation, also Triassic. The site
rests on 1 >edrock of sandstone.

Ai'i'i trance of Ruin Before Excavation

Broken K site was discovered l>\ Dr. William A. Longacre during the
course of our archaeological survey in 1961 1 %2 (Martin, Rinaldo,
Longacre, et al., 1962). It is composed oi lour house-blocks or groups of
contiguous rooms that are organized in the form of a large rectangle en-
closing an open pla/a. Entrance to the plaza was by means of a passage-
way loeated in the southeast corner of the pueblo. Nowhere were the
mounds making up the four-room blocks more than three or four feet
above the surrounding terrain.

Method of Excavation

I did not have enough money to dig the entire ruin. It was decided,
therefore, to utilize a systematic technique of sampling in order to obtain
a statistically random sample.

We first exposed all principal walls and crosswalks of all rooms. Ninety-
two rooms were thus outlined and assigned consecutive numbers. No
circular depressions were noted.

We decided that we should excavate at least 50 per cent of the rooms
in order to obtain an adequate sample. To obtain a random sample we
selected 46 numbers from a table of random digits; or, in other words,
the 46 rooms that were destined to be dug. (For complete discussion of
sampling:, see Chap. VI.)

Data recovered in this manner allow: confidence limits of greater than
90 per cent for ensuring a representative sample of the universe of rooms
and artifacts. Dirt was removed by natural levels and all culturally sig-
nificant levels were screened. Everything, except dirt and stray pebbles,
was saved, processed, catalogued, and tabulated. This included: lithic
waste (stone chips with and without "use" flaking), tools of stone and bone,
animal bones, beads, potsherds, wood, charcoal, seeds, nuts, and human
dun?. Only cultural debris found directly on the floors of rooms were
called '"floor materials."

In addition to the basic sample (46 rooms), we dug seven more pueblo
rooms, two subterranean kivas, and three outlying rooms, making a total
of 58 units. During the course of excavations we encountered two other
subterranean kivas (one under room 22 and one under room 3()); several

14 PREHISTORY OF EASTERN ARIZONA, III

pit houses that may be dated as Forestdale Phase (circa A.D. 600-800)
(Haury, 1940); and in the southwest corner of the plaza, a chamber cut
into bed rock. The function of the latter structure is unknown, although
it resembles an unfinished Kiva.

Evolution of Pueblo Architecture

The quadrilateral arrangement of rooms with plaza represents the
final step in a long tradition of building sequences that we have observed.
Earlier pueblos in the Pine Lawn-Reserve areas, as well as in the present
area, usually consisted of four to ten rooms strung out along one axis.
Later pueblos were constructed in the shape of a squarish U, often with
a kiva or kivas located between the arms of the U. Still later pueblos
clung to the U pattern but many more rooms were crammed into this
honey-comb-like arrangement. The culmination of this tradition is rep-
resented in the Broken K Site where the U becomes a rectangle with a
small entry-way at one corner.

The rise or invention of a row of contiguous rooms, with walls of poles
and brush and later of cut and shaped stones was a new combination or
synthesis of elements in the inter-active development of culture. The oc-
currence of a communal house of several rooms took place in the South-
west because the circumstances of culture growth and history in the early
centuries of the Christian era brought together the elements necessary to
this type of building in the Southwest. The pueblo building was an adap-
tive response to the total social and ecological environment.

It is possible that this evolutionary aspect of pueblo architecture (that
is, from pit houses to surface, contiguous stone-walled clusters of rooms
arranged in a cellular fashion), may be the result of indirect interaction
with the higher cultures to the south (the highlands of Mexico) in re-
sponse to a need. Then, this house building was pushed to new and
different evolutionary heights that bore no resemblances to the borrowed
prototype. In other words, the borrowed architectural form was reworked,
modified to fit the social and ecological milieu, and adapted to the par-
ticular needs of Southwestern cultural systems.

On the other hand, it must be admitted that this form of architectural
adaptation to Southwestern cultural environment may have been an in-
dependent development. A rather good case could be made for this latter
point of view but actually not enough evidence is at hand at the moment
to pass judgment on these matters.

The total social and ecological environment that produced the cultural
advance from pit houses to complex aggregates of rooms formally arranged
around a plaza is not known with certainty. But there is no doubt that

INI R< )Dl (1 ION 15

pueblo architecture was a technological discover) brought about l>\ the
fact that the culture processes had reached a point where such a discover)
was possible and probably inevitable.

If we view culture as an organized, integrated system composed of
aspects or subdivisions, we can distinguish three sub-systems: technologi-
cal, sociological and ideological. They are inter-related but the influence
of tli is reciprocal interaction is nol equal. Technology plays the primary
role

Architecture must be viewed in two cultural contexts: technomic, for
it deals directly with environmental problems, and socio-technic, for it
reflects the social relationships of the culture. In the case of pueblo houses,
seen technomicall) . the builders probably discovered that houses with
contiguous walls of stone were more efficient and easier to build, gave
better protection from winds and cold weather, served as better places
for storing and preserving fooch, and were- useful as protection from un-
friendly peoples. Within this technomic framework, architecture was
probably modified to lit the sociological structure which was evolving,
hut. remembering that culture is a system of interdependant variables
(technology, sociology, ideolo<^\ ). there can be no doubt that the sociology
was also modified to fit the architectural changes. The question of the
extent of these sociological changes is moot.

Thus we abandon the explanations that invoke "coincidence*", the
"achievement of a genius'*, ""migrations** or other whimsical ideas. When
certain factors and conditions are present and in conjunction, an inven-
tion or discovery takes place; when they are not present, the invention or
discovery does not take place. (White, 1949, p. 209).

Hence, I can say that the invention of pueblo architecture (contiguous
rooms, stone walls, several terraced stories) came about because the right
ecological, technological, economic, and sociological factors and condi-
tions were present and in association or conjunction.

PrOBLI Ms

One of the prime problems of the research on Broken K was to dem-
onstrate that archaeology can be raised to the explanatorv level by using
recent advances in methodology and by adopting a more useful concept
of culture. We focused our attention on cultural structure and processes
instead of on taxonomy and trait lists. We have combined some tradi-
tional methods with newer sampling procedures and with data-processing
by computer.

Hence, we were determined to make a contribution to the- growing
corpus of anthropological knowledge and theory; to augment our knowl-

16 PREHISTORY OF EASTERN ARIZONA, III

edge of some of the cultural processes that took place in the prehistoric
Southwest; and to demonstrate that the complex inter-relationships of
people, places, and artifacts of prehistoric sites can be better understood
through the use of diverse forms of statistical analyses.

We expected to make several primary contributions to knowledge:

1. a unique sociological analysis of a prehistoric site in the Southwest;

2. an important gain in our knowledge of the origins and causes of
contemporary social organization of the Pueblo Indians;

3. a demonstration of the use of the powerful tools provided by statis-
tical techniques and data-processing computers;

4. the value of a number of other research techniques for implementing
such goals as have been outlined here — notably involving the utilization
of other disciplines, such as palynology (Botany) and Zoology.

Note: This Volume includes the traditional treatment and descriptions of
architecture, and bone and stone tools, and a listing of pottery types with a
table showing gross frequencies by types. For interested persons these data
may be used for comparative studies.

The descriptions of the stone and bone tools by Longacre have been
tightened, clarified, and refined; and the types are treated functionally.
In addition, far-reaching and unique creative deductions have been drawn
from distributions, densities, and associations. Detailed counts, measure-
ments, and find-spots are published in Archives of Archaeology, Martin,
Longacre, and Hill, 1966, no. 27.

The companion volume by James N. Hill, (Broken K: A Prehistoric
Society in Eastern Arizona, in press) presents all of the data used in
testing the hypothesis that Broken K Pueblo represents an adaptive stage
in the evolution of contemporary Pueblo social organization. These data
include identification of (1) types of "areas" (work, storage, habitation,
ceremonial) and functions carried on there; of (2) the loci of uxorilocal
descent groups; of (3) the loci households; as well as analyses of social
organization and social processes, and of environmental deterioration as
indicated by palynological evidence. Statistical methods that were em-
ployed for perceiving associations and for testing hypotheses and infer-
ences will be given.

Paul S. Martin
June, 1964

II. Description of Architectural Details
/>') Paul S. Mari in

A. SECT LAR K< M >MS Surface
( Figs. 1 and 2)

Walls

Foundations. Walls based on reddish, sterile clay with no prepared
foundations or wall trenches except in ten rooms (rooms 1, 4, 5, 24, 33,
37-41). In tlu^e rooms, bottom course of masonry consisted of vertical
sandstone slabs set in trenches 5-20 cm. in depth. Note: Some walls with
vertical slabs in bottom course set directly on sterile clay or not in trenches.

Construction. Xo bonding at corners. Through-stone masonry with
stones reaching through from one wall surface to other side; no cores, no
composite masonry with two faces interlocked in center. Practice of
building four separate walls, all abutting on one another, for one room:
and of adding rooms without bonding, makes it difficult to follow building
sequence. Sometimes a wall will span one end and one side of a room
(room 23) and gives appearance of bonding at corner (northwest corner
of room 23).

Types of Masonry. — Masonry can be divided into two general el. isms:

(a) unbonded rubble (Figs. 3, 4); that which belongs to the core or nuclear
rooms and, in general, seems to be the earlier; consists of long and me-
dium-long thick slabs alternating with smaller slabs and spalls, all set in
thick cushion of adobe mud mortar; coursing, good to rough, apparent
on close inspection. In some rooms this class degenerates to mud-rubble
wall with large, unshaped stones, all more or less same size, set in mud
mortar with fewer spalls, and coursing not pronounced; or to walls com-
posed nt course of small slabs alternating with course of large spalls. W all
rocks vary from 25 cm. by 40 cm. to slabs 30 cm. by 20 cm. by 9 cm.;
to spalls, 5 cm. by 1" cm. Courses in latter, not maintained but appear
as a jumble of irregular stones or change in thickness; are irregular and
inconsistent. Sometimes these variations will appear in same wall.

17

Legend

▲ Datum

A

Ashpit

ma Wall

B

Bench

baa Wall abutment
Lb Wall bond

C
D

F

Pit

Deflector

Firepit

aa ■ Doorway

G

Burial

■■■■ Sealed Doorway

H

Bin

■i ■ Ventilator

K

Kiva

■■■i Sealed Ventilator

L

Loom Holes

::::: Collapsed Wall

M

Mealing Bin

r::z:: Early Wall (Removed
Earth Wall Inhabit

by
ants)

N
P

Niche
Pit -house

Postulated Earth

Wall

S

Slab

Not Visible From
'■'■ a Unexcavated

Surface

T
V

Trench
Ventilator

o Post Hole

(or Ventilator Tunnel )

• Post

W

Step

■ I '

t 2 3

J Meters

OUTLIER 1

O

BEDROCK
MORTAR

o

BEDROCK
MORTAR

CD

OUTLIER 2

Plan of BROKEN K PUEBLO, Ea

Fig. 1. Map of Broken K Pueblo, East Central Arizona.

OUTLIER 5

r

OUTLIER 4

I Arizona

OUTLIER 3

20 PREHISTORY OF EASTERN ARIZONA, III

(b) banded: found in many of later or latest rooms (Fig. 5). At its best
resembles a "Chacoan" type of wall — large horizontal slabs (30 by 60
cm.) laid directly on ground; on top of these and laid in even courses,
slabs and chips of laminate sandstone; and, alternative with the latter,
another course of large slabs (10 by 70 cm.). Stones matched in thickness
but not in length or color. Variations of this class occur: slabs (30 by 35 cm.)
set vertically and often in a prepared trench; on top of these, alternate
courses of mud and a few spalls (about 5 cm. thick), and horizontally
laid slabs (7 by 35 cm.) (South wall, room 39 [Fig. 6]). More imperfect
versions of this under "banded" class were found in north wall of room 64.

Walls not plumb, contain cracks and imperfections. In general, masonry
not impressive. It served to keep out rain, cold, heat, winds, and "var-
mints"; but was not a thing of beauty. Since rooms were only one story
high, more massive and better-built walls were not necessary.

Dimensions. — Greatest standing height of walls, 120 cm. (southeast
corner, room 54); thickness, average, 17 cm.

Materials Used. — Sandstone and mud.

Surfaces. — Rough hewn, undimpled; tooling absent.

Spalls. — Small stone flakes and small thin, flat, stone slabs used. Pot-
sherds not used. Spalls laid in mortar to level the courses, to fill voids, or
to keep wet mortar from squeezing out when next course was laid. All
spalls have one edge flush with joints.

Mortar. — A mixture of local clay mixed with sand; hard; served to
give strength and permanence to walls and to prevent stone-to-stone
contact.

Plaster. — Interiors of walls of most rooms were mud-plastered; several
layers of mud plaster were found in some rooms.

Doorways. — Thirty-six doorways located, 22 of which were sealed
(Figs. 7-9). All rectangular, located near center of the side or end walls
of rooms. No exterior doorways except those that gave on to the plaza.
Sills, of sandstone slabs, ranged from 4-40 cm. above floor level; average,

21 cm. Doorway in west wall of room 92 provided with steps 23 cm. to
70 cm.; widths averaged about 50 cm. Lintels of sandstone slabs and
sometimes wood beams; sides of doorways of masonry. In south wall,
room 49, two sealed doorways side by side, were found, making three
sealed doorways in this room. No "T" doorways. In some instances,
sealing of doorway may have occurred when addition was added. Other
reasons for sealed doorways not known.

It is assumed that rooms without doorways were entered through a
ceiling hatchway. The finding of several ring-slabs on floors confirmed
this impression.

1

6ma

jr-^ V~ *"

*»« *

-4

4

_

^ '.

m

~

A

1. Panoramic view of Broken K Site showing rooms and outside walls of north
and east wings outlined with trenches.

A.

^i*

"

• Jfc?

*v/>

Fig i. Masonry, south wall, room 11. Arrow (50 cm. long) points north: tnetei
stiek in background.

21

Fig. 4. Masonry, west wall, room 11. Arrow (50 cm. long) points north.

Fig. 5. Masonry, east wall of room 50.

22

\1<( III I 1 < II KAI. Dl.l \ll.s

23

Fig. 6. Masonry south wall, room 39. Arrow (50 cm. long) points north.

Ventilators in Rooms. — Eleven vents were located: two in room 1; one
between rooms 2 and 7; one between rooms 11 and 16; one between
rooms 47 and 49; one between rooms 60 and 64; one between rooms
78 and 79; one between rooms 78 and 75; one between rooms 80 and 81;
one in south wall of room 82 leading outside pueblo, roof corbelled
one in north wall of room 92. One was at floor level (82), others were
15-30 cm. above iloor. Dimensions ranged from: heights, 15-40 cms.;
widths, 12-20 cms. Some were placed in sealed doorways (after door-
ways were sealed?). Six of the vents were sealed.

Floors

Materials. — Sandy clay, usually hard; lavender, red. to gray in color;
rought to smooth in appearance, but not polished.

AlU Most rooms contained one floor only; eleven rooms out

of the 53 pueblo surface rooms were furnished with two floors. Distances
between these floor- varied from 3-15 cm. Most of the rooms with plural
floors were habitation rooms. It is believed that little time elapsed be-
tween the building of lust floor and the addition of the second. Usually
the later, added (adobe) floors covered previous storage pits, pits, nictate
bins, firepits and miscellaneous holes. Sometimes on the contrary, the later
floor (upper J had more features than the lower one.

Fig. 7. Sealed doorway center of north wall of room 11. Arrow (50 cm. long)
points north; meter stick in background.

24

\k( ill I I ( ilk \i Di. I AILS

25

Fig. 8. Sealed doorway in center of south wall, room 54. Arrow (50 cm. long) points
ninth: meter stick in background.

Bins (storage, mealing bins). Storage pits and or mealing bins are
found in most habitation rooms and sometimes in storage rooms. Mealing
bins occurred singly, in pairs; in one room (91) in a set of three, and in
room 92, in a set of four (Fig. 10).

Firepits. — Occurred in all habitation rooms (by definition) and some-
times in storage rooms. In latter, were often sealed over by added floor
and not used when food was stored therein. It is possible these were habi-
tation rooms converted to storage rooms.

26

PREHISTORY OF EASTERN ARIZONA, III

Three types: 1. Round or basin-shaped, earth-lined (Fig. 11); most of
these had been sealed over; were often not in center of room; and are
found only in storage rooms. Is it possible that these firepits were in exist-
ence before the rooms enclosed them or that the firepits had a different
function?

2. Hexagonal, slab-lined firepits were found in only a few rooms and
usually near a wall, off center (Fig. 12).

3. Slab-lined rectangular firepit (Figs. 13, 14), the most common type.
Some with fiat bottoms, some with concave, and some with a small sump
or pit at the bottom. Range in size from 30 by 50 cm. to 32 by 68 cm.
Depths ranged from 18 cm. to 52 cm.

Ceilings

Height. — Actual height unknown. Estimate, arrived at by inspecting
rock inside and outside rooms, six feet above floor.

Construction. — Construction or type unknown. From decayed or burned
portions of wooden beams and from impressions on burned adobe chink-
ing, believed roofs were similar to those of ancient pueblos: main beams,
smaller ones at right angle, splints and/or mats, mud.

*V

ft

Fig. 9. Sealed doorway in west wall, room 68. Arrow (50 cm. long) points north;
meter stick in background.

r

| j

M

I 10. Mealing bins, room 92. Slab-lined tin-pit, center; and older firepit, to south
oflatei one. Arrow (50 cm. long) points north; meter stick in background.

27

28

PREHISTORY OF EASTERN ARIZONA, III

' ' ',#'•'

•v

Fig. 11. Firepit, round, room 60. Manos and hammerstones found on floor. Sealed
doorway to left of meter stick. Arrow (50 cm. long) points north; meter stick in back-
ground.

GENERAL (Figs. 15-19)

Dimensions of Rooms. — Size of habitation rooms varied from 2.4 by
2.8 m. to 4.7 by 5.3 m., and are generally larger than 7.0 sq. meters in
floor area. Size of storage rooms ranged from 1.3 by 2.0 m. to 2.20 by
3.50 m., and are generally smaller than 7.0 sq. meters in floor area.

Number of Stories. — One.

Deposits. — Thirty-one rooms contained no trash; fill composed of soil,
sand, wall debris, water-laid soil, a few sherds. Twelve of these are habi-
tation rooms and 19 storage (not including outliers).

Twenty-one rooms contained trash, of which 17 were habitation units
and four storage rooms. (This count includes surface habitation rooms
later converted to kivas and rooms that may have functioned as habita-
tion and/or storage spaces).

The trash in some rooms was relatively thin, amounting to a few centi-
meters, and deposited on roof beams (room 11). A few rooms contained
more ash, sherds, and bones — but the maximum was about 42 cm. in
thickness (room 4). This does not take into account trash in subterranean
kivas or trash from Forestdale phase pit houses (Haury, 1 940) sometimes
encountered under floors of rooms (room 28).

1 12. Hexagonal tin-pit in corner of room 5. Miscellaneous masonry walls repre-
sent ventilator tunnel and shaft for surface kiva (4, 5, 6) and supports for bench thai
had been topped by Lai ge stone slabs. Arrow (50 cm. long) points north; meter stick in
Lground.

29

30 PREHISTORY OF EASTERN ARIZONA, III

When I compare the amount of trash that I have encountered in
rooms of other pueblos (room 8, Lowry Pueblo — Martin, 1936; and rooms
1A, 2A, etc., Hooper Ranch Pueblo — Martin, Rinaldo, Longacre, 1961a)
with that found in rooms at Broken K, I am surprised at the paucity of
it at this site. Several suggestions may be put forth to explain the relative
scarcity of refuse in rooms at Broken K: 1.) Much was put in the aban-
doned kivas (under rooms 22, 39, 41, northwest plaza kiva, and unfin-
ished subterranean building in the southwest plaza; 2.) trash was placed
on a midden or middens; if this be so, we never found them. Sheet-trash
was encountered on all sides of the pueblo and some in the plaza, but it
was thin (5-20 cm. in thickness). Perhaps if this volume were stacked,
it might make a respectable amount in cubic meters. 3.) trash was dumped
in the nearby wash; 4.) perhaps rate of accumulation of trash was slower
at Broken K Pueblo for some reason we do not understand. 5.) occupa-
tion of the pueblo was of short duration and vast amounts did not develop.

Burned rooms. — Six habitation rooms burned (rooms 21, 37, 53 [Fig. 20]
lower floor of 62, 78, and 80) and three more may have (rooms 20, 69, 73).
Room 19, a storage room may have burned. Several kivas show evidence
of destruction by fire — room 6, after being converted to kiva; kiva under
room 41, and northwest plaza kiva. The ceremonial room under room 22
may have burned.

It is amazing to me that relatively few rooms burned. My impression
is that many pueblos contain more evidence of conflagrations. Carter
Ranch Pueblo (Martin et al., 1964) yielded three burned rooms out of
23 dug. These rooms had really blazed and charred roof beams were
found on the floor. The fires in the rooms at Broken K Pueblo seem
"cool" by comparison. We found ash-covered, blackened floors and an
occasional piece of burned roof clay, but no masses of charred roof beams.

If habitation or storage rooms contained stored corn, one might have
hoped for a good, hot blaze — hot enough to bake the plaster on the walls
and to crack the wall stones. Is it possible little corn was on hand at the
end of the life of the pueblo; or, were the occupants very careful not to
let fires start? Insurance rates must have been low !

Functions of rooms. — Rooms at Broken K have been classed as follows:
1 . habitation (postulated on the basis of firepits, metates, grinding bins,
tools, occupation debris); 2. storage units (Fig. 22) (so classed because
they usually lack firepits, and because they are relatively smaller than
habitation rooms and contain few if any artifacts); 2a. rooms that might
first have been habitation rooms and then converted to storage or vice
versa — rooms 40, 48, 51, and 74. It should be noted that firepits in rooms
40, 48, and 51 were sealed by secondary or later floor; from which I

Fig. 13. Slab-lined tin-pit, room 2. Arrow (50 cm. long) points north: meter stick
in background.

k.

. . L_\

^ -

rfU j

1 ig. 14. Slab-lined re< tangular firepit, room 7. Scaled rectangular box to left, repre-
sents mealing bin. Arrovs (50 cm. long) points north: metei stick in background.

31

32

PREHISTORY OF EASTERN ARIZONA, III

t.f-

'w«*

Fig. 15. Room 27. Firepit in center; mealing bin east of firepit. Arrow (50 cm.
long) points north; meter stick in background.

deduce that function of rooms changed. This supposition is strengthened
by fact that after firepit in room 48 was partially dismantled, the area
was then covered by large slab and used as grinding bin or storage pit;
3. ceremonial (later stages of rooms 6, 29, and rooms under 22, 39, 41,
and in northwest corner of plaza.)

It should be added that evidence exists for supposing that roof tops
were extensively utilized for common tasks, e.g., grinding, weaving, paint-
ing pots, whittling, maintenance, and repair of tools, and gossip.

Nuclear or "core" units. — The order in which rooms were constructed
was exceedingly difficult to determine with much certainty. The chief
reason for this is that even the walls of primary rooms were rarely built
with four bonded corners. We thought ourselves lucky if we found two
adjoining walls bonded at point of contact. Abutments were the rule.

Nevertheless, prolonged study has brought forth certain conclusions
of which we feel fairly certain.

The primary rooms were (Fig. 23):
East wing: rooms 11, 12, 13
West wing: rooms 54, 55, 57

ARC III 1 l.( I I KAI. 1)1. 1 AILS

North wing: rooms 31 34

South wing: rooms 73, 74, 76, 77; and 80 and 82.

It will be noted that 1 have grouped the east and west wings. My col-
league Hill feels ili.u these were the first unit- to have been established
.it Broken K. Pueblo.

Shortly thereafter, (perhaps .i yeai oi .* fev\ years later) tin- units in

the north .mil south win1-! were built (Fig. 24).

Ai the moment, we feel that the earliest occupants li.nl .i quadrilateral
plan in mind and tint the pueblo, in that sense, was pre-planned. As the
nuclear families increased by means ol matrilocal marriages, rooms were

added as needed.

It is barely possible that all the rooms (almost 100) were occupied
simultaneously. The evidence on this point is not too clear; the available
data suggest, however, all sections were inhabited at peak or near end
oi. occupation. We visualize the pueblo as ,i living organism made up

of Cells (rooms) some of which were being created while others were

sloughed off.

I !' 16. Room 24. I hick slab-lined hrepit in center with pot-rest; mealing bin in
northwest cornei with mano "/ titu. Mealing Inn in southwest cornei with inri.it.- set
exactly vertical, held up by plastei around base, Mano in place. A now (50 em. long)
points north; tnetet stick in background.

Fig. 17. Room 62. Firepit in center. Hexagonal slab-lined firepit north of arrow;
fire-cracked stones next to it were found in pit. Arrow (50 cm. long) points north; meter
stick in background.

34

•&

Fig. 18. Room 79. Slab-lined fircpit in center. Alcove in southeast corner with
double grinding-bin north of it. .Arrow (50 cm. long) points north: meter stick in
background.

I 19. Room 80. Slab-lined fircpit in center; round pit in bottom. Doorways in
' w.ilN. Slab leaning against wall next to east door may have been used
to close door. Amow (50 cm. long) points north; meter stick in background.

33

36

PREHISTORY OF EASTERN ARIZONA, III

Fig. 20. Room 53. This room had burned. Four metates were found on floor. Two
posts were found in the walls. Shallow slab-bottomed pit next to the north wall is prob-
ably flour receptacle for a mealing bin. Unplugged doorway in the north half of the west
wall. Arrow (50 cm. long) points north; meter stick in background.

Alterations. — A few alterations will be mentioned. Rooms 5, 6, and 29
were converted from habitation rooms to kivas. Details concerning these
changes will be discussed under section on Kivas.

Rooms 31 and 33 were two separate units but were later thrown into
one. The firepit belonging to room 33 was sealed over at that time and
the extant one was built. The same order of events occurred in rooms
35 and 37.

The west wall of room 49 was placed over an earlier pi-gummi (Fig. 25).

Room 92 was at one time two rooms.

Room 64 (Fig. 26) was enlarged and the new east wall was built over
an earlier firepit.

Rooms 22, 39, and 41 were erected over earlier kivas, which were aban-
doned and filled with trash (Fig. 27).

The south wall of room 27 was built over a burial pit.
Pit houses belonging to the Forestdale phase (A.D. 600-800) were
found to have existed under rooms 33 and 48.

The kiva under room 41 lay on top of an earlier pit house floor.

I ig. 21. Room 69; burned. Firepit, with pot rests, in left (enter; ash pit adjoining
t'nepit to west. Upside down slab nutate east of directional arrow. Double mealing
bin in southwest corner. Single mealing bin near northwest corner. Arrow (50 cm.
long) points north; meter stick in background.

*,*

f

s

1

<

*

, <ft

'•,

Fig 22. Room 61. Various "storage pits" and possible post holes. Arrow (50 cm.
long) point! north; metei stick in background.

37

<■-

'-, /

/ 7
-J

r~

1

[54

1 5iJiil '

;"""' r

— -

R.

T

"M9j L — U -- -.

\

i Fl

64

ft

1-

' \ — ^^

1 /

\-"' \34\33

1 t_.

h

L__

Fig. 23. Earlier rooms (about A.D. 1150) some of which were occupied throughout
the life of the pueblo.

EL,

m

\ — \

Fig. 24. Later rooms, most of which were occupied until abandonment.

38

AK( HI I I ( I URAL DE I Ml.s

Fig. 25. Room 49. Shallow, round basin-shaped firepit in southwest quadrant.
Large slab-lined structure extending under west wall into room 48, a roasting pit. It
was sealed and antedates both rooms; contained ash, charcoal, and 30 sandstone and
basalt burned rocks. In south wall behind and to the west of meter stick is a plugged
doorw.tv. Arrow (50 cm. long) points north; meter stick in background.

I nder rooms 2 and 7 and the east wall of room 9, we found earlier.
pre-existing fin-pits.

Room 78 contained remnants of an oval-shaped wattle and daub struc-
ture (oven?) (Fig. 28). The upright sticks had burned.

PI. Contained unfinished (?) room (lava?), a bin or repository

(center), and nine firepits. Several of the latter may have been roasting
pits (Figs. 2') 31).

' I .'lying or detached rooms. — There were five of these: purpose unknown
Figs. 32-33).

B. KIYAS

G • '. Six kivas were discovered four subterranean and two sur-

fai e I "in- kivas of the six wen- excavated. In the southwest corner of the

pla/a. another depression was located. Excavation revealed a chamber
cut into bedrock. Although this depression resembles an unfinished kiva,
its function is not known.

Two of the excavated kivas were subterranean; two were converted

Surface rooms and were thus part ot the surface pueblo.

40

PREHISTORY OF EASTERN ARIZONA, III

Surface Kivas (Rooms 4, 5, 6)

The kiva, occupying spaces numbered 4, 5, and 6 on map, was not
built as a ceremonial room.

Briefly, the history of this area is: rooms 4 and 5, previous to conversion,
made up one habitation room. The slab-lined, rectangular firepit was
centered in the larger space, rooms 4 and 5, and later, was abandoned
and the east wall of room 5 was built over it. The hexagonal, slab-sided
firepit (Fig. 12) in the southwest corner of room 5 may have been an
exterior firepit used before the building of rooms 5 and 6. Room 6 had
likewise been a habitation room provided with standard rectangular fire-
pit, mealing bin, and storage pit.

When the people decided to convert the area (rooms 5 and 6) to a
kiva, the firepit and other pits in the floor (of room 6) were sealed over
by a second floor (about 5 cm. above the first one). What happened to
the occupants of former rooms 4, 5, and 6 is not known. They may have
died, they may have left voluntarily, or they may have been evicted.

The kiva (Fig. 34) was then constructed. In room 6, a secondary wall
on the south was erected and a ventilator opening southward was con-

Fig. 26. Room 64. Firepit in center with ash pit east of it. Sealed ash pit immediately
north of arrow. Note: Foundation of previous north-south wall immediately west of
meter stick and paralleling it. Arrow (50 cm. long) points north; meter stick in back-
ground.

I 2 Room* -41 and 43; room 41 lower part of picture : slab-lined firepit in

center; pit in southeast corner. It contained Snowflake Black-on-white, Snowflake

Kiva till extends under north wall; shovel standing <>n till matei ial ,iIh>\ <•

. . .i-4l . Room ■) 5 uppei slab-lined firepit in center; mealing bin on west wall.

Arrow 50 cm. long) points north; metei ^tiek in background.

41

42

PREHISTORY OF EASTERN ARIZONA, III

Fig. 28. Room 78. Mealing bin in northeast corner. Slab-lined firepit in center.
Slab near east wall probably used to close doorway in east wall. Circular structure in
the southwest part of room discovered below floor level. It is the base of a wattle and
daub structure. Purpose unknown; It may represent earlier structure. Arrow (50 cm.
long) points north; meter stick in background.

structed (similar to one in 41 kiva?). Later this was sealed. A firepit
(50 cm. deep, 29 cm. diameter) with round orifice (of stone) was con-
structed east of the earlier lower firepit. Several slabs were placed west
of the firepit; and four loom holes were constructed north of firepit —
three in slabs and one in floor. A ventilator (40 cm. wide by 10 cm. high
by 70 cm. long) was built at floor level in the east wall of what had been
room 6, with the flue shaft projecting upward in room 5. The opening
of the tunnel (on the kiva site) was provided with a slab facing pierced
by a rectangular opening. Immediately back of face-opening of this slab
a rounded (pecked) rock was found. This may have served as a damper
if needed.

Between the firepit and slab facing was the deflector (29 cm. long,
1.3 cm. thick, 7 cm. high [probably broken]) set east of firepit by 12 cm.

An "ash pit" was situated between the deflector and the ventilator
opening.

In the center of "room 5" and running north and south, a wall, com-
posed of inferior masonry interspersed with and strengthened (?) by

(

I ig. 2''. Plaza; firepit 7. No slab-lining; charcoal ami fire-cracked rock <>n flooi ol

ph. Arrow c"''1 (in. lone) pnmts north, meter stick at right.

43

Fig. 30. Plaza; firepit 4. Two levels: lower level, circular. On floor, charcoal and
fire-cracked igneous rocks.

44

\k< III I I i M R \l. hi I \ll.s

45

<

I

I ■ 31. Plaza; Brepits 1 and 2. Meter stick in pit 1. Top portion of vertical slab
siding is missing on north and west walls. East wall slumped. Vertical dividing slab
rests on horizontal worked slab which extends into both Brepits. Fire-cracked ign
rock at right of pits. Arrow (50 cm. long) points north; meter stick m background.

seven posts was built to a probable height of 70 cm. The space thus en-
closed between the east wall of room 6 and the new wall of poles and
rocks) was covered or roofed over by hum' slabs each 1.2<> in. loinj and
about _|1 cm. wide. Since the ventilator was roofed with poles and dabs
and the enclosure by the two large slabs, two hollow spates were thereby

:ated. The south one probably served as a cupboard and was closed on
the ki\a side cast wall of kiva) by a slab. The north spate was apparently
not used and was closed by masonry. In this masonry wall, a small
was placed purpose unknown.

The roofed over area described above thus provided a bench or "altar"
with slab flooring about ()0 tin. high, about 1.40 cm. deep and running
the north-south width of the room, or 2.75 ni.

•'Room 4" was filled with trash as were parts of "room 5" around
ventilator construction ^\^\ east and smith of the rear facings of the bench.

The space i reated b) the "new"" wall on the south side of room 6 and

of the west half of the north wall ol room 1 was likewise trash- fillet 1.

The walb ol room 6 were similar to those in rooms 7 and 11. The
dimensions of the ki\a were 2." m. b\ 2.8 m. The tool ol the kiva ma)

46 PREHISTORY OF EASTERN ARIZONA, III

have been about 2 m. high and was probably constructed of logs, poles,
branches, and mud. Entrance to the kiva must have been via a roof hatch.

After some months (?) or years (?) of usage, this kiva burned and some
trash was later deposited therein.

Surface Kiva (Room 29) (Fig. 35)

The secular history of room 29 before conversion to a "kiva" is not
known. To convert the room to a kiva, a bench, a ventilator opening in
the north wall and a paving of large, thin sandstone slabs were added.
After the bench or "altar" was installed, the floor area was reduced to
5.75 sq. m. (about 2.5 m. east and west; and 2.3 m. north and south).
The surface of the bench was made up of sandstone slabs; and the vertical
face was composed of stone slabs placed upright with horizontal masonry
on top of them. This kind of construction seems to occur only in late
additions and alterations.

Ventilation was obtained by means of an opening in the east face of
the bench and another opening to the outside in the north wall of the
partitioned area. We assume that the area inside the bench was hollow
and that the fresh air flowed into this area from the north opening and
thence into the kiva via the opening in the face of the bench. It is pos-
sible that a masonry or wattle-and-daub shaft connected the opening of
the north wall directly with the vent opening in the east face of the bench.
Excavation of the enclosed bench area failed to provide us with specific
information on this point.

Subterranean Kiva Under Room 41 (Figs. 36, 37)
Shape. — Rectangular.
Fill. — Ash, sherds, discarded stone tools.

Size. — 2 m. wide (north-south) by 3 m. long (east-west) including
bench. Available floor space (omitting bench) was 2.30 m. (E.-W.) by
2 m.

Depth of floor below floor room 41 and surface. — 2 m.

Walls. — Of native red clay over which lay several coats (3 mm. thick)
of sooty plaster of clay.

Floor. — Of native clay overlain by blackened clay plaster, and sloping
slightly towards firepit. The east portion of this floor was 3 cm. over floor
of earlier Forestdale phase pit house (A.D. 600-800).

Firepit. — "D" shaped, with vertical slab on flat, or east side; 35 cm.
east-west; 43 cm. north-south; 51 cm. deep. Lining of firepit of native
clay baked hard, red in color. Contained ash, charcoal, basalt, and sand-
stone rocks, sherds.

I ig. 32. Outlying room 2. Moor is bedrock. Rectangular tin-pit, slab-lined on two
sides with pot rest. I wo overlapping bowl-shaped pits (Iul; into bedrock to west of fire-
pit and later scaled. Arrow (50 cm. long) points north; metei stick in background.

47

Fig. 33. Outlier 1; floor is bedrock. Arrow (50 cm. long) points north; meter stick
in background.

48

AKl HI I I ( II R \l. hi I Ml v

49

I : J4. Surface kiva room 6 . Four loom-holes in floor on east side. Firepit, clay-
lined .unl provided with ring-slab for orifice, Easl ol firepit are deflectoi and ash-box,
A slab-faced ventilator-tunnel extended into former room 5; another slab-faced
ventilator-tunnel extended south with shaft opening in room 1. East ventilator system

was luiution.il; south system, sealed. Nictate, north ol firepit. Arrow (Si) (in.

points north, meter stick in background.

Deflector. A sandstone slab, upright; 4 cm. thick; 4(> cm. long; and
25 cm. high. Deflector placed 25 cm. cast of firepit and 90 cm. west of
\ entilator opening in bench.

Bet h. Carved or "sculptured" from aeolian (ill in earlier and aban-
doned pit house. Bench is 40 cm. high and 65 cm. deep (east-west). In
center of top of bench surface next to east wall, is a large pit, 4<> cm. by
^2 cm. and connects with ventilator tunnel. Where horizontal and ver-
tical hues of bench meet. ,i wooden beam (6 cm. thick) was placed as
edging to prevenl friable bench material from crumbling. Beam ran
entire width of bench and ninth end socketed in slot in wall.

I entilator. In center of bench hue. almost at floor level; 20 cm. wide
and 20 cm. high. < Opening to ventilator tunnel is horseshoe-shaped stone
slal>. The tunnel is 1 m. long to point where it connects with the vertical
shaft. Walls and roof ol tunnel ol sandstone slabs.

Viches. Two in number; one, centered ninth and south in west wall,
floor ol which is \§ cm. above Hour; dimensions: width. 70 cm.; height,
2 » cm.; and depth, 25 i m In it was fragmenl of a one-hand mano. I he

50 PREHISTORY OF EASTERN ARIZONA, III

second in the south wall in the southwest corner, the floor of which was
30 cm. above floor; width, 55 cm.; height, 30 cm.; depth, 33 cm. This
second niche had been sealed with sandstone slabs set in copious amounts
of adobe mortar, all of which had been neatly covered with mud plaster
3 mm. thick. This plaster coating had become blackened with soot after
which another coat of plaster (2 mm. thick) was applied to area around
sealed niche and this, in turn, became smoke blackened.

Pits. — Three in number; one, (pit G), in northeast corner of bench;
sealed with mud plaster. Contents: ash, charcoal, four flint chips and
two cores. Dimensions: diameter 18 cm. by 23 cm.; depth, 14 cm.

The second (pit H) in the southeast corner of bench; sealed with three
horizontally placed sandstone slabs and under them, a plaster plug. Con-
tents: ash, reddish-brown sand, four flint chips, and a flint saw.

The third pit (pit A) in center of top of bench next to east wall and
extending for 20 cm. under the east wall; sealed with adobe mortar.
Contents: seven pebbles of banded fine-grained stone; four of which had
been used as hammerstones and three of which had been broken or had
had flakes struck off one surface.

Entrance. — None found; assumed entry was gained by means of hatch-
way in roof and ladder.

Postholes. — One found, close to south end of deflector on east side;
diameter, 8 cm.; depth, 3 cm. Use, unknown.

Roof. — On and near floor, hundreds of pieces of baked roof clay bearing
impressions of beams and twigs. No beams found; but in absence of
vertical roof supports, it is assumed that main roof beams rested on the
surface of the ground and probably extended north and south, the short
dimension of the kiva. Then other beams, running east and west, were
placed on main timbers. Ceiling was probably about 2 m. above floor.

Secondary Ventilator (?). — Immediately (81 cm.) west of bench, in south
wall, and 34 cm. above floor was an aperture sealed with sandstone slabs
set horizontally and diagonally. Dimensions of opening when unsealed:
height, 30 cm.; width, 23 cm. The opening was explored and after remov-
ing soft dirt fill, was found to extend south for 1 m. ending in a vertical
opening 32 cm. in diameter. This in turn extended upward for 70 cm.
At time of excavation, it was called a secondary (?) ventilator, although
its position above the floor level (34 cm.) makes this unlikely. The entire
aperture is too small for use as a secret (?) or sacred (?) entryway to the
kiva, unless used by a small person or child. The construction of this
feature is a puzzle, especially the vertical shaft that probably had (at one
time) extended to the ground level.

I ig. 35. Surface kiva (room 29) with Hour of flagstones and rectangular, slab-lined
Grepit Bench on <-.i^t gide partially destroyed by vandals was incur, l by ventilator
opening; tunnel, however, ran north. Arrow (50 cm. long points north; meter stick

in t>.u kL'round.

51

52

PREHISTORY OF EASTERN ARIZONA, III

Fig. 36. Kiva under (later) room 41. Round firepit near center; east face, slab-
lined; pot rests(?) in firepit. East of firepit is deflector. Earthen bench provided with
edge-joint consisting of a log; is pierced by ventilator tunnel, face of which is horse-
shoe-shaped ring slab. Pit in top of east bench contained four hammerstones. Arrow (50
cm. long) points north, meter stick on bench.

General Remarks. — Since the walls and floor of room 41 were built over
the kiva, it is obvious that the kiva was built and used before the exist-
ence of room 41.

The floor of the kiva was laid on top of a floor of a pit house that
belonged to the Forestdale Phase (A.D. 600-800).

The rectangular shape of this kiva is similar to others of this time
period in the area of the Upper Little Colorado Drainage and, of course,
to kivas of the same period in the Hopi country. Many artifacts were
found on the floor and a few on the bench. From the floor: bone rings
and raw materials for making rings; a projectile point; 3 scrapers; 7 ham-
merstones; 3 choppers; 2 one-hand manos; fragments of 7 other manos;
5 stream pebbles, some of which were large enough to serve as seats;
3 bone awls; one marked slab of wood (altar piece?); splints of yucca; a
"brush" of twigs; a corn stalk with corn cob attached; corn cobs (25);
squash seeds; pieces of twilled matting; 2 ribs of a very large mammal;
one corrugated jar (small); 2 stone "tinklers"; on the bench: 1 deer antler
and 2 deer legs; 5 hammerstones.

I Kiva undo room 41 ; iict.nl ol > .ist end; pit in bench; log embedded in edge

d1 bench; horseshoe-shaped ring slab ventilator opening; deflectoi and east edg<
fircpit. Arro point- north; meter stick in near background.

53

54 PREHISTORY OF EASTERN ARIZONA, III

Northwest Plaza Kiva
Location. — In northwest corner of plaza.

Fill. — Entire kiva filled with rich, artifact- and sherd-bearing trash
except for top 35 cm. Depth of fill was 1.75 m.; of which 1.40 m. was
"artifactiferous."

Shape. — Square.

Orientation. — East and west walls, magnetic north.

Size. — 3.45 m. at floor level; only 2.90 m. at ends. Walls curve slightly.

Depth of floor below surface. — 1.75 m.

Walls. — Of native earth; plastered with two layers. At top of walls,
possibly two to three courses of masonry.

Floor. — Of native clay, with two layers of plaster, each 15 cm. thick.

Bench. — Composed of sterile red clay; may have been sculptured of
undisturbed and untouched earth. Width is 3.4 m.; depth (east-west),
85 cm.; height above floor, 50 cm.

Pilasters. — None.

Firepit. — Bowl-shaped with flat bottom. Diameter, 32 cm.; depth 21 cm.
Distance of firepit from north wall, 1.23 cm.; from west wall, 1.5 m.;
from bench, 80 cm.; lined with small thin sandstone slabs to a depth of
12 cm.; under them, native clay, burned red. Bottom of pit sunk into
bed rock for 3 cm. Contents: ash; some charcoal.

Ashpit. — Directly east of firepit; square in shape; sides and bottom, of
sandstone slabs; slabs on sides flush with floor; length and width of pit,
24 cm.; depth, 10 cm.; slab on west edge of ashpit (small deflector?) is
2 cm. thick and 10 cm. high. Contents: ash.

Deflector. — Makes up north side of ashpit; dimensions: 4 cm. thick;
20 cm. high. Between deflector is a slight depression, sides of which are
bounded by slabs set flush with floor. Dimensions of this depression are:
width (east-west) 40 cm.; length, (north-south) 46 cm. inside measure-
ments. Floor of depression of native earth.

Ventilator. — Opening unadorned with any special horseshoe-shaped
slab; runs almost due east for 85 cm., the depth of the bench, and there
joined a vertical shaft. Dimensions at face of bench, 30 by 30 cm. Floor
of tunnel slopes upward at an angle of 10 degrees. Roof of tunnel com-
posed of small cross-beams and slabs.

Cupboard (?). — Just south of ventilator tunnel was a storage area similar
in general position to one found in kiva 4-5-6. The north edge of the
opening was 10 cm. south of vent-tunnel. The floor sloped slightly up-
ward towards the east. The rear wall of this cupboard extended 10 cm.

\k< HI I I ( 11 R \I. Dl I \!l S 55

farther cast than the east wall ol the kiva; total length was 1.12 m. In
section, the pit \\ idened from 30 cm. .it the front to 39 cm. at rear, and
was shaped somewhat like ,i truncated teardrop. The compartment v.
75 tin. high, ami was roofed b) slabs set in the bench \> i i to the cup-
board was sealed w iih stone slab J6 cm. wide and 37 cm. high, and that
u.h -ci in floor ol kiva to .i depth ol 7 < m. Flooi ol cupboard u,h ol stone
-Kilts. I rider them were three "potter) rests" of coiled juniper bark.

Vichi One, in north wall near bench at heigh) ol 90 < m. from floor.
Dimensions: width, 70 cm. east-west; height, 2~> cm.; depth, m.

1 intents: nothing.

Pits in '.' i . None.

Postfi .■ One, in north wall, 50 cm. from north wall, and 80 cm.
above floor; depth, 20 cm. Ma\ have been horizontal post oi beam to
support a shelf.

/; Burned clay impressions on floor indicate root composed ol

beams, cross-beams, branches, and mud, same as roots of surface rooms.
B .mis might have been sel criss-cross at right angles to one another with
beam-ends resting on masonry; or beams could have been placed across
corners to form a cribbed roofing or dome. Height of ceiling on inside ol
kiva about 2 in.

/ ■ :■ ince. By means of a roof-hatchway.

(•■'ieral R< kiva had burned and then was filled with trash.

Egg shells were found in fill near the floor. Firepit contained fragments
orn i olis.

III. Artifacts

By William A. Longacre

A total of 2931 artifacts of stone, bone, shell, baked clay, and cordage
was recovered from Broken K Pueblo. The purpose of this chapter is two-
fold: (1) to present the basis for the descriptive classes and types of these
artifacts, and (2) to provide interpretive statements regarding these arti-
facts and their distribution in the Pueblo. Descriptive details and prove-
nience of individual artifacts are recorded in the Archives of Archaeology
series (Martin, Longacre and Hill, 1966). Inferences from the distribu-
tions of certain tools at the site are reported elsewhere (Longacre, 1966)
and will not be repeated here.

The artifacts here described were cataloged in the field by Miss Anna
Rose working under the direction of Martin, Hill, and Fritz. Because cer-
tain descriptive data were not recorded in the original cataloging and
because I wanted to check the basis of the typology, I undertook the task
of completely reworking the collection of cultural items recovered from
Broken K. This was only made possible by the careful work done in the
field by Miss Rose and the others in their exact recording and labeling
of individual specimens.

I was helped in this project by a number of people. Without their aid
this work and the resulting report could not have been completed. I
must, however, assume full responsibility for the shortcomings.

I acknowledge the aid of the following individuals and thank them
for their generous assistance. Miss Margaret Hardin, a graduate student
in the Department of Anthropology, University of Chicago, aided me in
the initial sorting and analysis. Miss Susan Wartell, a graduate student in
the Department of Anthropology, University of Illinois, spent an entire
summer working as a volunteer on this project. For her hard work and
insight-filled suggestions, I am most grateful. Mr. Allan Hammaker, a stu-
dent at Antioch College, helped in the sorting of artifacts and made many
of the metric observations. The tedious calculations involved in the sta-
tistical tests for significance applied to the metrical attributes of the tools
made on flakes (pp. 94-99, below) were undertaken by Mr. Elliott Simon,
a graduate student, Department of Human Development, University of
Chicago. The identification of the shells made into artifacts was carried

56

i - i i rinding surface from room 11. floor. Length is

13.- • . here arc both top and bottom faces (upper).

57

58

PREHISTORY OF EASTERN ARIZONA, III

Fig. 39. One-hand mano, two grinding surfaces from room 31, floor. Length is
17 cm.

out by Dr. Alan Solem, Curator of Lower Invertebrates, Field Museum
of Natural History. Faunal identification of artifacts of bone was made by
Dr. Joseph Curtis Moore, Curator of Mammals, Field Museum of Natural
History and Dr. Donald F. Hoffmeister, Museum of Natural History,
University of Illinois.

For the descriptive portion of this report, I have tried to follow Ri-
naldo's classification of artifacts from the Carter Ranch Site (Martin
et al., 1964b, Chap. Ill) as much as possible. This was attempted to aid
the comparative value of this report. I was generally successful in this
endeavor, but there are significant differences in the typologies which
are noted in the following pages.

GROUND AND PECKED STONE

Manos. — The 490 manos recovered from Broken K Pueblo were clas-
sified on the basis of size, form, and the nature and number of grinding
surfaces. There were 230 whole manos excavated and 260 fragments.

All were partially shaped by pecking and grinding and the edges of
many were roughed into shape by flaking (Fig. 38). Grinding surfaces
ranged from flat to convex with the latter being more frequent. Manos
with a single grinding surface were slightly more frequent (191) than

Ak I 11 \< I S

59

those having two grinding surfaces on opposite faces of the mano 1 2).
This is similar to the data from the Cartel Ranch Site Martin et al.t
lb, p. 6
Typologically, the in. mo- from Broken K Pueblo were segregated into
two major classes based upon size and shape. Those having a roughly
square or circular outline were classified as cobble oi "one-hand manos";

those with a rectangular shape were classified as "two-hand manos."

These major classes were further subdivided into types on the basis ol the
number and nature ol the grinding surfaces. The types and frequencies

ol manos .ire as follow SI

ine-hand manos, single grinding surface Fig J8): 4o (34 whole

examples; 1 2 fragments).

One-hand manos, two grinding surfaces (Fig. 39): -11 (36 whole; 5
fragments).

I wo-hand manos. single grinding surface (Fin. 40): 145 (77 whole; 68
fragments I.

Two-hand manos. two grinding surfaces (Fig. 41. Fig. 42): 131
w hole: 78 fragments

I ■- I wo-hand mano, single grinding surface from Broken K. Pueblo. One edge

has fij Length is 34.4 cm.

Two-hand manos. beveled type grinding surface beveled into two
planes on one surface Fig. 43, I. 12): 116 (27 whole; 89 fragments).

Mano "blanks" unused rectangular blocks ol' stone, roughed into
two-hand mano size and shape, hut showing no e\ idence of use: 3 whole
exampl

Fig. 41. Two-hand mano, two grinding surfaces with finger grooves worked into two
edges; from the roof of room 1 1 . Length is 23.3 cm.

Fig. 42. Manos, Beveled type, a; two-hand mano, two grinding surfaces, b. Length
of b is 22.5 cm.

60

AK1 II A( 1^

61

Inclctfnnin.ur in. mo fragments size, shape, and t\ pe indeterminati

* leneral comments: about one-fifth of the manos had depressions pecked
in their sides or with their sides slightl) incurved which ma) have served
.i- grips I ig. 40, Fig. 41). The presence ol so man) rectangular,
beveled type manos would seem consistent with .i relativel) late date foi
the i» t upation oi the site, perhaps into the beginning of the 1 4th century

cf. M.utin et <;/.. 1964b, |>. 65; Woodburj 1954, p. 70). If the general
correlation between lateness oi tune .mil increasing frequency ol beveled
type manos is valid, then it would seem th.it Broken K Pueblo would be
earlier than Table Rock Pueblo which seems to roughly bracket the 14th
century. At the latter sin-, about one-third ol the manos were beveled

Martin ami Rinaldo. 1960b, p. 12~)\ whereas onl\ about one-fourth oi
the manos at Broken K. were oi this type.

Metates. A total oi 65 metates was recovered from the Broken K site.

Whole examples numbered 45, while fragments accounted tor 2<> addi-
tional metates

The metates were of three basic types: hasin nictates irregular hlock
<>! stone with a circular, depressed grinding surface: 7 (5 whole examples;
2 fragments); trough type roughly rectangular hlock of stone, grindinj

I -'■ rwo-hand mano, beveled type with three grinding surfaces; room 33
.-. mealing bin) Length is 19.9 em.

surface forms trough through surface: (1 whole: 6 fragments); and
slab type rectangular hlock of stone, whole of top surface forms nearly
Hat grinding surfa< ' whole; 11 fragments). There was one- frag-

ment recovered that was too small to permit a reliable estimate of type.
It is significant that the majority (77 < j I oi the metates are the slab type
Rinaldo Martin et ol., 1964b, p. 67) has noted the observed trend of

62 PREHISTORY OF EASTERN ARIZONA, III

gradual replacement of other metate types by the flat slab metate (cf.
Bartlett, 1933, p. 26). This seems to correlate in a general sense with the
use of mealing bins (cf. Martin et al., 1964b, p. 67; Bartlett, 1933, p. 26).
None of the metates of the basin or trough types was found in association
with a mealing bin, however, 1 1 slab type metates were found in place
in bins.

There were many fewer slab metates at the Carter Ranch Site as well
as fewer mealing bins. This would argue for a later date for Broken K
Pueblo and as such is consistent with our other data. Rinaldo interprets
the high frequency of fragmentary metates and manos at the Carter
Ranch Site as indicating intentional breakage before abandonment (Mar-
tin et al., 1964b, p. 67). It may also be interpreted as an indication that
unbroken tools were carried out of the Pueblo for future use at another
location, creating an unusually high number of fragments in relation to
the number of whole examples. At Broken K Pueblo, the majority were
whole (45 out of 65).

The metate types from Broken K Pueblo are not illustrated. They are
basically similar to the types from the Carter Ranch Site (cf. Martin et al.,
1964b, Fig. 28).

Mortars. — Twenty-two mortars were recovered from the excavation of
Broken K Pueblo. These are stone cobbles or slabs with one or more cir-
cular depressions worked into the surface. One of the most important
attributes of this class of tools is the nature of the cavity itself. The de-
pressions of the mortars contain no angular facets; the sides and bottoms
of the cavity are curved and smooth. The class labeled mortars can be
divided into two types based on the metric attributes of the depression.

The first type has a cavity worked in the surface that is relatively small
and shallow. The circular depression on mortars of this type is fewer
than 6 cm. in diameter, and is less than 1cm. in depth. Mortars of this type
contained the stains of various pigments; red and black were the most com-
mon colors (Fig. 44, c). Ten examples of this type were excavated. One
mortar of this type was found which exhibited two worked depressions.
One cavity contained black pigment; the other was stained from a red pig-
ment (Fig. 45,6). In addition to these specimens, three were found contain-
ing a pecked depression of this size, but showed no signs of smoothing or
use (Fig. 44,a).

The second type of mortar has a significantly larger depression worked
into the surface of the tool. The diameter of the cavity of this type is
greater than 7 cm. and is relatively deep, greater than 1.5 cm. Interest-
ingly, there was no trace of pigment found on mortars of this kind. Eight
mortars of this type were found (Fig. 44, b).

\K i II \< Is

B

I M. Moi tai s. I ,ength of < is 17.5 cm.

\ functional inference ma) be made after the class labeled mortars is
egated into two major types based on size differences of the worked
depression. I he presence ol pigment on the examples of the first and
smaller type would seem to indicate their use in the processing of pig-
ments. Reducing lumps of pigments to .1 fine powder pi ior to mixing with
a liquid vehicle is the inferred use ol artifacts of this type (see paint pal-
ettes, pp. 64 66, below). The mortars of the second type were not used
leu grinding pigments, but their exact use is not known. The) probably

64 PREHISTORY OF EASTERN ARIZONA, III

were utilized for a variety of tasks, including the crushing of seeds and
other foodstuffs, as well as clay and temper for pottery manufacture.

Two bedrock mortars were found between outliers one and two at the
site (Fig. 1). These are probably associated with the Pueblo, but it is im-
possible to demonstrate this.

Type I mortars are what Rinaldo (Martin et al., 1964b, pp. 70-71)
refers to as "rough block mortars with a shallow cup." They are similar
to what Woodbury (1954, p. 113) terms grinding slabs. Significantly,
Woodbury {ibid.) reports stains from pigments on these items and infers
a function of processing paints for them.

Pestles. — A total of 1 1 pestles was excavated at the Broken K site.
These items were shaped to some degree. They are smoothed, cylindri-
cally-shaped stones with one or both ends worn from use. This class, like
the mortars, is divided into two major types on the basis of metric attri-
butes.

The first type consists of the smaller pestles, 7-9 cm. in length and
less than 8 cm. in diameter. There were eight pestles of this type re-
covered (Fig. 46a-c).

The second type is a larger implement, greater than 10 cm. in length
and more than 6 cm. in diameter. Only three examples of this type were
found.

The pestles recovered from the Carter Ranch Site were much larger?
on the whole, than those from Broken K (Martin et al., 1964b, p. 68)-
There is no apparent functional reason explaining this difference in size.
Style change reflecting a temporal difference would not seem to be an
adequate explanation as sites of a comparable age and more recent date
in this area have produced pestles of larger size, similar to those from the
Carter Ranch Site (Martin and Rinaldo, 1960b, pp. 243 and 247; Mar-
tin, Rinaldo, and Longacre, 1961a, p. 71, Fig. 41).

Paint Palettes. — Eleven paint palettes were found at the Broken K site.
All but one of these was made of stone; one was simply a large sherd.
The stone palettes were shaped and smoothed to create a flat or nearly
flat surface for the purpose of mixing pigments with a liquid vehicle, or
to otherwise process paints. All specimens contained the remains of pig-
ments. This class of items is segregated into the following types:

Type I: Irregular stone cobble or slab with a large, shallow circular
depression worked into one surface, with steep, angular sides. Six of this
type were recovered (Fig. 45, a). In addition, one example of this type was
found which contained three circular depressions in a single slab.

*m

■

^ •-*>,•*

Painl palettes (a and I and double mortal Length ol is 21

66

PREHISTORY OF EASTERN ARIZONA, III

Fig. 46. Pestles. Length of c is 7.3 cm.

Type II: Shaped rectangular slab with a square depression worked
into one surface which is well-smoothed. In the single example of this
type found at Broken K Pueblo, there is a narrow and deep groove or
trough worked into one corner to facilitate the pouring of a liquid from
the palette into a container (Fig. 45, c). It is similar to one illustrated by
Woodbury (1954, Fig. 23, a) from Awatovi, Pueblo IV date.

Type III: Thin, rectangular slab of sandstone, one area of surface
smoothed. Two palettes of this type were found. These are similar to
items called "lapstones" by Rinaldo found at Table Rock Pueblo (Martin
and Rinaldo, 1960b, p. 249).

Type IV: Large, shallow jar sherd (Snowfiake black-on-white, Carter-
ville variety) with the interior covered with pigment. No smoothing is
apparent. Only one example of this type was found.

The stone palettes differ from the mortars (pp. 62-64, above) in that
the depressions are angular with relatively fiat bottoms. All specimens
contained pigment; red and black were the only two colors noted.

Polishing stone. — Only two items that could be called polishing stones
were recovered. These are small, unaltered stream pebbles which were
smoothed from use forming one or more facets. A large number of small,
smooth pebbles were found at the site, but they were discarded in the
field.

[•} Hammerst • hopper (/ I, and block core with knife worked on

one i I I engtl i tn.

68 PREHISTORY OF EASTERN ARIZONA, III

Rubbing stone. — These items are shaped stones with one or two faces
smoothed from use. It is possible to divide this class into three types on
the basis of size, shape, and material.

Type I : Large, smooth stream cobble which is worked into rectangular
shape. Two examples were found.

Type II: Similar to Type I in size and shape, made from a fine-grained
sandstone with edges squared and smoothed. Four of this type were re-
covered.

Type III: Large oval cobble, carefully smoothed and shaped with all
specimens showing evidence of battering on their edges as well as use in
smoothing. Three of this type were excavated at this site.

The polishing and rubbing stones are essentially similar to those re-
covered from the Carter Ranch Site (Martin et al., 1964b, pp. 67-68)
and are not illustrated herein {ibid, fig. 29, p. 69).

Hammer stones. — A total of 247 artifacts of this class were recovered in
the excavation of Broken K Pueblo. The class was divided into three
major types as follows:

Type I: LInmodified stream pebbles, smooth and water worn, with
one or more surfaces or edges showing scars and flaking resulting from
battering. They are irregular in shape, ranging in size from 2 cm. in
diameter and 5 cm. in length to the largest examples that tend to spheri-
cal shape with diameters approaching 20 cm. A total of 133 examples of
this type was found (Fig. 47, c and g).

Type II: Similar to Type I in that these hammerstones consist of stream
pebbles, but they are modified with small pecked depressions on two sur-
faces, probably to serve as finger grips. Nine specimens of this type were
recovered (Fig. 47, a).

Type III: Exhausted core nuclei or workable core nuclei utilized as
hammerstones as evidenced by battered edges or surfaces or both. A
total of 105 was found (Fig. 47, d and e).

Hammerstones are among the most frequent artifacts recovered from
prehistoric sites, in the Southwest as well as other parts of the world.
Their frequency reflects their importance to a people dependent upon
the use of lithic tools. Hammerstones are probably one of the most im-
portant items in the stone knapper's tool kit. Generally, their importance
as tools for chipping flint has not been recognized in the Southwest (e.g.,
Woodbury, 1954, pp. 92-93). Various uses have been suggested for these
items including the pecking of stone as part of the shaping operation for
such items as metates, axes, and others. This is undoubtedly an important
function of these numerous implements. Another use for these tools has

\l< I II \( TS

69

J***$fc

•^

I ig. 48. Stone axes. I hree-quarters grooved, a-d; unfinished three-quai tei s grooved,
c full-grooved,/. Length of / is 14.5 cm.

been suggested (Woodbury, 1954, p. 93) — the pecking of grinding sur-
of milling tools to roughen them lor more effective grinding. This
suggestion i^ supported l>\ the presence of hammerstones .it the Broken K.
Site in association with mealing Inns in rooms 7, 33, 82, and 92. The in-
ted reader is referred t * > Woodbury's excellent discussion of ham-
merstones [ibid, pp. (ruin the prehistoric Southwest.

The use of stream cobbles .is well as core nuclei for hammerstones
appears to l><- t.iirK common in the Southwest. Both forms were found at
the Carter Ranch Site Martin et a/., 1964b, pp. ~2 73). Wheat (1
pp. 12"i 127) reports similar t\|><s from the Point of Pines ar<

Fig. 49. Mauls. Three-quarters grooved, a; full-grooved, b-e. Length of c is 10.5 cm.

70

Ak riFACTS 71

Worked & \ numbei oi shaped sandstone slabs w< overed

from this sit< S era] classes ol this category can be listed, but no attempt

has been made to further segregate these items into typ< 1 hi cl
arc as follows:

Ring slab I ire large ca, JO 60cm aero quare to circular

s|.d>s uith circular holes worked into the center. They probably served
as vent and hatchway capstones. Foui were found at the site.

Circular slabs: I wo roughl) circular slabs were found in association
with a floor pit in room 53. The) probably served as covers foi the pit.

"Floor smoothers": These arc rectangular in outline with roughl)
trimmed edges. The surfaces var) from rough to smooth, possibly from
I he) range in size from about 6 b) 1<> cm. to 25 b) l11 cm., and in
thickness from 1.5 3.5 cm. Their regular shape and smoothed surl
suggest .i use in household activities such as floor smoothing or as an
implement useful in plastering floors or walls. A total oi 12 ol these items
was found.

ooved -perforated slabs: Two examples of grooved slabs were
found, one of which had a hole drilled through its center. Both were
roughl) rectangular in outline: the drilled specimen had a shallow groove
pecked from the central hole to the edge, longitudinally. The other had
a shallow groove peeked in one surface, from end to end. Their use is
unknown.

Painted dale ( )ne example of a painted slab was found. It was approxi-
mately rectangular in outline with rough edges and unsmoothed surfaces.
( >n one surface was barel) visible a geometric design painted in black.

Axes, rhirteen axes of various types were excavated at this site. H\
far the most common types were the three-quarters grooved forms. The
nomenclature proposed by Woodbury (1954, pp. 25 26) is used here.

r*ype I rhree-quarters grooved, no lip around groove. Three speci-
mens with a smoothed, flat inner side (Fig. 48, a and b), and one example
with a rounded, rough inner side were found.

1 ■■ i ■• II rhree-quarters grooved with a lip around the groove. All
examples of this type had a smoothed, flat inner side (Fig. 48, i am

111: Full-grooved; two examples found (Fig. 48,/).

I ■;■■ 1\ Grooved on two faces only; "one-hall grooved axe." Onl)
one example found.

I .pe V: I nfinished three-quarters grooved axe. < hae specimen found
which is shaped through pecking but is not polished I ig. 48, e).

The high number oi three-quart »oved axes at this sin- supports

iggestion Martin et a/., 1964b, p. 74) that this form appears

72 PREHISTORY OF EASTERN ARIZONA, III

to be typical of the later Little Colorado ruins, although earlier (1959,
p. 284) he suggests that the full-grooved axe is typical of the "Zuni area."
In general, the three-quarters grooved axe appears to have spread from
the Hohokam area northward. Woodbury (1954, pp. 25-42) presents an
excellent discussion of typological considerations, distribution, and in-
ference regarding axes from the prehistoric Southwest.

Mauls. — Thirteen grooved mauls were recovered. They range in size
from relatively small (Fig. 49, c) to fairly large and heavy (Fig. 49, e).
There are two basic types:

Type I: Full-grooved. The most common type (11 found) covering
the entire range of size (Fig. 49, b—e).

Type II: Three-quarters grooved. Only two examples found (Fig. 49, a).

Unlike axes, the most common form of mauls in the upper Little Colo-
rado area appears to be the full-grooved type. There were no recognizable
examples of re-used axes as mauls at this site. As he does for axes, Wood-
bury (1954, pp. 43-49) presents an excellent discussion of mauls from
the prehistoric Southwest.

Arrow-Shaft Tools. — Fifty-one arrow-shaft tools were recovered from
Broken K Pueblo. They were produced on a stone cobble or on a re-
worked mano. All have certain features in common: smooth, flat bottom,
and one or more smooth grooves worked into the upper surface (the groove
is polished and generally blackened). Some had ridges worked into the
upper surface, and four had red pigment adhering to them. The following
typology is based on attributes such as the presence or absence of ridges
and the nature of the surface containing the groove.

Type I: One transverse groove on a surface that is worked flat or
nearly so. No ridge is present. The outline of the tool is either oval or
rectangular in about equal frequency. Thirteen examples of this type
were found (Fig. 50, h).

Type II: Single transverse groove worked into a rounded and smoothed
pebble. This is comparable to Rinaldo's "Truncated Triangular Type"
(Martin et al., 1964b, p. 77). There is no ridge present. All but two ex-
amples were sub- triangular in outline; one was oval in shape and the
other rectangular. Fifteen examples of this type were recovered (Fig.
50, /). One variation of this type was recorded; this was a single example
of this type with three parallel, transverse grooves (Fig. 50, g).

Type III: One transverse groove worked into a rounded and smoothed
surface. A single ridge is worked as a "lip" on one side of the groove and
parallel to it. All four specimens of this type were sub-triangular in out-
line (Fig. 50, e).

V

ndi1

'<;

Vrrowshaft tools, transverse groove typ( I n ih o( h is 6 9 cm.

v.- ■

B

■ ' £

'0#~-

D

■

Fig. 51. Simple grooved abraders (a-c) and polished and incised arrowshaft tool,
longitudinal grooved type, (d). Length of d is 8.6 cm.

74

\K riFACTS 75

I \ j e l\ One transverse groove worked into .1 rounded to flat surfa< c
I tere is .1 single ridge perpendicular to th< c 1 he ridge is poorly

denned and short I m Ml six examples have oval outline-. (Fig.

I :■ \ One transverse groove with .1 single ridge perpendiculai to
the groove. The ridge is well-defined and Ion >cm I and is worked

onto .1 plane which meets the flat surface containing the groove at about
an angle of 30 degrees. The outline varies among oval, rectangular, and
sub-triangular forms. Six examples of this type were found (Fig. 50 c).

Type V-b: Essentially similar to 1 ype V-a except that both the trans-
verse groove and the perpendicular ridge are made on the same, flat
surface. ( ml\ one such tool was found > Fig. 50, a).

1 ype V-c: Same as Type V-a except there are two, parallel transverse
grooves. One example of this type recovered (Fig. 50,

1 \ pe V-d: Essentially the same as Type V-a except that there is a ridge
parallel to the groove as a "lip" in addition to the perpendicular ridge.
1 wo examples were found.

Type VI: Single groove worked into the longitudinal axis of a care-
full) shaped and polished stone. '1 here is a cross-hatchured design incised
on the ends, sides, and top surface of the single example found of this
type Fig. 51, d).

In addition to these specimens, one fragment of an arrowshaft tool
that was too small to type with any accuracy was recovered.

Pointing out the experiments conducted by Cosner 1 1951, pp. 146 147),

Rinaldo suggests that these tools were used to straighten arrow shafts of

Martin et a/., 1964b. p. 77). It is possible, of course, that additional

us<s were made of these tools, and some not involving arrow shafts at all

e g . preparing ritual items such as pahoes or spindles, etc.). VVoodbur)

1954, p. 101 and pp. 108-110) suggests the term "shaftsmoother" for

these items. His arguments point to the multifunctional nature of these

tools. A possible typological sequence is suggested b) Rinaldo 'Martin

ft a!.. I ... 78).

Abraden These are generally rectangular blocks of
coarse-grained sandstone with one or more grooves worked into one
sUii.:. -• I hree examples were recovered at the Broken K Site. < >ne has a
single longitudinal groove Fig >1, Another has two transverse grooves
Fig. 51, I. Rinaldo documents their apparent late distribution in the
- ithwesl Mai tin et al., 1964b, p. ~v

Stone Pipi I wo pipes and an undrilled "pipe-blank" were recovered
from Broken K. Pueblo. All three are sub-conical in shape. One is a

Fig. 52. Miscellaneous stone artifacts. Pipe blank, a; pipe, b; irregular piece of sand-
stone with groove, c; "medicine cylinder", d; stone "needle", e. Length of e is 11.6 cm.

76

\k I 11 \< I S

Projectile points. Miscellaneous types. Length ol / is 2.8 cm.

small pipe made oi vesicular basalt (Fig. 52, b). I he other two are some-
what larger in m/<* ami made from a fine-grained stone Fig. 52,

( HIPPED SI ONE

/'• Point For typological convenience, this class has been di-

vided into three groupings. First are described those projectile points
which, stylistically, are associated with sites utilized during the pre-
agi icultural occupation ol the region. I hese are sites of the regional mani-
festation ol the De ert Culture in our area, known as the Concho Com-
plex M.ii tin et a/., 1 '"'2. pp. 1 55 1 64; fig. 69, p. 1 58 1. Projectile points of
e styles were found in surprising abundance at this late pueblo site
21 points). Because almost all of them were found in floor association

78 PREHISTORY OF EASTERN ARIZONA, III

or in the fills of abandoned rooms, it is probable that these early points
were collected by the inhabitants of Broken K Pueblo and brought to
the village. There was no evidence of an earlier, Desert Culture occupa-
tion of the site. This interpretation is supported by the complete lack of
other artifacts stylistically associated with the Concho Complex. This
situation prevailed at the Carter Ranch Site as well (Martin et al., 1964b,
p. 84).

The second grouping consists of eight fairly large corner-notched and
side-notched projectile points, stylistically associated with pit-house com-
munities dating prior to roughly A.D. 800. These, too, were found in
floor and fill context at the site. It is possible that these points were col-
lected from earlier sites in the area and brought to Broken K at the time
of its occupation. It is probable, however, that points of this style are
associated with an earlier component underlying the major occupation
of this site. Portions of two pit houses were excavated beneath the pueblo,
and early ceramic types were found (cf. Chap. IV).

The third grouping consists of the smaller, triangular projectile points,
stylistically associated with sites post-dating ca. A.D. 800 in much of the
Southwest. It is felt that these points represent the types actually made
during the occupation of Broken K Pueblo. It is this group of points
that is the most numerous (31 points).

Group I: Projectile points stylistically associated with the Concho
Complex (Fig. 53). Numbers in parentheses refer to quantity found at
Broken K Pueblo.

Type I: (2) Expanding stem, indented base, serrated blade, barb at
corners of base (Fig. 53, d and e).

Type II: (2) Triangular, indented base, serrated blade (Fig. 53,/).

Type III: (1) Straight stem, indented base, barb at base of blade
(Fig.' 53, b).

Type IV: (1) Expanding stem, straight base.

Type V: (1) Straight stem, notch in base (Fig. 53, c).

Type VI: (1) Contracting stem, excurvate base (Fig. 53, a).

Type VII: (1) Contracting stem, straight base.

Type VIII: (1) Contracting stem, indented base.

Type IX: (1) Expanding stem, indented base.

Type X: (1) Side notched, straight base.

Type XI: (1) Side notched, excurvate base.

Type XII: (1) Triangular, indented base.

Indeterminate Fragments: (6)

\r i ii \» rs

I i : Projectile points, miscellaneous types. Length ol is 1.2 cm.

Group 11: Larger notched forms, stylistically associated with pit house
sites, ca. \ D 200 800 Fi A Numbers in parentheses refei to fre-
quency u! occurrence for each type at Broken K Pueblo

Type 1 I) ( "i ner notched, straight base i Fig. 54, / ).

fype U Side notched, straight base (Fig. 54, a reworked).

rypi III I) < lorner notched, excurvate base i Fig. 54, /<).

Type IV: 1) Side notched, indented base Fig 54, d).

I . ; . \ I cpanding stem I i 54, b).

I ype \ I I) Coi nei not( hed, serrated blade I Fig. 54,

I peVU I ) Side notched, excurvate base, serrated blade I

Fig. 55. Projectile points, miscellaneous types. Length of p is 2.1 cm.

80

\K riFACTS 81

iup III Small, generally triangular forms, associated with pueblo
sites dating ca. A D 300 and later. Numbers in parentheses refer to fre-
quency tt the site.

r. : . 1 2 rriangular, side notch, basal notch Fi

I.;. II rriangular, side notched, excurvate base Fig 5 '•. I

I-.:, in 2 rriangular, double side notched, indented base 1

rype l\ 2 rriangular, indented base, side notched Fi

Type \ 10 1 i iangular, straight basi Fig 55, i

I \ pe \ I 5 I i iangular, indented base Fi •

I ype \ 11 -1 fragments) Straight stemmed, triangular blade.

rype \ 111 1' Triangular, straight stem, serrated blade (Fig. 55, m).

rype IX I1 rriangular, corner notched (Fig. 55, n).

l\j. \ 1) rriangular, expanding stem, straight bast Fig 55,

Type \1 I eaf-shaped, contracting stem (Fig. 55, /).

I ype XI 1 1 Diamond shape, side notch.

Indeterminate fragments: ' 1 ).

In his discussion of the projectile points from the Carter Ranch Site,
Rinaldi Martin <t al., 1964b, p. 84) presents some interesting observa-
tions. He notes the presence of the small, triangular points at South-
western sites alter about A.D. 900. He suggests the presence of lateral
notches about half the distance from the point to the base ma) I"- diag-
nostic of later Mte- in the Southwest. A total of 19 such point- were found
at ( larter Ranch. I he small number of projectile points with this chara< -
teristic found at the Broken K Site suggests thai this particular style was
relatively short-lived. The most numerous type at Broken K. Pueblo was

simple triangular shape with a straight base. I he presence of small,
side-notched, triangular points with a well-defined basal notch ma\ in-
dicate some level of interaction with the Jeddito area. This particular
form is one of the most abundant in that region after about A.D. 1100
Woodbury, 1954, p. 121 ).

Drill A total of twelve drills or fragments was recovered from the
ations at Broken K Ml were produced on Hakes, and were bi-

ialh worked.

I : • I Square or curved base, parallel side-. 'I hree examples of this
type were found I ig , t, and

I ;. II Flarin ;panding base. Five of this type were recovered

I . . . an<

I ; ■ 111 Broad flake with a drill or punch worked into one end. < )ne
spe< imen was found ' I ig. 56, a).

82

PREHISTORY OF EASTERN ARIZONA, III

Three fragments of drills were found which were too small to assign
to a type.

The drills from Broken K are similar typologically to those from the
Carter Ranch Site (Martin et al., 1964b, p. 86). In addition, their rela-

Fig. 56. Drills and punches. Punch, a; drill with flaring base, b, c; drill, d-g.
Length of g is 5.4 cm.

tive frequency compared to other items such as projectile points was
low at both sites.

Saws. — Ten saws were recovered at the Broken K site. Most of these
consisted of irregular flakes with several deep notches worked into one
or more edges.

Type I: Irregular flakes with regularly-spaced deep notches on edges.
Nine examples were found (Fig. 58, a, c, and d).

Type II: Reworked knife, leaf-shaped, with notches regularly spaced
on one side. Only one was recovered (Fig. 58, b).

I : 7. Kin\ e - Length oi a is 5

84

PREHISTORY OF EASTERN ARIZONA, III

Fig. 58. Saws (a-d) and notched side scrapers (e and/). Length of/ is 4.2 cm.

Similar items occurred in much greater frequency at the Carter Ranch
Site (Martin et al., 1964b, pp. 86-87). The marked decrease in quantity
may have temporal significance and perhaps points to a functional change
in the nature of the tool-kit after about A.D. 1200. Rinaldo (Martin et al.,
1964b, pp. 86-87) presents a summary of the distribution of saws in the
Southwest and suggests some possible uses.

Gravers. — Twenty engraving tools were found at this site. All were
produced on flakes or "shatter", and all were single-point gravers.

Type I: Small, irregular flake with a chisel-like point worked into
one end. Two examples of this type were recovered (Fig. 59, h and i).

AKI II \( I v

85

Graving tools. Graven i . g i; micro-gravers, e and/. Length oi
2.5 cm,

I -. ; '■ II Small, sharp graving point worked on broad Hake or small
piece "i shatter. Length of the flake varied from 2."> 4.2 cm. with a mean
length of 3.3 cm. Twelve of this type were found (Fig. 59, a-d and

I •. ! < III: Microgravers small narrow flakes with a small graving
point worked into one end. Flake length varies from 1.0 2.3 cm.; mean
length i- 1.9 cm. Six examples were found (Fig. 59, e—f).

Utilized Flakes.- This class of artifact was one of the most numerous
recovered at Broken K. h consists of all Hakes, including shatter, showing
no regular and purposeful retouch, but exhibiting irregular chipping
resulting from use. A total of 226 of these sharp-edged Hakes showing
signs oi use was recovered. The) ranged in size from about 1 to 9 cm. in
length and from about 1 to 7 cm. in width. Their frequency in terms of
shape categories is presented in graphic form in Figure 60.

Rinaldo Martin et a/., 1964b, p. 88) lumped the utilized flakes from
Carter Ranch with the Hake knives, showing purposeful retouch. I will
not follow this classification foi reasons that will be made clear.

/ .•• Knivi I his (lass ,,i items consists ,,i .,n flakes which have been
purposel) modified to form one oi more acute cutting edges fhe) are

86

PREHISTORY OF EASTERN ARIZONA, III

retouched, generally bifacially, but the shape of the flake is unmodified.
A total of 106 flake knives was found at the site. This class is divided
into the following two types:

Type I: These are flakes struck from a prepared core surface — the
Levallois Technique. Fifteen examples were recovered from the site
(Fig. 61, a-g). The shape of the flake was predetermined by removing
smaller flakes from the surface of the core and by preparing a striking
platform. The more classic examples of the Levallois Flake tend to be
somewhat oval in shape (Fig. 61,/ and g).

Type II: These are random flakes and pieces of shatter with retouched
edges. A total of 88 of these tools was found (Fig. 62, a-o). All types of
flakes were represented in this collection, including expanding, contract-
ing, and irregular flakes, as well as both primary and secondary decorti-
cation flakes (White, Binford, and Papworth, 1963, p. 5). The flakes
varied in length from 1.5 to 8.0 cm., and in width from 1.0 to 7.0 cm.
The frequencies of occurrence of the length and width dimensions are
shown in Figure 63.

Blades. — A caution is mandatory at this point as I am departing from
Rinaldo's use of this term in his typology (Martin et a/., 1964b, p. 84).
Rinaldo uses this term to refer to bifacially-worked, leaf-shaped artifacts.
In adopting this term he follows many workers in the Southwest (cf.

100

UTILIZED FLAKES

length

width

4 5 6

centimeters

Fig. 60. Distribution of dimensions of utilized flakes.

r-.

1 ...

-*

*

^>

•

g

61. Flake knh a made from prepared cores Length ol t is 4.02 cm.

87

Fig. 62. Flake knives. Length of o is 1.6 cm.

88

\u i ii \( rs

89

■ ■

- — —

•4 5 6

centimeters

I •. -n ibution of liimcnsions of random-llakc knives.

Woodbury, 1 954, pp. 121 123). Id nearly every other part of the world, the
term bladi is reserved for a specific type of thin Bake with parallel sides
and a width to length ratio below 0.35 or .40. Blades are the result of a
particular flaking technique and tools made on blades are quite distinc-
tive. I feel that the term blade should be used only for such specialized
flakes cf. White, Binford, and Papworth, 1963. pp. 18-23).

Blades are present in many Southwestern assemblages, but, because

the\ are not fated .is a flake type, we do not know how important

the\ were in the prehistoric tool kit. Blades and near-l iladcs or lamellar

flakes White, Binford, and Papworth, 1963, pp. 15 18) are present in

the assemblage at the Broken K site. They were almost all utilized in un-
modified form. A total of 21 unmodified hut utilized blades was found
I Two examples were recovered with steep retouch along one

edge and were classified as backed blades. All the blades and Lamellar
flakes from the Broken K Site showed irregular utilization flaking along
one oi both edg< Mean length of these items is 4.4 cm.; mean width
is 2. i ( m.

/. ■ I ... e are bifacially-shaped and trimmed artifacts generally

having a lanceolate shape. The) differ from Hake knives in that the basic
shape ol the Hake has been altered (see pp. 85 86, above). These are the
items that have often been termed blades l>\ various Southwest prehis-

90

PREHISTORY OF EASTERN ARIZONA, III

jh ^mi w *■' •

^SQi*

Fig. 64. Blades (a, b, e-i) and backed blades (c-d) Length of d is 3.0 cm.

torians (cf. Martin et al., 1964b, p. 84). This terminology is not employed
in this report for reasons discussed above (p. 86).

These items were probably hafted as cutting implements. The acute
bifacial retouch would suggest a cutting function. The majority of these
tools have flat or rounded bases with no modification for hafting apparent.
Twenty-two were recovered (Fig. 57, b-f, and h).

Two examples were found with basal modification to facilitate hafting
(Fig. 57, a and g).

Scrapers. — I use this term to describe a class of tools with a working
edge produced with steep retouch techniques. The angle of the working
edge is thus steep and not acute. Types within this class are determined

I .. ^ule scrap' i I hose made on Bakes I. g), on a core (/ I, and on .t cobble

I ■',,. •:.,:• i j l cm.

91

92

PREHISTORY OF EASTERN ARIZONA, III

by the location of the retouch. The shape of the flake or cobble on which
these tools were produced is essentially unaltered.

Side Scrapers. — The steep retouch appears along one or both sides
measured in terms of the axis of percussion of a flake or the longest di-
mension of a core or cobble.

60

50

40

u
§30

/
/

20

/

/

101-

SIDE SCRAPER

length
width

4 5 6

ce ntimet ers

10

Fig. 66. Distribution of dimensions of side scrapers.

Ten side scrapers were found that were produced on exhausted core
nuclei (Fig. 65, b) and six that were made on small cobbles (Fig. 65, a).
By far the majority of these tools were produced on flakes. A total of 112
side scrapers on flakes was excavated (Fig. 65, c-g). The distribution of
the flake sizes selected for modification into side scrapers is presented in
Figure 66.

Two specialized examples of side scrapers were recovered. Both of
these were made on fairly small contracting flakes. There is a broad,
shallow notch worked into one side with further modification in the
form of steep retouch (Fig. 58, e and/). These items have a series of
attributes that suggest their use as a spokeshave or similar tool.

End Scrapers. — The steep retouch appears at one or both ends of the
tool, measured in terms of the axis of percussion of the flakes and, for
those made on core nuclei or small cobbles, the longest dimension. Ex-
actly 50 end scrapers were recovered during the excavations at Broken K
Pueblo, and all but seven of them were produced on flakes (Fig. 67).

\kin \< i -

f

>..

**c

<>r*-

I ' \ liid scrapci- 1 hose made on flaki , mi a con and on .1 cobble

(h). Length "i h is 4.0 cm.

1 hnly three end scrapers were found that had been made on exhausted
nut lei. All three were cores of the split cobble type, and all three
had steep secondar) retouch at one end I ig. 67, t

I mil were produced on small river pebbles. These tended to he thin
ami nerall) rectangular shape. Steep secondary retouch was present

ai the end of the tool measured as the longest dimension (Fig. 57,

94

PREHISTORY OF EASTERN ARIZONA, III

END SCRAPER

length
width

4 5 6

centimeters

Fig. 68. Distribution of dimensions of end-scrapers.

A total of 43 end scrapers made on flakes was found at the site. The
steep retouch appears at one or both ends, and the shape of the flake is
unaltered. The flake shapes selected by the knappers for end scrapers are
presented graphically in Figure 68. The mean size of these tools is about
4 cm. long and 3 cm. wide.

Ovate Scrapers. — These implements exhibit steep retouch on two or
more edges. The prepared edge is not limited to an end or side. The
outline of these tools tends to be oval or square (Figs. 69 and 70).
Most of the ovate scrapers were made on flakes; there were 123 of these
out of a total of 163 ovate scrapers. Only 24 were produced on a core
nucleus and all of these were made on cores of the split cobble type.
Nine ovate scrapers were produced on small stream pebbles.

Those made on flakes had a mean size of approximately 4.5 by 3.5 cm.
Ovate scrapers made on cores and pebbles tended to be less uniform in
size and shape. The basic shape of the flake, pebble, or core was unaltered
by the retouch (cf. Fig. 71).

Shapes of Scrapers and Knives. — In order better to understand the knapping
technology present at Broken K Pueblo, it was decided to examine the
shapes of flakes and tools made on flakes and, in some cases, small stream
pebbles. We wished to know if certain shapes that were available were
being selected for particular kinds of tools.

\r i ii \< rs

I ig. 69. Ovate scrapers. Those made on cores (a and /<), and on cobbles i an

I > iiL'tti of d is 4.8 cm.

Ideally, it would have been best to compare the various tool types
produced on Hake- where the outline of the Hake was essentially unaltered
with the unused lithic debris. But we did not have an) measurements on
th<- thousands ol examples of flakes that were the products of knapping.
In essence, we had only those Hakes that had been selected for use, either
in unaltered form as "utilized flakes" or slightly modified with retouch
,i- various types <>l scrapers <>i Hake kniv<

We <!<•( ided to i ompare a number of i\jk-s ol such items among them-
selves i" see ii there were significant differences in shape. I tilized flakes
were held .i- the "control" in the assumption that they would probably
be closest t<< the total range and variation in propei relative frequencies

96

PREHISTORY OF EASTERN ARIZONA, III

to the actual range of shapes available to the tool maker in the total popu-
lation of flakes possible with the knapping technology.

The results of these investigations do suggest a selection of flake shapes
for particular types of tools; but also suggest that "utilized flakes" may
have been selected with certain sizes in mind for certain tasks and, there-
fore, might have "skewed" our control. In addition, the factor of thickness

Fig. 70. Ovate scrapers made on flakes. Length of k is 8.4 cm.

aki ii \< rs

OVATE SCRAPER

■

4 5 6

. ■

1 i Distribution of dimensions of ovate scrapers.

of flake is probably important, but this dimension was not considered in
our original experiment. Future tests should control the shape and thick-
ness variation of the total sample of both tools and lithic waste debris.

Five types of tools were utilized in this experiment: utilized flakes,
side scrapers, end scrapers, flake knives, and ovate scrapers. Graphs por-
traying tin- shape and frequency of each type are presented in Figures
60, 63, 66, 68, and 71. Table 1 present- the results of the statistical tests
ot significance for the variations in shapes of these tool types as well as
summary figures of the range and variation in length and width for each
t\ pe.

Both the /•' and ! tests for statistical significance were employed
Moroney, 1956, p. 227f and 233f). The / test was found to be more
sensitive in this particular experiment. The results (Table 1) indicate
ral generalizations. The flakes -elected for ovate scrapers were sig-
nificantly different in size and shape from the Hake- utilized for other
tools. '1 hi- particular size and shape was in the mind of the knapper before
retouching the working edge. Also, the width of the flake- selected for
cud scrapei - was a signify ant factor. I hus, the width of the working edge
was .i < i itical aspe< t in the selection oi production of flakes for end scrapers.
This suggests that the specific tasks for which end scrapers were used were
quite specialized .\n^ these tasks were clearly in the mind of the knapper
in fashioning these tools.

cn o <n r- oo

J3

-C

0

Q

CO

O rn d fl Cl

2^

CO

2

CO

2

fa

oS

A

03

u

U

fa

CO

CO

D

w

w

*4

X
h
Q

i— i

z

<

Q

oo r-~ cn m ,—

< «

> be
O c

CO

2

> be:

O c

co
2

2

u

M N N (O to

fa

fa

fa
fa*

£

s

0

ri

M

s

z
<

X

JZ

0
fa

oo r- c\ vo oo

p to pm N

N (N M to to

z iS"

m in

o o

z "^

S£

CO

U4

u:

w

a

a

Oh

X

2

>H

►J *->

< -C

H

Ci, be

CO CO

fa 5P

CO CO

0

c

V

22

c

22

fa

fa

0

H

PONi'l't

Q

in m in r~- c\

«

co

>

i— i

fa

■M

2

X

a.

CO CO CO

222

"O
0.

CO

2

*-> in
o o

2

H

z

<
as
o

<

0

O

<

o \o to oo r-~

1— 1

H
<

2

5

oo o cs t-h <s

CO

CO

fa

tO-* Tf ^ Tf

q-5

«J3

i— i

<
>

co £P

CO

CO CO

" be

co

CO en

22

22

fa

fa

fa

0

z

CM CN Tf Tf tO

fa

U

ON N'ftn

U

s

■* rf Tl- ^t Tt

2

J3

-C

<

33

u.

co

CO

CO CO

Si

fa

co

CO

S2

u

p— 1

fa

zzzz

22

2

Q

Q

O

>— i

CO

Q

<

W

N

CO

w

-J
a,

2

ONOOOrH

CN ^ O <N rj-

U
N

H bC

fa1 a

CO CO CO CO

2222

w

N

h be

fa <u

co co
22

S£

(M ,-> ^H rn

fa

fa

-t-»

CO

fa

<

CO

a
a
o

fa

U3

<

'c

X

be

CO

1.

CD '—

>-

u

>.

h

CO

o

*— <

w

-1
n
<

w

Oh

><

H

^ rt c O «

S

e

s*
u

a

"S

fa fc

« n

u
3

s
s

3

'S

u

u ta

2

N CO y V CO

3

Si

o

M<

co £

o

^

co J3

CO

H

+-»

CO

V

u co

CO

CO

V

o co

2

^ 1; _* nj -q

V

4J

-ii

"3 T3

C/T

CU

_<:

rt -c

_

■m "0 _2 > j;

D io fa 0 fa

-g

M

0 fa

CD

^r

-

> c

<,

CO

fa

*-»

CO

fa

0 fa

98

Ak riFACTS 99

The shapes and sizes ol the other tools tended to be more homogeneous.
Significant variation among these types would probably be demonstrable
if the variable ol thickness had been utilized and II these types were com-
pared to the thousands ol Bakes which had not been sele< ted foi modifica-
tion into tools.

Choppei rhese are large, bifacially worked chopping-cutting tools.
In all cases the sharpened edges shovA battering and irregulai chipping
resulting from use. In man) cases, the unworked surfaces show battering
.i-. well, perhaps indicating their use as hammei ton* In summary, the
tools classified as choppers appear to be multi-functional implements
utilized in cutting and chopping operations as well as for tasks requiring
hammering.

'I he class labeled choppers was segregated into three types based on the
form ol the --tone -elected for modification. Since almost .ill of these items
were assigned field numbers and were not brought to the Field Museum
ol Natural History for further analysis, I have had to rely on the cards
prepared for each artifact by Miss Rose for the typology.

The first type consists ol water-worn stream pebbles with one end or
side trimmed bifacially with percussion flaking. A total ol 58 choppers
of this type was recovered from the excavations (Fig. 47, •

A second type consisted of exhausted or workable cores with one edge
retouched bifacially. All 95 examples ol' this type showed use chipping

and battering on the chopping edye.

A third type of chopper was recovered from the pueblo. These choppei i
were formed on large, thick Hakes with bifacial trimming forming the
working edge. All 37 examples of this type exhibited use chipping and
batterin

' ■ \ total of 59 cores was found at the Broken K. Pueblo. All hut

five of these were split cobble cores. A stream pebble, usually of chert,

elected and split in two. The weathered surface was then removed

using the tlat split surface .is a striking platform. Then llakes vvere struck

from all surfaces of the prepared core often employing prepared suikim;

platforms until the core was either exhausted and the nucleOUS discarded

oi fashioned into ,i tool, or in some cases the workable core was discarded
re it became exhausted, split cobble cores are illustrated in Figure
~2 in \ arious <>i use.

Nothing (an he said regarding tores in the prehistoric Southwest
because they generally are not studied, n cores are recognized, they are
usually simply noted .is present with no attempt al description or analysis
. \\ oodbury, 1 954, pp. 1 34 1

Fig. 72. Cores. Split cobble type (a, d-g); exhausted nucleous (c); and block core
(b). Length of g is 5.6 cm.

Kin

ar i ii \( rs 101

< >nl\ five examples of a second type of core were recovered. These are
block cores which are rectangular blocks ol stone from which lamellar
(lake- or blades were removed using one or two striking platform I

I >nc example ol ,i block core was found with a bifaciall) trimmed
cutting edge worked into one side (Fig. 47, !

Miscellatu ^ •• A A total of 72 stone artifacts was recovered

which do not fall within neat functional classes. For some of these items,
.! "ceremonial" use has been suggested b) various prehistorians, but for
man) of these spe< imens, there is no clear indication as to fun< tion.

Nine large, thin slabs oi sandstone which had been roughed into cir-
cular shape were found. The) ranged in size from 1 1 to 2~ cm. in diame-
ter. Six smaller discs were recovered which were also made ol sandstone,
ranging in size from 5 to * cm. in diameter (Fig. 73, b). Quite similar
items were reco\ ered at the Carter Ranch Site where the) also occurred
in two size modes (Martin et al., 1964b, pp. 78 81; fig. 34). Rinaldo
/• . . fig. 34 ) calls them "pot covers", but lists other inferred uses includ-
ing covers for storage pits. He (Ibid., pp. 78 81) presents details of the
distribution of these items in the Southwest.

Three small, irregular slalis of sandstone were found, each with a hole
drilled through the center (Fig. 73, c). Their use is unknown. A single
example of a thin piece of sandstone that had been roughly shaped into
a lanceolate outline was recovered. Its edges were unsmoothed and its
function is undetermined.

Three fragments of large, rectangular "tablets" of stone were found
at the site. The edges were smoothed and squared and both top and
bottom surfaces were well-smoothed. Their reconstructed sizes are ap-
proximately Id by 12 15 cm. Their use is not determined.

Three "medicine cylinders" of vesicular basalt were recovered Fin.
52, (/). < 'in- example of a similar shape and form produced in fine-grained
sandstone was found.

I wo -tone tools occurred that may have functioned in weaving opera-
tions. ' )ne was an elongated piece of fine-drained sandstone with a deep
>ve worked all the way around the item, lengthwise (Fig. 32, <). The

other was a "stone needle", a thin, lanceolate-shaped piece of smoothed

stone with one end pointed and the other end lihmted. A hole was drilled
through the blunted end I ig 32, e).

< me large andstone dis< was found with a hole drilled near its edge.
both surfaces ol the disc were smoothed and the edges were smoothed
.md rounded. Its use is unknown.

Maturall) occurring, unusuall) shaped stones were found at the site
1 hese included a fragment oi a circular stone concretion and two ex-

Fig. 73. Miscellaneous stone artifacts. Large sandstone disc with edges smoothed and
hole near edge, a; circular sandstone slab, b. Diameter of a is 15.7 cm.

102

I ig. 74. Miscellaneous stone, bone, and shell artifacts. Petrified wood "tinkler'
water-worn stone, shaped, b; stone "tablet", c, water-rolled stone, polished, d; quartz
crystal, e, stone tablet, incised lines on surface, '. irregulai piece of stone, edges squared
and polished, g; small polished stone table! irregulai pie. e ol shell, edges squared,

'.. polished portion ol bird long-bone, /. "medicine ilise." m; irregulai Hal stone, edges
squared, n. I <ength ol n is 2.9 cm.

104 PREHISTORY OF EASTERN ARIZONA, III

amples of quartz crystals (Fig. 74, e). These items probably functioned
in a ceremonial context, but their exact use is not known.

A number of small, shaped, and polished stone items were recovered
during the excavations. Their exact use is undetermined, but they prob-
ably were utilized in ceremonial or gaming activities. Six small rec-
tangular "tablets" made of sandstone with edges that were smoothed
and squared were found. The flat surfaces were unaltered. They were
all about 0.5 cm. thick and ranged in length from 4 to 5 cm. and in
width from 2 to 3 cm. (Fig. 74, b). A similar item was found made of a
hard, dark colored stone which was polished on all surfaces with two
incised lines on one face (Fig. 74,/). In addition, six examples of a much
smaller rectangular stone "tablet" were found. These were narrow rec-
tangular pieces of fine-grained stone with smoothed and squared edges
and all surfaces polished. They ranged in length from 1.5 to 3.6 cm.,
in width from 0.5 to 1.0 cm., and in thickness from 0.2 to 0.5 cm. (Fig.
74, h). Four examples of small, water-rolled stones which had been
smoothed and polished were found. There was some purposeful shaping
evident (Fig. 74, d).

Four examples of small stone discs were found. These were thin, cir-
cular pieces of stone with polished surfaces and squared and smoothed
edges. Diameters ranged from 2 to 4 cm. (Fig. 74, m). Their use is not
known.

A total of 17 small, irregular pieces of thin, flat stone was found. The
edges of these pieces had been squared and smoothed and their surfaces
had been polished (Fig. 74, n).

Two examples of stone "tinklers" were recovered at the site. These
were both fashioned from splinters of petrified wood and were notched
at one end (Fig. 74, a).

Similar miscellaneous stone artifacts were found at the Carter Ranch
Site. Rinaldo presents excellent descriptive notes and details of their dis-
tribution in the Southwest (Martin et at., 1964b, pp. 78-81; 96-97)

BONE, ANTLER, AND SHELL ARTIFACTS

Bone Awls.- — A total of 89 bone awls was recovered from the Broken K
Pueblo. These tools have been segregated into a series of types on the
basis of the particular bone selected for the implement and the nature
of the alteration. Bone awls were considerably more abundant at the
Carter Ranch Site where 120 were found (Martin et al., 1964b, p. 97).

There were ten bone awls produced on ulnae (Fig. 75). Six of these
were made on larger ulnae from such animals as deer and antelope

*

'"""^r ^5

r

r

i ig. 5. Bone awls, ulna type. Length ol i is 4.9 cm

105

Fig. 76. Bone awls. Split long bones and cannon bones, condyles intact. Length of
d is 9.6 cm.

106

I J. Bone awls made from splinters. Length >

107

108 PREHISTORY OF EASTERN ARIZONA, III

(Fig. 75, c and d), and one was found which was quite small (Fig. 75, e).
Three were recovered which either had the condyle removed or nearly
so (Fig. 75, a and b).

There were six awls found of the split long bone type with the condyles
unaltered (Fig. 76, a, b, d, and e). One split long bone awl had its con-
dyle abraded (Fig. 76, c).

The most abundant type consisted of bone splinter that had been
sharpened. There was a great range in size of the 68 awls of this type
(Fig. 77, a-h, j-l).

One awl was found that had been produced on the tip of an antler
tine (Fig. 77, i), and three awl fragments that were too small to place
into a category with any certainty were recovered.

Bone Needles. — Four bone splinters were found that had one end squared
with a hole drilled through and the other sharpened to a point. Two
were fairly large with lengths of from 15-17 cm. (Fig. 78, b and d), and
two were relatively smaller with a length of about 8.5 cm. (Fig. 78, e
and /). One bone needle is reported from the Carter Ranch Site (Martin
etaL, 1964b, p. 99).

Weaving Tools {?). — Seven bone tools were found made on splinters
or split long and "cannon" bones that may have functioned in weaving.
The points of all these implements were purposefully blunted and there
were no perforations present (Fig. 78, a and c).

Miscellaneous Bone Artifacts. — A total of 14 bone items was found that
did not fall into a system of typology. This category included such items
as splinters of bone that had been smoothed, partially split long bones,
fragments of deer ribs that had been polished, and pieces of split long
bone shafts that had been polished and carved (Fig. 74, /).

There were many fewer bone tools found at Broken K Pueblo than
occurred at the Carter Ranch Site. Perhaps of significance is the lack of
the Grooved Bone Awl type at the Broken K Site, a type that was found
in some abundance at Carter Ranch Pueblo (Martin et al., 1964b, p. 99).

Miscellaneous Shell Artifacts. — Five small, irregular fragments of shell
were found, all of which had squared and smoothed edges (Fig. 74, k).
Their use is not known.

Antler Flakers. — A total of 14 antler-tine flaking tools was recovered.
All of these implements had beveled or rounded ends and all exhibited
evidence of battering (Fig. 79, c and e).

Antler Wrench. — These items are basal fragments of large antler shafts
which have been split and one, two, or more holes drilled through. The
holes in these tools are approximately 1 cm. in diameter. Three antler
wrenches were found at Broken K Pueblo (Fig. 79, a, b, and d).

r*

I

r*

■ *

f

- Miscellaneou bone artifacts. Blunted tools, a and c; needl<

in.

Fig. 79. Antler flakers (c and e) and antler wrenches (a, b, d). Length of c is 15.5 cm.

110

Ak riFACTS Ml

CEREMONIA1 IT! MS, I >K\ Wll \ I S,
l'I< All NTS Wl » MINI R \l S

A total ol about 500 items of this general category was found .it Broken
k pueblo. The) have been si ted into classes and, in some cases,

into typ

/■ The 22 ornaments of this class have been classified on the

basis oi material (itiii.iniir.nimc: bone, shell, stone, or pottery.

'I luce pendants made from bone were recovered: one was .1 section
.1! the shaft ol .1 long bone with both ends beveled and smoothed, and .1
deep incision made .ill around one end of the shaft I one 1 on-

sisted of a thin, small spall ol bone which had been shaped into an oval

"teardrop" shape with .1 small hole placed in the smaller end I
. and the other bone pendam was .1 flal piece ol bone which v.
carved into symmetrical shape. 'I he last has .1 rectangular center section
with a single rounded projection on either side and .1 small hole drilled
in the top of the central portion Fig. 80,

Twelve pendants made from shell wen- found. ( >f these, one was ol
.1 "teardrop" shape with a hole drilled in the smaller end (Fig. 80, if).
< me consisted of .1 small bivalve shell (probahh Glycymeris maculata) with

hole chilled at the "hinge" (Fig. 81, //). Five pendants made from
conical shells were found. Most of these appear to have been made from
* ' nus) ximenes, but at least one was produced on Turritella

ma. These pendants exhibited little modification beyond a single
perforation or notch .n\<\ smoothed edges (Fig. 81, e, t. and /). One was
produced on a triangulai section from a large, unidentifiable shell. I
edges are squared and smoothed and a hole appears at the apex. A
zoomorphic figure has been carved or etched on the convex surface
I . v". I our fragments of shell pendants were found.

\ single example ol a stone pendant occurred. It was fragmentary,
but appeals n> be a thin piece of stone having irregular outlines with
smoothed edges and a hole drilled at one edge.

six ceramic pendants were found at the site, lour ol these were cir-
1 ul. u in outline, with squared and smoothed edges and with a single hi

illed near the edgi I ' 80, g). Two of these were Show L.OW Blaek-

on-red, one was an indeterminate black-on-white, and one was a brown
ware with a polished smudged interior. I wo examples were sub-triangular
in shape with squared and smoothed edges and a hole placed at the
smaller end. One was Snowflake Black-on-white, Ha\ Hollow Variety
I and th«- othei was made on a plain, brownware sherd with

,i smudged interioi \ geometric design has been incised on the exterior
surfaci I

112

PREHISTORY OF EASTERN ARIZONA, III

Fig. 80. Pendants. Bone pendants, a, b, e; shell pendant, d; carved and etched shell
pendant, c; incised sherd pendant (Plain Brown Ware, smudged interior), /; sherd
pendant (Show Low Black-on-red), g; sherd pendant (Snowflake Black-on-white,
Hay Hollow Variety), h. Length of h is 3.8 cm.

Beads. — Approximately 400 beads of various kinds were recovered.
These included about 30 beads of whole Olivilla sheels. But the majority
of the beads were of the small disc type of either shell or stone. Approxi-
mately half of the beads found at the site were in association with burial
number one from room 27. The remainder were recovered from trash,
floors, and pits.

Shell Bracelets. — Four examples of bracelets were found. All were frag-
ments and all were produced on Glycymeris shell. One example had a

g

Fig. 81. ( M 11. it units ,,i bone and shell. Bone rings, a d; shell pendants, ?,/, h, i; shel
bracelet fragment, i . Length of - is 2.3 cm.

Pigment (red hematite) and irregulai lumps ol turquoise, i and i.

11 5

Fig. 83. Bone ring material. Length of a is 4.5 cm.

114

ARTIFACTS 115

raised portion at the top of the bracelet in .1 triangular form, and an
incised geometric design on the bod) I ig 81, g).

r Nine examples ol pigment lumps were found. Seven ol

these were irregular hunks ol red hematite with one or more worn and
smoothed facets I i B2, b and 1 I ( toe example each ol azurite and
yellow limonite were excavated.

/..'.. . Four irregular lumps ol turquoise were recovered, rhere
was mimic evidence of smoothing <>n each, bul there wen- no perforations
inn ,tn\ definite shaping Fig. <s.\ a and </). All but one example were
found in floor context.

Isbestoi A single, irregular lump ol this mineral was found on the
floor ol room 6< '

/.' • Rings. These items are circular sections cut from long bone
shafts with beveled edges and generally smoothed surfaces ; I 'ig. 81, a d).
1 hree whole examples were found and 12 fragments were re< overed. ( >ne
ol the latter had an incised geometric design (Fig. 81, . I. < )ne fragment
of a ring made of shell was found. Rinaldo presents a detailed discussion
of hone rings and points to their possible use .is beads rather than finger
rings Martin ei a!., 1964b, pp. 93-94).

In addition, 42 examples of long hone fragments from which the bone
rings had been cut were found. Often, there were incisions marking uncul
rings and the end of the shaft shows a beveled cut. A total of 25 was re-
covered with the conchies intact and 1" which had the condyles removed
and rings removed from one or both ends of the shaft (Fig. 83).

6 •• Tinklers. — Eight tinklers were found at the site. They consist of
short sections cut from the shafts of long hones near the articulating end

Their shape is roughly that of a truncated cone. All examples were hol-
lowed with ends squared and smoothed. A small hole appears through
the flaring end, ne.tr the edge (Fig. 84, a 1 >.

Painted Bom Items. A total of 11 examples of various hones which
\ia<{ been painted was found. It h not known if these items functioned
in gaming 01 in ideological activities such as divining. All hut four were
found in the till of kivas. Eight were scapulae of deer ami antelope with

netric designs painted in black on both surfaces. The designs tended

lo be wav) lines, bars, broad lines, and circles (Fig. 85, ' and . I. I wo

oients ol ungulate mandibles with geometric designs painted with

red and hlack paints were recovered. < >ne example of a skull fragment
..1 Lepus californicus was found with hlack paint applied to the skull vault.
'I he design was indetei inmate > I iu. s.">, ,/).

Fig. 84. Bone tinklers, a-c; notched sherd (Show Low Black-on-red), d; sherd ladle
(St. Johns Polychrome) e. Length ofe is 12.0 cm.

116

I Painted bone. Skull fragment ol /./..> Calijornicus, a; fragmentary scapulae

.Hid i .

117

118 PREHISTORY OF EASTERN ARIZONA, III

MINIATURE VESSELS, WORKED SHERDS,
AND OTHER CERAMIC ARTIFACTS

A total of five miniature vessels was found at the Broken K Site. The
vessel forms were pitchers and jars and one miniature ladle was un-
earthed. They were quite similar to those found at the Carter Ranch
Site (Martin et a/., 1964b, p. 104).

Miniature Pitchers. — Two miniature pitchers were recovered. Both of
these were Snowflake Black-on-white, Snowflake Variety. In both cases
the handles were missing (Fig. 86, a and b).

Miniature Jars. — Two miniature jars were found. Both examples were
brown indented corrugated (Fig. 86, c).

Miniature Ladle. — One miniature ladle was recovered. It was a plain
brown ware ladle and was in fragmentary condition.

Worked Sherds. — A total of 131 worked sherds was found at the Broken
K Pueblo. This class of items was further segregated into four types on
the basis of such attributes as the nature of the worked edge and the
shape.

Type I: Sherd Blanks. These are sherds roughed into circular or other
regular shapes with no evidence of use or smoothing. The edges are rough
and appear to have been shaped through spalling. A total of 14 such
blanks was found (Fig. 87, f-j).

Type II: Sherd Blanks in first stages of use. These sherds exhibit slight
and sporadic patterns of wear on the edge. Most of them are circular
and appear to be roughed into shape prior to their use. The wear pattern
is beveled indicating their use in scraping activities. A total of 25 sherds
of this type was found (Fig. 88,/-/?).

Type III: Pottery Scrapers. These sherds have edges that are well-
worn through use. The pattern of wear is beveled indicating a scraping
function. A total of 54 such tools was recovered of which 34 were circular
in outline and 20 were roughly rectangular (Fig. 88, a-e).

Type IV: Pendant Blanks. These are well-shaped sherds, generally
with a circular or rectangular outline. The edges of these items are ground
to a straight or squared profile. There is no evidence of wear on these
sherds; they were not used as scrapers. They are identical to the sherd
pendants (above, pp. 111-112) except for the absence of a perforation.
A total of 38 sherds of this type was found (Fig. 87, a-e).

Not a single example of a pottery disc with a central perforation
("spindle whorl") was found at the Broken K Site. A total of 14 such
worked sherds was found at the Carter Ranch Pueblo (Martin et al., 1964b,
p. 104).

J*

^••^•fr f

I 36 Miniature vessels. Pitcher, Snowflake Black-on-white, Snowflake Variety,
. pitcher, Snowflake Black-on-White, Snowflake Variety, '. Brown Indented Cor-
I [eight of b, 6 < m.

11"

Fig. 87. Worked sherds. Pendant blanks (edges ground to square profile), a-e; sherd
blanks (roughed into regular shape, no evidence of smoothing), f-j. Length of j is
4.1 cm.

120

\R I II \( I S

121

l -- Sherd scrapers. Scrapers with edges beveled from use, a t; scrapers in first
ol wear, I h. Length of h is 5.2 cm.

/ I wo sherd ladles or scoops were found. < me was made

from a liin sherd of a large St. Johns Polychrome howl. The edg<
the sherd had been i arefully shaped and smoothed I ig. 84, • i, I he other
example was fragmentary; it had been produced on an indeterminate
black-on-white sherd.

,\/: ellai Sherd 1 One irregular sherd ol Show how Black-on-

Red v\ it h on<- <t\'j<- beveled and smoothed was found. This edge had .i
series ol regular, shallow notches worked into it. Its use is not known:
n could be a tool foi grooving or stamping clay (no examples of stamped

122 PREHISTORY OF EASTERN ARIZONA, III

or grooved ceramics were found), or it might have served in plant proc-
essing in some fashion (Fig. 84, d). It is similar to two such items recovered
at the Carter Ranch Site (Martin et al., 1964b, p. 105).

CORDAGE

Two short fragments of cordage were found in a jar recovered from a
pit in the floor of room 33. They are small pieces, but both appear to be
Z-twisted cord or string.

LITHIC WASTE

A total of 5868 items of lithic waste was found during the excavations
at Broken K Pueblo. These include both flakes and chips as well as
"shatter" that were discarded and not utilized in the various stages of
preparing stone tools.

In general, lithic waste is not considered an important part of the data
from prehistoric sites and as such is often not collected in the field. When
it is collected, often it is simply noted as present and then discarded. As
far as I can tell, this lack of attention to workshop debris is based on the
a priori assumption that there is little to be gained from an analysis of
such items compared to ceramics and stone tools to augment the picture
of an extinct culture. In so far as I know, there has never been a syste-
matic and thorough analysis of such debris from a prehistoric site in the
American Southwest.

Binford and Quimby (1963, p. 277) point out that "Since tool pro-
duction is a process, the techniques and motor habits of which vary
stylistically and according to their relative efficiency, it should follow that
variations in processes of tool manufacture are as important to our under-
standing of extinct cultural systems as the variations in the tools them-
selves." They further suggest (Ibid., p. 278) that isolating the way in
which the various steps in tool manufacturing were carried out, by what
social groups and at what locations at the site, is extremely important to
an understanding of an extinct cultural system.

The nearly 6000 items of lithic waste from Broken K Pueblo were
sorted into categories and counted in the field under the direction of
Mr. John M. Fritz. His classification and counts form the basis for the
analysis reported below. I stress the fact that this analysis is very crude
and incomplete. But even at that, the suggestive findings would seem to
indicate that further, more sophisticated analyses are in order.

There were no metrical observations available on the debris. The field
sorting was in terms of a series of classes of waste such as "core prepara-

rOTALS

\ ■::

Stage

I

II

111

hil \1 -,

\k I II \< I S

2 1 Kl Q\ 1 V [ES I >I I.I I HIC u.\s i i BR< >KI \ K PI I BL< I

■ '■ i

Fill 1

I 338

II 1160

III 312

185o 1226

1 ill Floor

73

151

59

West Wing

1 ill 1 loor

■<■ I 152 149

Stage II 429 416

Stage III 109 68

totals 690 633

South 1 1

Fill Floor

Stag I 10 42

II 22 161

Stage III 3 36

roTALs 35 239

itter", '"tore preparation flakes", and these categories were further
sub-divided on the basis of fine-grained and coarse-grained raw material.
Limitations of time imposed die necessity for lumping certain of these
categories. When this was done, a gross, three-fold classification resulted.
All lithic debris was segregated into three major classes: core preparation
material: primary shaping debris; and secondary shaping debris.

The core preparation material includes both Hakes and shatter re-
sulting from the preparation of cores for the production of flakes for tool
manufacture. Included in this category are decortication flakes and pri-
mary shatti White. Binford, and Papworth, 1963, p. 3; Binford and
Quimby, 1963, pp. 286 288). The primary shaping debris includes both
unutilized flake blanks and secondary-flaking rejects (White, Binford and
Papworth, 1963, p. 7; Binford and Quimby, 1963, pp. 296 299). Secon-
shaping debris include- chips and small flakes resulting from both
primai \ and secondary retouching of flake blanks. These three categories

North

East

South

West

Wing

Wing

Wing

Wing

23

19

7

7

42

49

27

21

13

22

6

3

124 PREHISTORY OF EASTERN ARIZONA, III

represent three major stages in the manufacture of stone implements.
The first step is the preparation of the core; the second, the striking of
flakes from the core on which will be produced tools (projectile points,
drills, scrapers, etc.); the third step sees the actual shaping and finishing
the tool on the flake blank. Hereafter, simply for convenience, these

Table 3.— MEAN FREQUENCIES OF LITHIC WASTE PER ROOM,

BROKEN R PUEBLO

Floors

Stage I
Stage II
Stage III

three categories of debris will be referred to as: Stage I, Stage II, and
Stage III, respectively.

For this analysis, the site was treated as consisting of four major room
blocks or wings: North Wing: (rooms 24-40); East Wing (rooms 1-23);
South Wring (rooms 72-91); and West Wing (rooms 41-71).

Table 2 presents summary figures for the actual frequencies of lithic
waste by wing for all floors and for lumped fills. From these data it is
clear that: (1) in any room, stage II debris has the highest counts from
both floor and fill; (2) in all wings but the east wing, stage I is second
in abundance; in the east wing, stage III is the second most abundant
class; (3) the north wing has the most material representing stage I,
followed by the east, west, and south wings in that order; (4) for both
stages II and III, the order is east, north, west, and south wings. Arguing
from total density figures, the order is east wing, west wing, north wing,
and then south wing. If only floors are considered, the order is somewhat
different: east wing, north wing, west wing, and then south wing.

Several generalizations are suggested from these data. Arguing from
gross density of lithic waste, by far the most knapping activity was carried
out in the rooms in the north wing. Looking at mean frequency of waste
materials on floors per wing, the following pattern emerges: north wing
and east wing had between two and three times the number of waste
items on floors compared to rooms in the south and west wings. This
bimodal distribution suggests that two to three times as much knapping
activity was carried out in the north and east wings as was true for the
south and west wings. This has special impact when compared to the
distribution of certain classes of tools which functioned in the preparation
of arrows and points, such as antler flakers and wrenches, shaft smoothers,
and so on. These items, too, have a patterned distribution at the site.

\K I 11 \< Is 125

From these data, I have argued elsewhere Longacre, 1966) thai there
might have been a pattern ol reciprocal exchange in operation at the
site. I In- waste debris counts might add support to that suggestion, but
is no cleai correlation except for the north wing which is high in
l>uili such tools and litlm- waste

I : ■ s pattern held true when the counts were broken down l>\

stage ol manufacture as well. That is. no one wing was high in I

debris and low in waste representing stages II and III liable presents

me. in i: counts l>\ wings and l>\ stages ol manufacture. Essentially

the same pattern held for till counts as well.

1'.. far the most striking pattern was observed when location <>! the
rooms in terms ol wings of the pueblo was ignored, and room type was
examined. Hill 1965) has defined two types of rooms .it the mic on the
basis ol such attributes as floor area, presence of floor features, and arti-
fact densities. These he labels storage rooms and habitation rooms. A
third type, ceremonial structures or kivas, was also defined. When floor
densities of lithic debris were examined using this typolog) a striking
pattern i merged.

I < 23 habitation rooms at the site had a mean densit) of lithic debris
- items. Hie 25 storage rooms had a mean count of only 30. This can
l>e further refined by noting that eight of the storage rooms had floor
features. The average count for this sub-type of storage room was 73,
a figure nearh as high as the habitation room mean. The remaining 17
featureless storage rooms had a mean density of onl) In items on their

I rom these data, it is clear that knapping activity was essentially con-

: lain rooms in the pueblo, and that these tended to occur in

the north and east wings. High counts were also noted on the flooi

kivas, except for the plaza kiva. This, in turn, ma) suggest that such

divides tended to be carried out in both the habitation units

and the kivas during the earlier occupation of the site. Then

in the later period of the site's occupation, knapping was not carried out

in the large, central kiva, but only in the habitation units. This ma)

indicate that such male activity was at first carried out by groups ol men

in both habitation rooms and kivas, a pattern that became somewhat

altered in the latei periods when knapping appears not to have been

ied out in the large central kiva. This ma\ have implications for our

understanding of the changing nature ol village integration (cf. Long-

.n re, !

IV. Pottery

By Paul S. Martin

Excavations at Broken K Pueblo, Carter Ranch, Eastern Arizona
(Martin, Rinaldo, Longacre et al., 1964) brought forth a total of 26,082
sherds and 28 whole or restorable vessels. Twelve of the whole vessels
came from pueblo rooms and 16 from a burial of an adult male (grave
placed in plaza and over it, later, room 27 was built.)

The following description is given in the usual taxonomic fashion.

The listing of pottery types and a tabulation of sherd totals for the
entire site regardless of location or level follow this discussion. Since
Longacre (in Martin, Rinaldo, Longacre et al., 1964, pp. 110-122); has
given a description of the major decorated pottery type — Snowflake
Black-on-white, and its varieties — I shall not duplicate that. Citations for
all types are also given in the volume.

The design elements of Snowflake Black-on-white pottery were ana-
lyzed by Mr. Stevens Seaberg, artist, formerly on the staff of the De-
partment of Anthropology at this museum. One hundred seventy-nine
elements were recognized. Some of these were used by Hill (1965, manu-
script) in his analysis of their spatial distribution in order to set forth
inferences concerning the social organization of the Pueblo. The distri-
butions of the design elements were related to function of rooms, the loci
of social groups, and temporal distributions.

A complete discussion of the uses of the pottery types and the design
elements is given in Chapter V of: Broken K: A prehistoric society in eastern
Arizona by James N. Hill, dissertation submitted to the Department of
Anthropology, University of Chicago, 1965; to be published by Field
Museum of Natural History in Fieldiana: Anthropology.

The whole pots (except for those with the burial) all came from the
north, west, and south wings. Their exact location, shapes, names, and
chronological placement are given in the listing at the end of this Chapter.

Binford has called my attention to the fact that taxonomic classifica-
tion varies in usefulness, depending on the problems being studied.
Analysis of the variations in forms — manufacture, shape, designs on the
finished products — might help to understand the function of the pots, and

126

POT] IKV

127

ocial context in which the) were 1 reated. Indeed, such .1 stud) might
give us clue- .is tu i h<- social structure oi a given village; might help explain
processes ol social change, and give us some hints as to the articulations
xtinct cultural s^ stems.

89. Bowl. Cat. No. 282877. Snowflake Black-on- white; Hay Hollow vai
associated with burial «1.

\ stud) on the morphological characteristics of potter) as suggested
b) Binford was begun b) Phillips (1965 M.A. dissertation, University
ol Chii ising pottery from Broken K. Pueblo.

I :iv functional analysis was hampered l>v three difficulties: I.) The
textured potter) had been classified in the traditional manner and (limit-
ed, but the sherch were discarded, .is had been our custom for 30 years.
Thus, thousands ol sherds and inferentiall) hundreds ol shapes and si/(s.
were ex< hided from the stud) because of adherence to tradition. 2.) I he
numbei ol decorated sherds (Snowflake Black-on-white [all varieties],
Show-Low Black-on-red) that wen- large enough to show a complete
design element and vessel shape was small (4500 sherds); 3.) and when
numbei was again cut b) using decorated sherds from floors only,
urn sample size was disappointingl) ver) small. In spite ol these draw-
bat me useful data were ret orded.

Fig. 90. Bowl, Cat. No. 284074, Springerville Polychrome; room 92; floor #1; bin.

Fig. 91. Bowl. Cat. No. 282971, St. Johns Polychrome; room 48; level A.

128

POl II RY

A lew significant conclusions are herewith given:

1. Black-on-red bowls were more than twice as abundant .1- Bl.uk-

oii-w hite on<

2. Black-on-red jars and pitchers were not present at the Broken K

P

I

I 2. Bowl, Cat No. 282974, St. Johns Polychrome; room 53; il<><

3. Black-on-red bowls are more frequent in late rooms than are Black-
on-white bowls. This may be a functional replacement as Black-on-
white jars arc abundant in both earl) and late rooms.

\ male who was from 18 to 30 years old was buried in a deep 1 m.)
pit under the south wall ol room 2". It was very clear that the interment
had taken place before that room was built, and thai the burial was
placed in what had formerl) been part of the plaza.

In addition to a bracelet and necklace of stone beads, 16 potter) vessels
were found with this burial. A lisi of the potter) follows:
I >ow Is; Si. Johns Black-on-red

1 jar; Snowflake Black-on- white; lla\ Hollow variet)

1 howl: Snowflake B on-white; Ha) Hollow variet)
Snowflake Black-on- white; Snowflake variet)

1 howl: Snowflake Black-on-white; Snowflake variet)

Fig. 93. Bowl, Cat. No. 283476, St. Johns Polychrome; room 69; level B.

Fig. 94. Bowl, cat. No. 283477, Pinto Polychrome; room 69; level B.

130

I Bowl, Cat. No Querino Polychrome. Associated with burial 1.

I i ,1,1 Cat. N ; Snowflake Black-on-white : Carterville variet)

a BO, floor "1 .

1 si

132

PREHISTORY OF EASTERN ARIZONA, III

1 bowl; McDonald plain corrugated
3 bowls; McDonald indented corrugated
3 jars; Brown indented corrugated
1 jar; Patterned corrugated.

No midden or trash areas, as such, existed, although we searched for
them on all sides of the pueblo, and in the plaza. A few sherds in ash
soil were noted on the east side of the pueblo but the layer was thin (5-10
dm. thick). This deposit overlay the sandstone bed rock which, every-
where around and within the pueblo, outcrops at the surface or is buried
by only a few centimeters of soil. Bed rock was found in the rooms some-
times at, or just below, floor levels.

If a thicker layer of trash existed centuries ago and has been washed
away, we found no evidence of it.

Although Broken K Pueblo is larger (by over 50%) than Carter Ranch
Site (Martin, Rinaldo, Longacre et a/., 1964) about 30% fewer sherds
were recovered from the former. (Broken K Pueblo, about 26,000 sherds;
Carter Ranch Pueblo, about 34,000). All dirt was sifted at Broken K
Pueblo, none at Carter Ranch Pueblo; and yet fewer sherds at the former!

No adequate explanation for this discrepancy is at hand. I suggest
that the difference in quantity of sherds might be explained, 1.) by a

Fig. 97. Jar, Cat. No. 282879, Snowflake Black-on-white: Snowflake variety;
burial #1.

POTTERS

133

I - fai Snowflake Black-on-white: Carterville variety; room 69; levels B I

and flooi I; parts ol the rim from room 4. level B

shorter occupation; 2.) by the manufacture ol less pottery, since potter)

have declined in importance due to difference in storage and ci
ing techniques, or due to the fact that there ma) have been less food to
ok.
Along with these explanations, I would note that onl) two burials v»
found at Broken K. Pueblo, as against 34 at Carter Ranch Pueblo. Where
interments w< not known.

I l-I ( )1 Win i] I ( >R
l'l I Bl I I

K1>K)RI.I) POTTERY VESSELS, BROKEN K

Bowl Snowflake Black-on-white, II. i\ Hollow variety ited with adult

male bui tal i i noi tli w .ill of room I

Bowl Springerville Polychrome; loom n2. floor, .it west ed| inns

I
Bowl St. fohn Poly( hromi room 18; fill, level A 182971 ) (lati I

Bowl Si foluu Poly< hrome; room 53, flooi I 1) (late) I

Bowl St fohns Polychrome room 69, fill (lev. B ti 1

Bowl Pint irome; i el B 177) (lat< I

Bowl Querino Poly* iroi ted with adult male burial I 1); undei south

1
( an teen Snowflake Black-on-white, Carterville variety, room 80, flooi I: (284075)

I
|.u Snowflake B i-white, Ha) Hollow variety; associated with adult male

burial ah wall ol room l

Fig. 99. Jar, Cat. No. 282970, Snowflake Black-on-white: Hay Hollow variety;
room 69, east wall trench.

Fig. 100. Jar, Cat. 282868, Snowflake Black-on-white: Snowflake variety; room 41;
floor #1 ; Pit A.

134

Sgji^

101. Jar, Cat. No. 284076, Snowflakc Black-on-white: Snowflake vari
room 19; level B.

I .| Cat. No. • Snowflakc Black-on- white : Snowflake variety;

1 1 "in i 33, pit • >.

135

136 PREHISTORY OF EASTERN ARIZONA, III

Jar — Snowflake Black-on white, Showflakc variety (also called Tularosa and/or

Pinedale Black-on-white?); room 69, fill, (levels B, C) and floor; and rim
portions from room 4 (across plaza) fill (level B); (283475) (late). As noted
here, pieces of this jar came from two different areas separated by the
width of the plaza. Why the pieces were hurled in this fashion is not known.
Obviously, it may represent a part of the terminal occupation of the pueblo
(Fig. 98)

Jar — Snowflake Black-on-white, Snowflake variety; from trench along inner

east wall of room 69 (282970) (early?) (Fig. 99)

Jar — Snowflake Black-on-white, Carterville variety; room 41, pit A, floor (282868)

(late) (Fig. 100)

Jar — Snowflake Black-on-white, Snowflake variety; room 39; fill (level B);

(284076) (late) (Fig. 101)

Jar — Snowflake Black-on-white, Snowflake variety; room 33, floor, pit O (283245)

(early) (Fig. 102)
— Snowflake Black-on-white, Snowflake variety; room 33, floor, pit O (283244)
(early) (Fig. 103)

Jar — Snowflake Black-on-white, Snowflake variety; male burial (#1); (282887)

(Fig. 104)

Fig. 103. Jar, Cat. No.
room 33; floor #1, pit O.

283244, Snowflake Black-on- white: Snowflake variety;

Table 4.— SHERD* TOTALS FROM ALL ROOMS (FILL AND FLOORS),
FROM THE PLAZA, FEATURES OUTSIDE OF THE PUEBLO PROPER
AND FROM THE FOUR OUTLYING ROOMS— BROKEN K PUEBLO,

Decorated Pottery Types

Houck Polychrome
Kiet Siel Polychrome
Pinedale Black-on-red

No.
4
1

1

POT] ERY

137

• Seed jar, Cat. No Snowflake Black-on- white: Snowflake

vai ial I .

11 k-un-w hite
laic Polychrome
■ i Polychrome

Polychrome
is Black-on-red
St. John Pol; ■ hrome
Show l.uw Black-on-red

Bl ick-on-white, Snowflake variet\
Snowflake Black-on-white, lla\ Hollow variety
sii<>\\ i!.ik<- Black-on-white, Carterville variet)
^ 1 1. .\% on-white, Broken K variety

Black-on-white, Tularosa variety
Springcn ille PoK i hrome
ii-w hite
Wii m-red

I tei unii.it>' Black-on-red

U i minate Blai k-on-w hite
I terminal R
I terminate White-on-i ed
Indel ite Poly< hrome

b- total 1 >< •< hi . 1 1 f-< i I \ p

19

21

115

1
14

-
•■. n plain > on ugated

■■•II plain nudged mt<-i m Is i

Brown indented < hi ■

\

8101

138 PREHISTORY OF EASTERN ARIZONA, III

Brown indented corrugated, smudged interiors (bowls) 595

Gray indented corrugated (jars) 261

Gray patterned corrugated (jars) 10

Incised corrugated (jars) 4

McDonald corrugated, plain (bowls) 74

McDonald corrugated, indented (bowls) 722

McDonald corrugated, patterned (bowls) 38

Orange indented, corrugated (jars) 12

Patterned corrugated (jars) 283

Patterned corrugated, smudged interiors (bowls) 31

Punched, smudged interior (?) 2

Punched, corrugated (jars) 6

Red indented corrugated, smudged interior (bowls) 2

Sub-total— Textured Pottery Types 1 1 ,578

Plain Pottery Types No.

Alma plain 376

Alma smudged 9

Forestdale plain 971

Forestdale smudged 144

Forestdale red 24

Lino Gray 62

Reserve smudged 28

San Francisco red 1

San Francisco red, Smudged interior 12

Woodruff smudged 25

Indeterminate 51

Sub-total — Plain ware types 1,703

Grand Total 26,082

* Complete breakdown of all sherd counts by types and levels is published in Ar-
chives of Archaeology No. 27, Society for American Archaeology, and the Uni-
versity of Wisconsin Press, 1966.

Note: All the plain pottery types came from pit houses, the floors of which were
about 2 m. below the present surface. These earlier houses were beneath rooms 33,
48, and 41-kiva. Since the plain wares did not pertain to the pueblo proper, Hill
omitted them in his counts.

\ . Dating of Broken K Pueblo
/;. Paui S M vrtin

Sixteen samples of wood or charcoal were selected for dating out ol
the 35 obtained.

Six were sent to Isotopes, Inc.. Westwood, New Jersey ; five to Dr. J I
\ igel, Director, Natural Science Laboratory, Royal University, Gronin-
gen, The Netherlands; and five to Dr. Bryant Bannister, Laboratory ol
rree-Ring Research, University of Arizona, Tucson, Arizona. We grate-
fully acknowledge the aid we have received from these laboratories.

lour samples sent to the Laboratory of Tree-Ring Research were not
datable.

ie sample from Room 11 — a north-south roof beam was cut into
three pieces and sent to the three laboratories. The three dates from the
three laboratories on this one room are close (GRN-4555; ULC 75-76;
I 15'

The -ample numbers, laboratory, provenience, center dan-, one
standard deviation date-, and relative dates are given in Table 5.

My colleague, Dr. Hill, ha- dated Broken K Pueblo (Hill, 1965,
Chap. II) at A.D. 1150 1280. These date- were based primarily on four-
teen potter) types that have been previously dated by association with
-ring dan-- and on ('-14 date-.
The range of the assigned dates for these pottery-types is A.D. 1050
to A.D 1385. For statistical reasons, Hill divided these types into
two groups: 1) relatively cub. A.D. 1050 1250; (2) relatively late,
\ D. 1250 1385. He then averaged the dates for each group and this
ive him a nie.in early date of A.D. 1175, and a mean late date of
\ h 1283 He felt that these dan- were fairly good considering the limi-
tation- of this dating method. 1 Ie was encouraged when the\ cone-ponded
well with the sue radio carbon date- from Isotopes, Inc. and the single
tree-ring date; but, was puzzled b\ the initial four date- from the ( rronin-
gen Laboratory, Hie Netherlands. The latter date- are too early, and
Hill felt that they wen- probably not valid.

I < repan< \ between four of the Groningen date- ami those from

I ..tope- laboratory cannot be hilly explained.

1 VI

tfl o

U -M

0 rt

•S 2

J-S

II J

Z "i

s* u

:>

10

S
CO

1

10
|(D

CM

—in

o §•

1

. o

K

•— ** tO

c "
.2 II

5

10

10

O

8^

10 L

n1

h

1=

0)

in

.2

> IT

C

«j C

•« 2

5

0

0

<*

13 H

0<N

ID

en

1

ifM

01
— m

00)

1

-a

K

™ c

to O

5

0

0

n

N

Of)

CVII

1

J CO

Q

0)

— in

0*

en1

1

E

^"^ t*-i

< ,

0

0

in

0).

2

1

100

0)

in

1

o w

CL

>

to 'S

1

0^

K

0
(0 1

-i|--

U ■

boD
u

CO 1*

2

0

1 °

Z«i

V

-w to

.2 be

Or-

10 1

(j. in

0$

o-
cr

oil

a.c

5
0^

o-

0
(0

— 1—

1

5

— in

V I

5

Oco

10

I

to

2«
„co
0«

J*3

ON

a.

I

0)

O 0

+J

KIVA

UNDER

ROOM

41

10

I

I1"

0«

2 S

^ 2

2

I

01

0 II

5

ON
On

0

CO,

gi

H-

0
10)

CD

Cl n

(D*

E/5 i_

to *^
bo to

5

o

1

0

ZO

Is

00
0 CM

I

1

i

1 1 1

i i

I

s —

CO

1 1

H—

1

— u

N

1 1

(D?

£ cd

2 ^

X CJ
co _c

0

0

0 0

0

8

01

0

0

■4-)

0

0

to

0 0
PI -

0
0

0

n

0

s-1 ^

o ^

a v

SdV3A

. H

LD

o c

rl V

. bo

be fl

• — ( »2

M O

s-

o

140

— a

3 i

n —

-

Q -~

- =

--

:- =

= -'

-

—

.

z

—
£ 5

- z
-

Z
<

z

-
-

-

-

Id

_

—

x

-

-.
/

s-

>-

j>

.

-

—

m

■—

—

—

-

-

«

—

_

-

-

-

-

i/-

—

—

»™

r

in

ri

■•

*—

r

■"■

' —

c

-

-

-

-

Q

Q

Q

<

<

<j

---

<j

«jj

.--

I/-.

r

1/1

—

r

12

3

"■"

■~~

'

■

~~

-H

+1

+1

-H

r

c

—

r

.<-.

—

—

u-,

r

c*i

—

tN

n

cs

~

x

^

— »

^'

— '

—

Q

~

w

—

*■■

—

"~~

~~

—

--•

<*

^

^"

rf

^*

c

f

1

Z

ed

,g

- _

- 0

- c

id

C
-
-

y n
i X

—

Z C

E :
~ E

id

0

-

rv

-.-'■

> u

4-1

"a

Id

u

. . -

j£ —

_ c
• -

—

-

r

i

_ -

C]

T

r 0

~ .

n

— w

E-

z -

: —

= -

3 —

—
E -

~ -
— U

E

3

SI

- 0

— £

- -

0 r

zt -3

- >j

Z E

3

y

j

/

I

/

y

tit

-

a

Q,

Q,

Q,

z.

g

:

0

3

3

B

*-

:

3

r

3

3

~

2

j

j

j

J

'

/

-

_

<~\

T

1/1

•

-

X.

ir.

ir.

a-.

if.

in

u-.

-

—

—

—

—

—

—

_"

141

■a

V

3
3

+->

c

o
O

<

<
Q

O

H

Q

<

o

Pi
<

Id
►J

m
<

rt

J

o

CO

o

T

o

00

Q
<

o

+1

o
o

Q

Si

a
v
u

U5

CO

0 X!

SO
m

o

in

o

T

in

Q
<

o

+1

m

nO

CN

Q

o <*

2e
^ s

-a
a a

3 o

c

s

V

V

be

be

C

c

a

c

o

o

S-.

-

0

O

m

o

rt

J

m

CN

o

in

o

CN

Q
<

in

NO

+ 1

o
m

o

p

<

a

u

05

cc

s

o
o

Pi

s

V
be
3

"3

o
0

i

o

O
m

CN]

I

o
in

Q

o
in

+1

o
o

(N

Q
<

g S

2 «

Si

W TO

" ^£

o o1^

..,05 HH

^ 3 V
U

£

c
o

Pi£> %

3
V

be

a

"c

o

o

m
in
in

u

o

it
td

3
o

J2

bc~

z<

>
>

-0 be

cn > j3

PnI _ o

'-1 o <**

•J2 0

5 c

bC

^ O 3

"S E

o <a

u v

A*

3 8*

e 5-

in

OOH

3 O ^

..05

3 gS

tf.o'ao

CS 4) CC

Ofn be p

-° 3 3

M l-i .3 B3

>l&Pi %

NO

in
r--

13

142

DA riNG Ol SIT]

The difference between these two sets of dates might reflect earlier,
unknown activities at Broken K Pueblo. Since, however, there was nothii
peculiar about the four samples from which Groningen obtained measure-
ments, except that the) were .ill contained in one single shipment, we
suspect that contamination occurred in shipment or in packing that uni-
formly altered .ill four dates.

It should be emphasized that there was no reason a priori to suspe< i the
foui Groningen dates .ill to be different since theii contexts within the
Pueblo were in no wa) peculiar. The samples were not all drawn from
kivas or storage rooms, or from an) other unique locus. We argue, there-
fore that the four aberrant dates do not represent .1 tenden< \ on the part

the Indians to reuse beams from abandoned, old structures in an) par-
ticular type ol room.

1 wanted however, to pursue the matter .1 bit further l>\ plotting all
the radio-carbon dates Isotopes, Groningen) as ranges of time, take
into account the calculated standard deviation (Fig. 105). I hoped there
might be an overlap of ranges of time and that this might pull all the
dates together and produce a more plausible time sequence.

( >n .1 whole, the results were disappointing, although it did show that
the six Isotope dates, the tree-rim; date, and one date from Groningen
were internall) consistent.

1 noted that the center dale for 1 [ill's ceramic dating span is A.I >. 1 229

11™" I *> w "l

mean earl) A I). -§— + mean late A.I). -g"". This center date is ver)
close to the centei dates for all the Isotope measurements and to one
measurement Room 11) Irom Groningen. It is also very close to the one
tree-ring date part of identical los* from Room 11 that was divided into
three parts and sent to the three laboratories).

I 01 the purposes of experiment A (below) I did not use the four meas-
urements from Groningen that appeared to be too earl) b) several hun-
dreds ol years. It should be noted again that these measurements were
derived from foui samples all of which were sent to Groningen in one

shipment and were run as one series. Experiment A was performed to
what the range of dates would be outside of this block of dates.

A. 1. I found the mean ol the fractionated C 14 dates for all Isotope
1 uns an< 1 one 1 un li om ( rroningen.

2. I aho found the mean ol all the standard deviations referred to in
No. 1.

rhesc two operations yielded a possible time span of from Al>.
IK 1 to, I',,,, ken K Pueblo,

144 PREHISTORY OF EASTERN ARIZONA, III

B. I next averaged all the fractionated C 14 dates and all the standard
deviations and came up with a range of dates from A.D. 1024-1202 for
Broken K Pueblo.

Since Hill's center date fits well into the range of the dates from ex-
periment A, and also well with the tree-ring date, and since the range
attained from experiment B does not include the tree-ring date and does
not accord well with the dates obtained from pottery, we then accept
Hill's chronology— A.D. 1150-1280.

\ I. I he Problem of Sampling

By ] ames N 1 In i

l\"l K< >i)l ( I |( )\

Before discussing the nature ol the sampling techniques employed at
Broken K Pueblo, it will be worth while to examine brief!) some ol the
important aspects of sampling theory which are of concern to archaeolo-
gists. While I make no pretense to presenting a highly sophisticated expo-
sition ol .1 complicated subject (for which I am not competent), it is im-
portant that some consideration of sampling theory be presented as back-
ground material for understanding and evaluating the procedures em-
ployed at Broken K Pueblo. This discussion may also be of some interest,
in itself, since the problem of sampling is one of the archaeologist's fun-
damental concerns, and since there is still a great deal of confusion sur-
rounding the subject. Excellent theoretical discussions of archaeological
field sampling are found in Vescelius (1960), Binford (1964), and Rooten-
berg (1964). I am indebted to these writers for much of the following
discussion.

THE( >RY

It i^ perfectly evident that the best way to obtain accurate and complete

information about an archaeological site is to excavate it completely.

:. when this is done, ol course, it is impossible to recover all of the

i.h t- and materials that may be of interest in the present or in the future.

i most th.it can be asked is that we collect the data necessary to the

solution ol our problems, as well as such other data as oui colleagues

have a right to expect. In ,i sense, then, a site can never l»- completely

in ated.

In mo ( ' i i . even an approximation to complete excavation is not
.. epi perhaps in the case of very small sites | he exigencii
available time, money, and personnel generally require that
some kind be taken However the sample is chosen, and whatever its
-\/r. a usually i .i base from which to make inferences about

site .is .i v.

145

146 PREHISTORY OF EASTERN ARIZONA, III

If one is interested in obtaining a truly representative sample of the
remains of an entire prehistoric community, then it is clear that not all
of the available sampling techniques are equally useful. It would, for
example, be inadvisable to regard the material from a single test pit or
trench as representative of the range of variation in form throughout a
community. This would be analogous to regarding the contents of a
modern kitchen as representative of the artifacts and activities of other
rooms in a house. Clearly, most of the items found in a kitchen will not
be very informative with respect to the activities carried out in a bath-
room, bedroom, living room, or service porch — and we would not want
to make inferences about these rooms (or the house as a whole) without
sampling them. The same kind of argument holds, of course, for sampling
within any universe or population. If the sample is not truly representa-
tive of the universe being studied, then it is dangerous to use it as a basis
for generalization about that universe.

One of the most common types of sampling technique in the South-
west is "block" excavation. This involves opening up one or more large,
contiguous portions of a site, leaving the remainder untouched. This
method of sampling is clearly more desirable than excavating one or two
test pits or trenches; but as it is generally practiced, it suffers from most
of the same drawbacks. Unless these excavated "blocks" are strategically
located in all major portions of the site, the information they yield will
almost certainly be biased. The possibility exists that one could excavate
half of a site, and recover information about half of a sociocultural system !

A vast amount of energy has been expended in attempts to describe
and compare prehistoric "cultures" on the basis of biased samples. It is
frequently the case, for example, that a small sample of sherds is taken
from a site; this sample is then considered to represent the "norm" for
pottery at the site, and this "norm" is compared with a similar biased
"norm" from other sites. If two or more sites appear to contain very
similar pottery, then they are considered closely "related" in some sense;
if the samples are very different, then the sites are more distantly related.
This kind of comparison is dangerous, in part, because it is possible that
the samples drawn from the various sites are not representative of all of
the pottery existing at the sites. In some sites the sample may have been
drawn from a burial area, while in others it may have come from a cook-
ing, butchering, or trash area. One would expect the samples of pottery
(and other artifacts) found in these functionally different areas to be
different from one another, simply because different kinds of activities
(and possibly social units) often employ different kinds of pottery. Re-
garding a sample drawn from one or two of these activity-areas as repre-

SAMPLING 147

senting .1 "norm" simpl) covers up the probability that their is important
functional and spatial variability within a site (Binford, 1964, p, 433).

It i-~ apparent, then, that we will not be able to get at this variability
h\ excavating a small, localized area of .1 site. If we are profitably to
describe and romp. tic prehistoru sociocultural systems, we must obtain
representative samples ol the sites involved. H .1 site must be sampled (as
opposed to complete excavation), then it should be sampled in such a
w.i\ that all portions ol it are represented in some degree

I ii^ is not to sa) that test pits, trenches, "block" excavations, and tin-
like have no worthwhile uses; the) ma) profitably be employed in .1
number ol specific situations. In general, however, the) are nol sufficient
for deriving information about complete prehistoric communities. The)
are most frequentl) used b) archaeologists who want to obtain relativel)
sin. ill samples ol sherds or other artifacts to use in studies ol taxonomy,
chronolog) and the spatial distribution ol traits. Even for these kinds of
culture-trait studies, however, there is some doubt that a biased sample

would I >e \ er\ useful.

Another point with respect to sampling-theor) is also ol interest here.
There are man) instances in which archaeologists have been highl)

telive of the kinds of data they recover and analyze. While some selec-
tivit) is obviousl) necessary, it is easy to create serious biases b) empha-
sizing certain kinds of cultural material to the virtual exclusion of others.
An emphasis on pottery, projectile points and burials, for example, will
restrict the analyst's inferences to certain aspects of those sct^ of data. It

would, in such ,( , ase, he impossible to learn very much about a complete
iocultural system (Binford, 1964, p. 433; Rootenberg, 1964, p. 187).
Given this interest in complete systems (structural-functional analvsis).
it is also important that all areas within a site are sampled in a comparable
in. mner. < Jtherwise, it will be extremely difficult to compare various sub-
areas within a site with respect to their differential frequencies of cultur-
ally relevant materials. If, for example, the excavator screens most of the
soil deposits in half of his site with a one-fourth-inch screen, but does not
een the- other half, it is likely that there will lie a great difference in
imounl and even the nature) of materials recovered in each half.

inparisons ol the differential densities ol artifacts between the halves

would then be misleading li. on the other hand, the sampling proce-
dures are essentially the same in all portions of a site, then one can begin
to interpret differential densities of materials in terms of differences in
function (activity), differences in social units, or temporal changes in

these things. In short, il Samples are obtained in a comparable manner.

then ..ue of the possible soun es of sample variation and en 01 is eliminated
and this, of course, reduces the complexity involved in interpretation

148 PREHISTORY OF EASTERN ARIZONA, III

It appears, then, that a truly representative sample is one that covers
all areas of a site to an equivalent degree. One way of obtaining this kind
of sample would be to use some form of systematic sampling. One could,
for example, excavate all of the odd (or even) numbered rooms or grid-
squares at a site. This would, in general, be superior to the "block" ex-
cavation technique. It has its drawbacks, however, since it is possible that
the units not excavated might consistently contain important classes of
information not found in the excavated ones. While this seems somewhat
improbable with respect to many sites, it could certainly happen if the
site involved had been laid out in a systematic way by the prehistoric
inhabitants (e.g., a grid or checkerboard system).

It is becoming increasingly apparent that the easiest, most efficient
way to obtain a relatively "unbiased" sample is to utilize some form of
probability sampling (random sampling). Although this does not pre-
clude the possibility of "sampling error", it clearly minimizes such error.
J. G. D. Clark (1960, p. 125) views the situation as follows:

... if we are to apply quantitative methods of analysis to settlement debris suc-
cessfully, we must either totally excavate a site, which is only possible for certain
rather primitive states of culture as a rule, or we have to devise some system of
sampling which approximates, as far as is possible, perfect random sampling.

Such a method, by definition, gives each of the units to be samples (e.g.,
rooms, grid squares, etc.) an equal chance of being chosen for excavation
(cf. Parten, 1950; Walker and Lev, 1953, p. 10). It virtually eliminates
the possibility that the excavator will consciously or unconsciously select
areas to excavate which he feels, for one reason or another, are most
"desirable." In other words, it serves to prevent a common form of sam-
pling bias. It also allows us accurately to measure the reliability of a
sample, and it permits us to predict the numbers of various kinds of cul-
turally related materials that we would probably find in the unexcavated
areas of a site. In short, random sampling permits us to make "probability
statements"; and these are often not possible if one uses other forms of
sampling.

It is not possible to go into detail here with respect to the theory and
applications involved in probability sampling. These are amply discussed
in a number of publications (cf. Deming, 1950; Vescelius, 1960; Binford,
1964; Rootenberg, 1964). It will be worthwhile, however, to discuss
briefly the two basic kinds of probability sampling — "simple" and "strati-
fied" —and to attempt to answer some of the criticisms that are frequently
levied against the use of such sampling in archaeology.

A "simple" random sample is one in which all of the units of the sam-
pled universe (i.e., site in this case) have an equal chance of being chosen.

SAMPLING 14')

All ol the units or items In the population, whether the) be rooms,
squares or artifacts, are treated as equivalent Each should be the same
and the) should be independent ol one anothei (for other require-
ments, sec Parten, 19 i0 I a< h sample unil is then assigned .1 number in
.1 systematic manner. Then, after the desired sample-size has been deter-
mined, the units that will actually be selected are determined with the
aid ol a table of random digits, such as that published in Arkin and
ton, 1961.

It 1- not possible, ol course, t<> sample in random mannei .ill ol the
cultural materials in an archaeological site. In ordei to do this, it would
first be necessary to excavate the site and numbei .ill ol the artifacts and
other materials in the process of selecting the sample, This would defeat
the purpose, since the entire site would already have been excavated
ami .1 sample would be unnecessary Rootenberg, 1964, p, 182). None-
theless, it is possible to use .1 simple random sampling technique on an
archaeological site it one chooses natural units (such as rooms 01 depres-
sions) or artificial units (such as grid squares) as the sample units making
up the population. This would yield a random sample of the units chosen,

of all of die culturally relevant material at the site.

\ stratified sample is one in which the universe being studied is not
considered as an undifferentiated whole. Instead, the analyst divides the
univ< into different sub-populations, and then resorts to taking

a simple random sample ol equivalent units within each subdivision sepa-
rately. This is desirable whenever previous knowledge of a site permits
dividing it into two or more different classes of units. Recognizable house
depressions, for example, would not be lumped with the undifferentiated
squares ol a grid system; they are clearly different populations, and should
be treated separately. Such a method of stratified sampling can increase
precision without increasing the total sample needed (Binford, 1

p. a:

An interesting and sometimes useful variant of random sampling has
been suggested by Vescelius I960), Binford (1964), Rootenberg 1
and others. 'I hi- consists in selecting the first sample unit (e.g., grid square,
in a random manner, then selecting all subsequent units in terms
oi a pre-established interval. For example, one mighl select a single grid-
square foi ex< avation by using a table of random dibits, and then exca-
vate every third square, counting from the initial one This sampling
technique has the advantage ol ensuring equal dispersion oi sample units
throughout a site, and it is particularly desirable when one is interested
in determining density clines and comparisons of the densities ,,f various
culturally related items in space 'I Binford 1964, p. 435).

150 PREHISTORY OF EASTERN ARIZONA, III

Regardless of whether this sampling technique or one of the others
mentioned is used, the problem of sample size must be faced; and this
has been given only scant attention in the archaeological literature. It
would seem that, in general, the size of the sample selected will be a func-
tion of the archaeologist's problems, and his estimate of how large the
sample must be to provide enough information for their solution. Ve-
scelius (1960, p. 462) notes that "... a sample of 5 or 10 per cent" of
the units of a population is enough. Rootenberg's survey of the statistical
literature, on the other hand, leads him to believe that there is no single
optimum sample size, and that the size chosen should depend on the
degree of heterogeneity within the population. Although this entails
knowing (or estimating) something about an archaeological site prior to
excavation, it is clearly the most reasonable approach. If we assume, for
example, that a site is extremely homogeneous with respect to its contents,
then a very small sample should be adequate; if the site is suspected of a
high degree of heterogeneity, then a larger sample is required. The method
used in determining sample size at Broken K Pueblo is outlined in the
succeeding section of this paper, and it seems adequate for the purposes
for which it was designed. A method for determining sample reliability
after the sample has been collected has been given by Vescelius (1960).

Before closing this theoretical discussion, there are two frequently heard
criticisms of the applications of probability sampling in archaeology that
should be discussed. One of these is that archaeologists who employ such
techniques are blind slaves to the techniques themselves. In other words,
it is often believed that the employment of random sampling obviates (or
suppresses) archaeological expertise. This criticism is, of course, unjus-
tified as long as such sampling is not considered an end in itself. The
purpose of random sampling is primarily to show the excavator what
kinds of things will likely be found (spatial distribution). Random sampling
is simply a quick and relatively inexpensive way to do this. The initial
sample serves to point out those aspects of the site which may require
additional excavation. This sample should shed more light on the struc-
ture of the site than was apparent prior to excavation. One might then
profitably subdivide the site in accordance with this discovered structural
differentiation, and structure simple random samples of each subdivision
separately (stratified sampling). This process of discovery and re-struc-
turing of the sample ("phase" or "stage" sampling, Binford, 1964, p. 438;
Vescelius, 1960, p. 461) could, theoretically, go on indefinitely — or until
the excavator is satisfied that he has learned enough. It may be that,
upon reaching a certain point, he will want to abandon random sam-
pling altogether.

SAMPLING 151

h is likely that as the site is being excavated in .1 random manna
number of the excavated units will contain only part of an important
feature of some kind. The remainder of such .1 feature is then to be found
in another unit but outside the random sample. Are we to excavate onh
tint p.ut of a feature that happens to fall within the initial sample? ( llearly
inn. In most cases it will be decided that complete features will be 1
vated. One might simply decide, ahead of time, that an) feature en-
countered in a randomly selected unit will be excavated in its entirety
except in those cases in which the information to be derived by don,
is insufficient to warrant the effort.

Probability sampling, then, need ruoei obviate archaeological experti

When it does it is, I t . being misused. It should onh he used in sam-

pling populations ol items (of any nature) that can he assumed to be
relatively homogeneous (as tar as the excavator can determine). In other
words, it is used to sample a universe that is unknown with regard to
content. When content is known, the sample can he structured or stratified
a< 1 ordingly. It is important that there he a continuous interplay between
held work and analysis.

The second frequently heard criticism of this kind of sampling is that
:i he equated with "haphazard"' or "grab"' sampling. This is an un-
informed and wholly unjustified criticism, however, since probability
sampling is not at all equivalent to "haphazard." I he terms "random
sampling" and "probability sampling" are technical statistical terms to
designate a highly systematic sampling technique. The rules of the game
.1! e -tint and measures of reliability arc easily derived. If one of the mi >st
"interesting" structures in an archaeological site should not happen to
fall into the initial sample, this does not mean that it cannot he exca-
vated. If it is unique, it can be sampled as an independent, homogeneous
unit— a sub-structure of the site-. As previously mentioned, the sample is
useful in elucidating what is in the site — it docs not govern the exca\ ator's
judgement of what is important.

SAMPLING .VI BROKEN K PUEBLO

since Broken K Pueblo was a rather large site nearly 100 roonisi. it

not possible to excavate it completely. Furthermore, since we were

'ined with learning as much as possihlc about the activity structure

and social organization of the enure community, it seemed necessary that

our sample I OVC1 the entire site to some degree (even coverage being most

desirable). I hi- clearly precluded the use ol a few test pits, trenches ,„

"block' ations foi most purposes at lea^t. Although consideration

given to the- idea ol excavating every other room, in checkerboard

152 PREHISTORY OF EASTERN ARIZONA, III

fashion, it was decided that a random sample would give adequate cov-
erage. In addition, it would be much less biased than any other sampling
system we might devise, and it would permit the making of probability
statements. It should be mentioned, however, that our application of
probability sampling was not as well done as it might have been; and this
will become apparent to the reader. It was, nonetheless, relatively suc-
cessful; and it may be as instructive to others as it was to us.

The first "bias" to creep into the situation involved the definition of
the universe to be sampled. This, of course, was the site. Instead of sam-
pling the entire site in a random manner, however, it was decided that
this technique should be applied to the rooms only (Fig. 1). Other portions
of the site, such as the plaza and the sheet trash surrounding the room-
blocks, were sampled with trenches. As a result, these areas were not as
systematically or as thoroughly sampled as were the rooms — and proba-
bility statements concerning them would not be easy to formulate. In
short, the population chosen for random sampling consisted of rooms
only (all except the outliers); so the opportunity of stratifying or struc-
turing the entire site in terms of probability sampling was overlooked.

The first step in setting up the sample of rooms was to uncover as many
of the walls at the site as was necessary in order to make a rough ground
plan of the site. This was done in about three weeks time, and it is be-
lieved that no more than two or three surface rooms were missed. Fol-
lowing this, all of the rooms were assigned numbers consecutively, from
1 to 92. This could not include subterranean rooms, of course, since their
existence was not known prior to excavation.

It is worth noting that one of the surface rooms (room 92), which is
located in the west wing, was not discovered in the initial wall trenching
operation. It thus did not receive a number during the systematic num-
bering procedure. It was not possible to renumber the rooms to include
it in sequence, since excavation had already begun in a number of rooms.
Renumbering would have excluded some of the rooms from the sample
that were already being excavated. Their numbers would have been
changed such that they would no longer be a part of the selected sample;
and conversely, other rooms would have had to be substituted. Further,
it would not have been statistically legitimate simply to select an entirely
new sample.

After all of the walls had been located, it should have been obvious
that some of the rooms were very small and others were very large. There
were, in fact, two statistically significant classes of rooms, and possibly
three, based on size alone. This bimodal distribution was not observed,
however, until after the site had been excavated. An opportunity to

S \M1'I.I\<.

stratify the population ol rooms was thus missed. Instead, .1 "simple"
random sample was selected. This i- clearly .1 case in which there was
lack ol feedback, or interplay, between excavation and analysis

Nonetheless, the large size <>t the simple selected proved to be adequate
to compensate foi this error ol omission even though it was ne< essai ) to
avate .t larger number of rooms than would otherwise have been nei -
M \ in achicN e the same results.

The sample size (numbei ol rooms) necessary to give the desired results
w.iv determined with the aid of tables of the > umulative binomial distri-
bution Aiken. 1 I hese tables indicated thai .1 sample of 50 percent

the rooms (46 rooms) would give us .1 90 percent chance probability

ol : discoverin ol an) item which might occui <ml\ five

tinier ,u the site (in five different rooms); and there was .1 11 percent

I finding at least three of these hypothetical items. It was also

determined that then- was .1 probability of .99 of discovering at least one

any items occurring in only ten rooms, and a probability of .61 ol
getting one of an) item occurring onh twice at the site. This seemed
ample for our purposes since, for statistical reasons, we were primarily
interested in obtaining a representative sample ol relatively • ma-

terials at the site. A sample of 50 percent virtually assured us that such
items would be well represented; and even many of the less common

itein^ would be represented as well.

Accordingly, 46 rooms were chosen from a table of random digits

Aikin and Colton, 1961), and the selected rooms were excavated in the

random order indicated l>\ the table. Although this order of excavation

is in no wa\ mandatory, it served as insurance against the possibility

that we would he unable to complete the entire sample in the time avail-

B doing it in this manner, it would have been possil,l<- to quit work

.it any time and still have acquired a random sample even though it

would have been somewhat smaller than the size desired.

'11. /ated sample was not considered as an end in it sell'. It served,

ted, to point up the fact that a number of other rooms needed to
he ited if we weie to gain .1 more or less complete understanding ol

the total range of variability within tin- site. Thus, eight more rooms
were dug, bringing the total number ol excavated surface rooms to 54.
I is, ol course, increased the chances of finding a given number .m\
particular class ol item. In fact, a 54-room sample yielded a probability
ol .95 ol finding at least one ol an) item occurring only five times at the
site. '1 hi- 1- .1 very rough approximation, however, since the additional

ivated looms were not chosen in a random manner.

154 PREHISTORY OF EASTERN ARIZONA, III

Even though additional rooms were excavated, the time and resources
available did not permit the excavation of a rather large block of rooms
on the east end of the south wing, and a somewhat smaller block in the
middle of the east wing (Fig. 1). None of these rooms fell into the initial
sample. In short, the site was not sampled as evenly as had been wished;
and it was necessary that these blocks be excluded from subsequent analy-
sis. It seems likely that this would not have happened if the sample had
been stratified initially, as mentioned above. We now believe that the
"ideal" way to have done it would have been as follows:

1. Structure the population of rooms into four separate "wings." The
four wings of the pueblo form relatively distinct units; at least this
might have been assumed prior to excavation. Thus, each wing
should have been sampled separately.

2. Within each wing, the small rooms (2.5-6.6 sq. m. in floor-area)
should have been separated from the large rooms (ca. 6.6-16.0
sq. m. in floor-area). Thus, the two statistically significant size-
classes of rooms should have been sampled as separate populations.

This would have constituted a stratified sample of the entire universe
(rooms at the site), which would have included eight separate populations.
It would have been almost impossible to obtain uneven coverage in this
way. The plaza and the surrounding sheet-trash could have been struc-
tured as separate random samples, and these might also have been di-
vided internally.

In spite of these drawbacks, the sample was quite productive. One of
the important by-products of it was that after the sample had been exca-
vated, it was possible to predict (approximately) the total number of any
given item in the entire site, even though all of the rooms were not exca-
vated. For example, there were two surface room-type kivas discovered in
the initial 50 percent sample (room 6-kiva and room 29-kiva). Since the
sample was random, we can expect the unexcavated half of the rooms to
contain approximately two more such kivas. This same kind of prediction
can be made for any other category of item, and the accuracy of predic-
tion increases when dealing with items that occur in relatively large
numbers.

It is worthy of note here that if random sampling had not been em-
ployed, we would probably have missed several interesting and significant
portions of the site. We would have missed, for example, at least three of
the six kivas discovered. These kivas were located in areas of the pueblo
that were either very badly preserved or had been "potted" by amateurs.
The employment of traditional methods of sampling would have led us
to select only those areas which were well preserved.

SAMPLING 155

\ previously mentioned, the plaza and surrounding sheet-trash at
Broken K. Pueblo were sampled by trenching. In addition, the plaza was
scraped with a tractor blade in ordei to remove overburden and expo
cultural features. Although the sheet-trash was sampled, no one midden
area was located outside the ni.nn portion oi the pueblo. Two lai
trenches, and a numbei ol exploratory "holes" wen- dug, bui sterile soil
hi bedrock was found in most places at .1 depth ol less than 20 cm. Al-
though there was no time to obtain .t "representative" sample ol th<

eas, 11 is almost certain that no extensive midden .n<-.i exists there today.

There were, ol course, other important sampling problems faced at
.< M k Pueblo, but these wen- ol a somewhat different nature. In addi-
tion to determining the rooms and other areas which were to be exca-
vated, it was also necessary to decide on the specific techniques ol sampling
each of the units that were chosen. It was particularly important that

1 h oi these units he sampled in a < omparablt manner, so 1h.1t ditlerenti.il
densities of culturally relevant materials would lie readily apparent.

All aie. is of the sue were excavated in terms ()| natural levels, since
such levels were easily observable, and since they were directly referable
tu rather discrete episodes of natural and cultural deposition. All trash
ash le\eU of the till, .is well as all materials on the floors, were put
through a screen oi one-fourth inch hardware cloth. The floor levels were
defined as including everything resting directly on the floors or clearly
associated with them. I his definition of "floor" was employed in an
effort to obtain materials which were definitely associated with the rooms,
and to exclude those which had been thrown in the rooms after their
abandonment. Such a procedure is mandatory if significant inferem
concerning room function are to he made. In a few rooms, notably rooms
1 . 40 and 69, there may have been some mixture of flooi and till materials
during excavation; hut on the whole, the procedure yielded much more
reliable information than could have been obtained by our previous
method oi including the Id cm. overlying the floor as "floor."

'1 he 1 non is ranged in depth from aboul 2<i cm. to 110 cm. The "typical"

room was about 60 cm. deep, and usually contained three natural levels,

including the floors, as follows:

I 1 . el 1 lii< km ( 'I1.11 .11 ti-i i-in s

\ Surt Brown humus .nut sand: windblown and watcrwashcd;

411 (in. fallen wall stones; no trash or cultural debris (except an

in 1 .i-iDii. il item ili.it had been washed in).

I; in Red sand or clay; deposited primarily l>\ water; trash

... 1 asionally 1 sometimes rool beams.

I ...1 .<i(in 1 1. 11. 1 packed red sand or grey plastei . 2 cm. thick

iH-tiiiH-s sonifuli.it thicker, and including ■< tempo-
1 a] ac< umulation ol del n

156 PREHISTORY OF EASTERN ARIZONA, III

A few rooms had more than one floor, with later floors having been plas-
tered over earlier ones; and in some rooms there was a trash or sterile
layer intervening between floors.

All floors were troweled and swept, and features were carefully exca-
vated and recorded. Artifacts recovered from features were "bagged"
separately, although this separation was not found useful in the analysis.
After a floor had been photographed, and plans and notes completed,
trenches were dug along the walls in an effort to locate any deeper (earlier)
floors or sub-floor features.

Charcoal samples were taken from all rooms having any charcoal,
and this included most of the "habitation" rooms (the large room category)
at the site. Very little charcoal could be found in storage rooms, because
these rooms rarely contained firepits, and usually had not burned.

An effort was made to collect wood samples suitable for dendrochrono-
logical analysis, but nearly all of the wood at the site was Juniperus and
was found to be undatable, given present techniques. A single date
(A.D. 1260) was, however, obtained in this way.

Pollen samples were taken from the floors of most excavated rooms.
There were 115 samples taken in a total of 43 rooms — an average of nearly
three samples per room. Only 53 were processed and analyzed. Although
most of the analyzed samples were taken from floor surfaces (floor plaster
or hard-packed earth), two of them were from mealing bins, three were
from the bottoms of floor pits, two were from the grinding surfaces of
metates, one was from a sample of fecal material found in rooms 31-33,
and one was from an infant burial underneath room 34. In addition to
this, pollen samples were taken from the natural levels in two stratigraphic
sections, and some additional ones were taken from the present land sur-
face (for control).

Not all of the samples yielded pollen, however. Nine of the sampled
rooms fell in this category (rooms 25, 36, 41, 43, 54, 64, 65, 69, and 82).
Thus, out of a total of 43 sampled rooms, 34 rooms yielded pollen. This
is considered a sufficiently large sample to use in distributional studies
of pollen types within the site (p. 159).

No materials were consciously discarded during excavation. Every-
thing was saved that could conceivably be of use in studying the site.
Even such items as bone, fine chert chips, and seeds were carefully saved;
and all such materials have proved exceedingly useful.

After all the materials had been classified and counted, some categories
were discarded. It was felt that such items as pebbles, fire-cracked rock,
and textured-ware sherds would not require re-examination once they

SAMPLING 157

had been counted I nfortunately, the textured-ware sherds could have
been used in studying vessel form and function; but the) can never be
!<-. i ivered.

the whole, we feel that the site w.is reasonably well sampled. 1 here
rtain gaps, already mentioned, in the evenness ol the covei
but these do not seem to have been as detrimental to the results as we
might have expected I 1 c rooms and areas that were excavated were
done in .1 comparable in. inner, and it has been possible to use the data
obtained in studies of differential density and spatial variability in arti-
facts, non-artifact materials and features. Some ol the results of these
studies are reported in this volume, while others await publication.

I is experiment in probability sampling m.n be useful to others be-
cause it presents some of the rationale for using such a technique, as well
as .1 number of the specific procedures which should be followed in the
attempt. It also suggests a numbei ol procedures which should not be
followed and in some respects we have learned the haul way. It would
xtremely interesting to have the opportunity ol excavating the re-
maining unexcavated portions ol Broken K Pueblo to compare the ma-
terials and conclusions derived with those we already have. This would
seem m be the easiest test ol the validity of our sampling methods. I sus-

t the results would be roughly comparable.

Another interesting experiment would be to choose .1 sample ol an as
yet unexcavated site choosing the sample in some "traditional" manner),
and then actually excavate the site in terms of probability sampling. It
would be very simple to point to many of the features and materials
which would have been missed if probability sampling had nut been em-
ployed. Both the "traditional" sample (unexcavated) and the random
sample excavated) would have to be held equivalent in terms of die
amount ol earth involved (size of the sample). If one had the necessary
resources, he could then excavate the "traditional" sample and compare
results with the first one. It is evident that a number oi archaeologists
will not he willing to accept the application of probability sampling in
archaeology until several such experiments are made. I'or this reason,
thev will he Worth the effort.

The present discussion has \t,-rU concerned with the sampling of .1
single archaeological sin-: and it is geared to sampling individual sites
rather than populations ol artifacts, features, 01 large regions. The prin-
ciples are, in general, the same, however. An excellent discussion ol re-
gional sampling i- given l>\ Lewis Binford (1964

VII. Structure, Function and Change
At Broken K Pueblo

By James N. Hill

A recent trend in archaeological research consists in attempting to re-
construct prehistoric sociocultural systems, much as ethnologists describe
them. This concern seems to involve two primary tasks: (1) the description
of past cultural systems in time and space, and (2) the explanation of sta-
bility and change in these systems. Although there is some concern over
the degree to which these goals can be achieved, it is apparent that a
certain amount of success is now possible.

One of the prerequisites in this pursuit is the view that prehistoric
communities can be studied as "whole" systems, each with an intimately
inter-related set of functional parts. The view that culture can be regarded
as an assemblage or aggregation of individual and comparable traits, on
the other hand, does not lend itself well to structural-functional questions.
This has, of course, been pointed out by a number of archaeologists, and
it needs little support here (cf. Taylor, 1948; Martin and Rinaldo, 19506;
Sears, 1961; Binford, 1962, 1964, 1965).

The description of "complete" systems becomes especially important
when one is interested in explaining cultural stability and change. The
reason for this is that the forces promoting stability and change are
almost certainly operative on complete, ongoing systems rather than on
aggregates of individual traits. To understand the processes of sociocultural
change, it seems likely that it will be necessary to study whole systems
(insofar as possible) and their relationships to the causal forces involved.
Eventually, it may be possible to determine regularities in change; and
these may constitute significant contributions to anthropology.

This brief report serves to illustrate some of the kinds of interpretations
which structure-function oriented researchers can make of archaeological
data. Broken K Pueblo (to which this entire volume is dedicated) is the
subject of concern. It should be made clear, however, that this is simply
a preliminary report of some of the results that have been achieved in
the study of Broken K Pueblo. It is necessarily too brief to admit either
detailed interpretation or the inclusion of the "raw" data necessary to

158

STRUt ll RE, 1 I \( I 1< >\ WD I II VNG1

support the interpretations \ omewhat more detailed account will be
found in Hill. . but the raw data used in this monograph as well

in Hill's publication are published in Archives <>i Archeology,

M.ii mi. I ,onga< re, and 1 [ill I, no. 27, 1 '

I ie basic theoretical model employed in the analysis may Ik- stated
in the form of two postulates: 1 I Since human behavior is patterned
structured, the spatial distributions ol cultural materials are also pat-
terned non-random), and will be so within an archaeolo ite Th<
patterns reflect the loci * > i patterned behavior that existed in prehistoric
times. 2- Cultural items and ^t\li^tir elements change in form and rela-
tive frequency through time, as their associated fun< tions < hange in nature
01 relative frequency. Change occurs primarily through the action
selective pressures. In other words, culture is man's extrasomatic means
ol adaptation to tin- total environment physical and social) cf. White,

\ large portion ol tin- data from the site u,h quantified and manipu-
lated statistically. Three multivariate analyses (factor analyses) were per-
formed on tin I B \1 . 7094 com 1 unci' at the University of Chicago. 1 Th<
analyses permitted the development of non-random clusters of pottery-
types and ceramic design-elements; and the clusters or "factors" were
used in various distribution studies.

Such distribution studies could not be adequately performed without
first controlling the temporal variable within the site. Various lines ol
evidence, both architectural and stratii^raphic, were used in this inti
site dating effort. It w.ts found that the southern portion ol the site is. in
■■ . . earlier than the northern portion, although there is some evidence
it the entire site was occupied simultaneously near the end of the
occupation cf. Hill. 1965, 1966).

It is perhaps significant that lossil pollen data were found useful in

this respect. A detailed pollen chronology lor eastern Arizona indicates a
sliilt in the relative proportions of pollen-types during the time in which
iken K Pueblo was occupied, and this shift was documented within
the site itself.2

For a 1 11 "! l.H tm analysis, see 1 ruchter, 1954.

\ pollen chronology for tin- area (Schoenwettei 1962; ll<\U 1964) indicated .1
J temporal shift in tin- relative proportions ol pollen types, characterized pri-
111.111K by 1 ol arboreal pollen (especially /':■■■ and increas

percent non-arboreal pollen during this tunc- period (ca. A.D. I

Mil ol the rooms .it tin- site which, on the basis ol other evidence, had been consid-
mtained K) percent Pinui pollen; while "late" ro neralh

than 20 percent ["he "late" rooms contained significandy mi
Hon-. 11 hiii .-.il pollen (especiall) Compositae, Chenopodiaceae, Amaranthaceae and
It would not In- valid to claim th.it pollen data can be used widely in
in 1 1 ■ 1 furthci experimentation sc-nns called for.

160 PREHISTORY OF EASTERN ARIZONA, III

The next step in the analysis was the establishment of statistically
valid room-types. It was discovered that there were two discrete modes
of room-size. The small rooms (2.5-6.5 sq. m. in floor-area) generally
contained few features or artifacts. They did, however, contain large
amounts of the pollen of "economic" plants.1 The large rooms (6.6-16.0
sq. m. in floor-area)2 were, on the other hand, associated with firepits,
mealing-bins, ventilators, artifacts (including sherds), lithic waste, animal
bone and seeds; but they contained very little "economic' pollen.3 I
called these rooms "habitation" rooms; while the smaller ones were desig-
nated as "storage" rooms. A third class of rooms was considered "cere-
monial", since they contained features in common with Hopi and Zuni
ceremonial rooms.

This taxonomy of rooms was designed primarily to permit a determi-
nation of variation in the functions of the rooms, insofar as these could be
discovered. Several of the functions of the rooms were determined by
examining the differential spatial clustering (mutual covariation) of both
artifact and non-artifact materials. The functional "meanings" of these
materials were derived from direct ethnographic and world-wide com-
parative evidence. The results may simply be mentioned here:

1. The habitation rooms probably served in the preparation of food?
eating, the storage and use of water, and the manufacture of hunting
tools. They may also have served as centers for the manufacture of
pottery, ground and pecked stone implements, and ornamental items
— but the latter inferences are much less than certain.

2. The storage rooms were used in the storage of plant foods (pollen
evidence) and non-food items; they also apparently served as work
areas.

3. The ceremonial rooms were indeed used ceremonially. They also
served as centers for weaving and the manufacture of hunting tools.
Ethnographic evidence suggests they may also have been male
"clubhouses" or meeting places.

The centrally located plaza shows fairly clear evidence of ceremonial
utilization, in that it contained a small, slab-lined "box" that seems analo-

1 Economic plants are denned as those for which there is evidence of their intro-
duction into the site by men rather than by natural agencies. In the present case, they
include ^ea, Cucurbita, Cleome, Opuntia, and several others.

2 Seven rooms exceeded 16.0 sq. m. in floor-area; one was as large as 33.5 sq. m.
The distribution did not suggest a definite third mode, however.

3 These associations were established on the basis of a series of Chi-square tests of
association. The .05 level of significance (and above) was accepted as indicating
significant association.

STRUCTURE, FUNC1 1< >N AND I II VNGE 161

i- to similar features in present-day pueblos rhese features arc used,
rti ihically, to contain various ritual materials during ceremonial

il.u. ind the) arc presumably symbolic ol the place from which the

peoph ding to le d from the underworld ("sipapu"

"earth navel"; MindelefT, 1891, pp. 71, ~2: Parsons, 1936, pp.

- ch boxes are apparently found in all modern pueblos (Alfon
( hrtiz, native ol San Juan Pueblo, personal communicatioi

In an\ event, nearly .ill of the functions mentioned here .u<- carried
out in analogous types ol rooms found among the recent Hopi and Zuni
Indians.

It is significant that were ii not for the pollen data reco I from
nearly all rooms .it the site, it would not have been possible to I"- sure
that tl ims actually served in a storage capacity. The demon-

stration that pollen data can be used in isolating functionally specific

.i- within .i site may represent an important methodological advam
ment. It may, in the future, be possible to isolate functionally different
sites and seasonally occupied sites) l'\ this method.

e uxorilocal residence units1 apparently existed at Broken K Pueblo,
provided that our interpretations are correct; and these can \n- groupi
into two, more inclusive units Fig. 106 for their locations). These units
were tentatively demonstrated as follows:

1. Non-random distributions of ceramic design element-, pottery-
types, firepit-types, storage pit-. "chopper"-types, and animal bone
indicated discrete localizations within the pueblo (which could not
he explained in terms of functionally specific areas). An example
of i these distributions i- given in Figure 107. A summary ol
all of the distributions i- given in Table 6.

2. Through the use of ethnographic evidence, it was found that th<
item- and stylistic element- were probably associated with female

tivities except perhaps choppers and .mini.il hone, lor which
re i- no clear evidence). No item- clearly associated with m
activities were found to cluster in localized areas of the pueblo.

3. All of the female-associated item- (above) were found to have been

Me in tin- da\-to-da\ maintenance ol .i residence unit.

4. Each unit w.i- found to have had temporal continuity .it least

..n the fact that the entire occupation approximated
130 \e.n-. ,, that it i- likely that both "halves" ol the pueblo w
ipied tor .it least hall ol the occupatii

I a defini in which husband and

wife livi :nt\ ..i tl.<- wife's maternal relati

\

^

D

Fig. 106. The locations of residence units and their sub-units. No rooms were
excavated in the east end of the south wing.

162

^

D

l I Distribution of Factoi Numbei 1 "I the factoi analysis oi (» <i(fi \ types;
from flooi "i rooms, only.

163

164 PREHISTORY OF EASTERN ARIZONA, III

This evidence seems sufficient for the tentative establishment of the exist-
ence of uxorilocal residence units, ' especially in the light of the fact that
such units are characteristic of the recent western pueblos.

It was possible to show that these "residence units" were probably
residence groups (rather than simple aggregates of women). This is sug-
gested by the fact that each unit shared a number of stylistic elements
not common to other units; and the distribution of ceramic design-ele-
ments in kivas suggests the possibility that each unit maintained its own
kiva. This is at least true with respect to the two large units (Units I and
II, Fig. 1).

We may also present the hypothesis that certain non-movable property
was transmitted from generation to generation within each unit (rooms
in this case). This is suggested by the fact that the rooms inhabited by
each residence group seem to have contained the same cluster of stylistic
elements throughout the occupation of the site; and this may indicate
that the same group of people occupied each unit for at least 65 years.
If there had been a high degree of group movement within the pueblo,
the stylistic elements would presumably not have clustered as tightly as
they did. If residence stability is indicated, then, it seems likely that the
rooms in each case were inherited through time. This suggests the hypoth-
esis that the groups were "corporate"2 in some sense.

It is not possible to intimate that matrilineal descent was involved,
since it is not believed possible for archaeologists to excavate descent
rules (or any social rules or norms). Furthermore, social anthropologists
are not even in agreement on the definitions of the various types of de-
scent.

Upon comparing the residence units at Broken K Pueblo with those
discovered at a slightly earlier site in the same valley (Carter Ranch Site;
cf. Martin, Rinaldo, Longacre et al., 19646, and Longacre, 1963, 1964),
I found that both communities had contained residence units of a very
similar size (ca. 20 rooms per unit). Broken K Pueblo, however, con-
tained over twice as many such units as did Carter Ranch site, and they
formed sub-units within two much larger units (i.e., ca. 40-50 rooms per
unit (Figs. 106, 107). Since the earlier site contained two small units only,
it is possible that that site was, as a whole, equivalent to only one of the
largest units at Broken K Pueblo. It may be that as villages increased in

1 The first demonstration that residence patterns can be discovered by studying
the distributions of ceramic design-styles was made by J. D. F. Deetz, in a Ph.D.
dissertation (Deetz, 1960). This study was later published (Deetz, 1965). A subse-
quent study of a somewhat similar nature was made by W. A. Longacre (1963, 1964).

2 A "corporate" group is defined here as a group of related people who co-oper-
ate economically and transmit nonmovablc property within the group.

\

\

X

\

X

X

\

X

X

\

X

X

\

X

X

\

X

X

\

X

\

X

\

X

X

\

\

X

X

STRUCTURE, FUNCTN >\ AND < HANG!

I DISTRIBUTION Ol STYLISTH CATEGORIES USED IN

[S< 'I \ I IM. 1 111 LOCI ( M kl MM \< I UNITS

Stj I 'U Residi n< ■■ I nita

1\ in [( II \ [IB

1 irepits, type I \ X X X X

I .u tOl -'. DOttCr) -t\ pes, fl<

i 1 . pottei \ -is pes, fl<
I .n ti 'i '. .'. i ei .linn design, flooi s
I .n ti i '.i ej .nun design, flooi s

• >iii
Factoi 13, < ei amic design, flooi s
Firepits, I \ i ■ I
Flake ( Ihoppers, floi
I .K t cei amic design, floors

Mine design, fills
I'l.iu I I Bone, flooi s
I ai i' mm design, fills

I ai t« •! 6, i eramic design, floors

I .K tm i. potter) -ts pes, floors
I u tm 4, potter) -t) pes, floors
I .u toi 1 . ceramic design, fills
1 ,u ti a 2, < ei amic design, tills
Firepits, n pc 1 1 1
1 .u toi ;. i 'i amic design, Him, is
1 ,n i .nun design, floors

Mountain Sheep Bone, floors
Bone, tloors
iits, t\ pe \
1 .ii tm 4, i ei amic design, floors
I .H tm \ i ei amic design, fills
1 irepits, rypi II
Mm age Pits, none pi esent

size through time, there was an increasing number of "large" units per
village, each of which included at least two sub-units.

The modern Hopi and Zuni also have a hierarchy of social organiza-
tional units lineages, clans, and so forth, Eggan, 1950) It is possible
thai the largest units at Broken K Pueblo were equivalent to clans oi
phratries), while the sub-units were equivalent to lineages (oi clans)
'1 his cannot be demonstrated, ol course, and perhaps it is unnecessary.
It may, in the future, be possible to compare prehistoric pueblos with
one anothei with respect to "equivalent" social units; and it is likel) that
non-puebloan sites ran be compared in a similar manner

\ numbei ol sociological and demographic changes were occurring
between VI) 1050 and 1 JO0 in the Southwest, and the) ma) have been
promoted b) a minor environmental shift. The existence ol this shift is
well known, and has been demonstrated l>\ palynological, dendrochrono-

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

166 PREHISTORY OF EASTERN ARIZONA, III

logical, and physiographic evidence (Bryan, 1925; Douglass, 1929; Hack,
1942; Schoenwetter, 1962; Hevly, 1964; Schoenwetter and Eddy, 1964;
and others). One of the possible responses to it was a general decrease in
population which was in full force by A.D. 1250. At about the same time,
many villages were abandoned, and people appear to have aggregated
into fewer but larger villages along major drainageways. These sites are
not generally in "defensive" locations, and the fact that they distribute
along waterways suggests the possibility that this change in settlement
pattern was related to a factor in the physical environment. Perhaps
water resources were a critical factor.

There may also have been an increase in the scope of inter-village inte-
gration, as indicated by the fact that "Great Kivas" became more common
between A.D. 1000 and 1200 (cf. Vivian and Reiter, 1960). These kivas
may have been associated with inter-village ritual institutions. Further,
there is evidence of a possible broadening of the scope of intra-village in-
tegration, as suggested by the fact that the ratio of ceremonial rooms to
other types of rooms became continuously smaller through time (Steward,
1937). All of these trends are noted in the vicinity of Broken K Pueblo.

In addition to the hypothesized increasing scope of integration and
aggregation, however, there is direct evidence (too detailed to include
here) that the people of Broken K were responding to a shift in the physical
environment. This evidence suggests that agriculture was being replaced
by the gathering of wild food-crops through time, and hunting was be-
coming extremely unproductive.

After Broken K was abandoned, the supposed processes of aggregation
and integration continued, both in eastern Arizona and throughout major
portions of the Southwest. By 1540, there were only a few large pueblos
remaining. An examination of the ethnographic evidence indicates that
the Hopi and Zuni Indians have a much wider scope of intra-village in-
tegrative mechanisms than is apparent at Broken K Pueblo. It seems at
least a likely hypothesis that an environmental shift made subsistence so
difficult that previously separate family or lineage groups were forced to
aggregate for mutual support.1 This aggregation, in turn, may have led
to the development of broadened integrative mechanisms which served
to bind the larger groups together (cf. Longacre, 1963, 1964). In short,
it seems likely that a number of aspects of western Pueblo social organiza-

1 As an alternative hypothesis, it might be suggested that under the environmental
conditions hypothesized, the inhabitants of the area would disperse rather than aggre-
gate. Still, once these people were "committed" to a stable, sedentary agricultural
economy, it seems likely that they might have attempted to maintain it in the face
of conditions that would probably promote dispersal among less sedentary peoples.
The impression is given, in the ethnographic literature, that the Hopi and Zuni would
find it difficult or impossible to revert to a strictly hunting and gathering existence.

STRIN II RE, M W I I< >\ WD I II ING1 167

tion can be explained in terms oi adaptation to .1 shifting physical en-
\ iron men t.

1 ie interpretations offered here are hypotheses only; and a^ such,
ilic\ ina\ be useful. This papei does, however, present some >>i the ways
in which archaeological data ma) I"- interpreted

VIIL A Brief Appraisal

By Paul S. Martin

This volume is another in our series on the Pre-history of Hay Hollow
Valley, eastern Arizona — from circa 3000 B.C. to A.D. 1400. The authors
of the several chapters have competently covered their subjects. Para-
phrasing their essays is unnecessary.

The chapters by Hill (Sampling and Structure, Function, and Change) and
by Longacre (Artifacts) represent innovations in our archaeological re-
porting and in my orientation. A word, then, in explanation of this shift.

Since my intellectual life spans several decades, I can trace the history
and changes in anthropological stances in my writings. Not unsurprisingly,
through the years my theoretical orientation parallels that of my peers.
Briefly, permit me to review the theoretical orientations that have guided
them and me.

When I was a student, my professor, Fay-Cooper Cole, urged us to
gather facts and to let them speak for themselves. He urged us to employ
speculation sparingly. Emphasis in archaeology was on historical recon-
struction.

In 1924, Kidder published his Introduction to Southwestern Archaeol-
ogy. He summed up (p. 353) the relevant facts of Southwestern prehistory
by means of historical reconstruction.

Strong (1935, p. 6) set forth an expanded notion of archaeology by-
stating: "To the writer [Strong], ethnology and archaeology, far from
being isolated studies, are actually two inseparable means to an essential
end — the attainment of the most complete understanding possible of
human culture at all places and in all times. The more complete data of
ethnology permit certain deductions concerning the past, but these can
be objectively checked only when the archaeological record is known."

In Rediscovering Illinois, (1937) Cole and Deuel stated: "The avowed
aim of archaeology is to make the past live again. It seeks to extend his-
tory beyond the written record . . . archaeology likewise tells us of the
growth of cultures" (p. 1); and ". . . When . . . resemblances [of traits]
were sufficiently numerous or striking, the sites were classed together as
a "culture' . . ." (p. 33).

168

vppr as \i

Nelson 1938, p 146) said: "Archaeology ma) be defined as the science
devoted to the stud) ol the entire bod) i>t tangible relics pertaining i<> the
origin, the antiquity, and the development ol man and his culture . . .
archaeolog) in .1 narrower sense is concerned with the stud) ol th< remains
of human handici aft

In 1949, (l\il<- Kluckhohn, .1 true practitionei ol interdisciplin;
studies, said p 50 "The interest of modern archaeoloi ed upon

helping to establish the principles ol culture growth and chani

1 1. \\ Haury, in .1 paper given in Detroit in 1954, expressed .1 -i< m-
ishment at the fact that lew archaeologists had tried to extract socii
litical factors and interpretations ol settlement patterns in the Southwest.
He went on to say: "Obviously, what is needed is the formulation ol
problems and the invention ol procedures for gathering and studying
comparable data. This will mean a shift in emphasis from historical to
research italics, mine] and a more intensive dependence upon
the work of the ethnologist. ..." 1 1 laur\ . 1 956, p. 3 I.

In a mere matter of 30 years (1924-1954) then, one can trace the shift
in the basic philosophy ol archaeologists from an emphasis on chrono-
logical, particularistic studies cast in an historical mold to an emphasis
on sociopolitical-functional researches.

From the 18 0s to the 1950s, no matter where one looks, one ,
clearl) see these changes: from Bandelier and Boas through Culin, Cole,
Fewkes, Fowke, Holmes. Moore, Putnam, Uhle to Kluckhohn. Haury and
most recentl) to Binford, Deetz, Longacre and Hill; running the scale
from histor) to the stud) of patterns of behavior and their functional sig-
nificant e

But. implicit in the works of all earlier American archaeologists like
the ground hass of a Passacaglia— is the interest in description and analysis
of cultures and in cultural variations, lor the most part, however, a par-
tiality for historical narration and reconstruction outweighs all other
interests in explicit statements. Such orientation stressed, naturally, traits:
homes, pottery, stone ,1\( ;s, projectile points, mortals, and s,, on. I nique
historical events were underscored; historical accidents were invoked to
explain differences and similarities. Nowhere does one find a search foi

laws and regularities, although the historical school claimed to he scien-
tific How did this tendenC) to dwell on culture elements arise' \\'h\ did

archaeologists follow this path?

'1 he answer is that everyone all anthropologists, almost without ex-
ception were wedded to the Same theoietn.d oueniation

What were most anthropologists ol the same time period doing? I

■ 1 a lew titles oi scholarly subjec t^ Means. Musical Instrw

170 PREHISTORY OF EASTERN ARIZONA, III

of the Incas; Culin, American Indian Games; Mason, Aboriginal Basketry;
Sprinzin, The Blowgun in America, Indonesia, and Oceania; Loire, Sun Dance
of the Crow Indians; and Gifford, Culture Element Distributions: Apache-
Pueblo."

The anthropologists of the time considered it their task to find the dis-
tribution and variants of each culture element. Careful analysis often
permitted the scholar to reconstruct the distributions of one particular
trait and trace its spread and these in turn sometimes permitted him to
create a reasonable hypothesis that would explain the place of origin.
For example; the use of copper, of bronze, of iron; the rise and spread of
agriculture, pottery, writing, the couvade, stories of the flood, the origin
and spread of the double-headed eagle; and the connections, if any, be-
tween games, such as patolli, parcheesi, string figures.

In short, everybody was doing it ! Archaeologists were the products of
their times and the empirical approach was popular.

What triumphs are to be credited to our predecessors ! No matter what
"school" they followed, many excellent, sound, necessary, and invaluable
results were obtained and indeed still stand today, unchallenged. One
has only to remember the brilliant work of my former teacher, Edward
Sapir, (1915, 1936) in tracing Athapascan linguistic elements from Canada
to parts of the United States and Mexico; or of my former scholarly chief,
Berthold Laufer, in showing the interchange of culture elements between
Iran and China (Laufer, 1919).

I feel humble when I recognize the achievements of the giants because
without them we could not make advances nor go on to newer interests.
If we hope to arrive at cultural regularities, varied approaches and mul-
tiple hypotheses (Chamberlin, 1965) are essential.

With such an array of methodologies and an accumulation of results
on which to build; with the advent of recently devised statistical tech-
niques for handling data quantitatively, and with a focus on culture as a
system of inter-related variables, we are now in the happy position of
being able to study cultural systems as a whole, to investigate cultural
processes and the reasons for cultural changes, and to contribute to the
science of man. Without the fullness and richness of past researches, we
would be lost.

In his Detroit paper, Dr. Haury stated with great prophetic insight:

" it does appear now, with the trend toward broadened horizons

that inference as to the nonmaterial aspects of archaeological groups must
be as much a part of our reports as is the description of architecture and
pottery. The crux of the matter is, of course, how far we can go in making
such inferences and interpretations." (Haury, 1956 p. 10).

APPRAISAL 171

This volume is an example <>i some of the recently devised techniques
and multiple approaches and ma) be .1 partial answer to "how far we
( .111 go "

Bibliography

Aiken, Howard H. (Director)

1955. Tables of the Cumulative Binomial Probability Distribution. The Annals of
the Computation Laboratory of Harvard University, 35, Harvard University
Press, Cambridge.

Aiken, Herbert and Raymond R. Colton

1961. Tables for Statisticians. College Outline Series, no. 75. Barnes and Noble,
Inc., New York.

Bartlett, Katherine

1933. Pueblo milling stones of the Flagstaff region and their relation to others in
the Southwest. Museum of Northern Arizona, Bull. 3.

Binford, Lewis R.

1962. Archaeology as Anthropology. American Antiquity, 28, pp. 217-225.

1964. A consideration of archaeological research design. American Antiquity, 29,
pp. 425-441.

1965. Archaeological systematics and the studv of cultural process. American An-
tiquity, 31, pp. 203-210.

Binford, Lewis R. and George I. Quimby

1963. Indian sites and chipped stone materials in the northern Lake Michigan
area. Fieldiana: Anthr., 36, no. 12.

Bryan, K.

1925. Date of channel trenching (arroyo cutting) in the arid Southwest. Science,
62, pp. 338-344.

Chamberlin, T. C.

1965. Method of Multiple Working Hypotheses. Science, 148, pp. 754-759.

Cl.ARK, J. G. D.

1960. Discussion. In Application of Quantitative Methods in Archaeology, Robert
F. Heizer and Sherburne F. Cook, eds. Viking Fund Publications in Anthro-
pology, no. 28, pp. 125-128.

Cole, Fay-Cooper and Thorne Deuel

1937. Rediscovering Illinois. LTniversity of Chicago Press, Chicago.

Cosner, Aaron J.

1951. Arrowshaft-straightening with a grooved stone. American Antiquity, 17,
no. 2, pp. 147-148.

Curtis, Edward S.

1922. The Hopi. The North American Indian, vol. 12. F. W. Hodge, ed. Univ.
Press, Cambridge.

Deetz, James D. F.

1960. An archaeological approach to kinship change in eighteenth century Arikara
culture. Unpublished Ph.D. dissertation, Harvard LTniversity, Cambridge.

172

BIB] i< ",k \i-m

1 f si .1 i ri ami* s. Ill m

Anthropology, no ; l niv< I Pre i rbana

I )| MING, Wll 1 1 AM EOW \.

Some 1 Sampling [< >hn W iley and 5 I

Douglass \ I

lecrct ol • Southwest solved In talkative tree rings Nationa l
iphii . 5o. pp

i . Fred

nization ol the western Pueblos I nivei it} ol < I

I I M Kl S, | \\

al expedition to Arizona in 1895. Bureau ol American I
nolog) . 1 'ill Ann. Rep., pi

I I w ii Summers Work in Pueblo Ruins. Bureau ol \merican I thm
Ann Rept., 1900 1901, pp. 1 I

1 1111 K. Bl NJ \M1\

1954. Introduction to factoi analysis I > Van Nostrand ( ompany, Inc., Princeton,

N I

■ ■ I I
1942. rhe changing physical environment ol the Hopi Indian ona Pa

the Peabod) Museum ol American Archaeology and Ethnology, no
i ambridge.

'. . l.MIl W

1940. Excavations in the Forestdale Valley east-central Arizona I niversit)
Arizona Bulletin, 11, no. 4. Social Science Bulletin, no. 12.

195 Spi i illations on prehistoric settlement patterns in th( Southwest. Prehisti
Setdement Patterns in the New World, Gordon R. Willey, ed., Viking Fund
Publications in Anthropology, no. 23. New York.

1 1 \i RY, 1 \\ and I .'■ NDl IN I .. 1 1 VRGR \\ 1

1931. Recently dated pueblo ruins in Arizona. Smithsonian Mis< Coll., 82,
no. 11.

II: \ 1 Y. Rli HARD 1 I( ii Ml S

4. Pollen analysis of quaternary archaeological and lacustrine sediments fi
the Colorado Plateau. Unpublished Ph.D dissertation, rhi in

\i i/nii. i I ii -on.

Hill, James N.

1965. Broken K A Prehistoric Society in Eastern Arizona. I npublished docti
dissertation, Department ol Anthropology, University ol Chii

19C)',. A prehistoru community in eastern Arizona Southwestern fournal ol
thropoli -'.'. no. 1 . pp. 9-30

Kent, Kati Peck

The cultivation and weaving of cotton in th<- preshi l
l nited States Frans. Amer. Philosophical Soc, new series, 47. pi

Kidder, Alfred Vinceni

24. An introduction to stud ithwestem archa with a preliminary

ount n! the excavations -it Pecos Published foi Philli] the

Yale Pr< New Haven A re-print ol ihi>, with an introduction b\ I

k<,i)-.- I '62, ,* ale ' nh ei sir) Pi

Fhe artifacts ol Peco Papers ol the Southwestern 1 xpedition, no <■ Phillips

\> aderm . \ '• I

176 PREHISTORY OF EASTERN ARIZONA, III

Steward, Julian Haynes

1937. Ecological aspects of Southwestern Society. Anthropos, 32, pp. 87-104.

Strong, William Duncan

1935. An Introduction to Nebraska Archaeology. Smithsonian Misc. Coll., 93,
no. 10, Washington. D.C.

Taylor, Walter W.

1948. A study of archaeology. Memoirs of the American Anthropological Associa-
tion, no. 69.

Vescelius, G. S.

1960. Archaeological Sampling: A Problem of Statistical Inference In Essays in the
Science of Culture, in Honor of Leslie A. White, pp. 457-470 (eds. Gertrude E.
Dole and Robert L. Calneilo.) Thomas Y. Crowell Company, New York.

Vivian, Gordon and Paul Reiter

1960. The Great Kivas of Chaco Canyon and their relationships. Monographs of
the School of American Research and the Museum of New Mexico, no. 22,
Santa Fe.

Walker, Helen M. and Joseph Lev

1953. Statistical Inference. Henry Holt and Company, Inc., New York.

Wheat, Joe Ben

1954. Crooked Ridge Village (Arizona W:10:15). University of Arizona Bulletin,
25, no. 3. Social Science Bulletin, no. 24.

White, Anta M., Lewis R. Binford, and Mark L. Papworth

1963. Miscellaneous studies in typology and classification. Museum of Anthro-
pology, Anthropological Papers, no. 19, University of Michigan.

White, Leslie A.

1949. Science of Culture. Farrar, Straus and Company, New York.

1959. The Evolution of Culture. McGraw-Hill Book Company, New York.

Willey, Gordon R. (editor)

1956. Prehistoric settlement patterns in the New World. Viking Fund Publications
in Anthropology, no. 23. Wenner-Gren Foundation for Anthropological Re-
search, New York.

Woodbury, Richard B.

1954. Prehistoric stone implements of northeastern Arizona. Reports of the Awatovi
Expedition, Report no. 6. Papers, Peabody Museum of American Archaeology
and Ethnology, 34.

INDEX

Alexandci . Ed, I 1

Anthropology, significant contributions,

\i« haeologit .il expci tisc, need for,

Bias, m Broken K sample, I
Bimodal <Iimi ibution,
"Ml.. i k" ex< i\ ation, 1 46
Burial, I

( .it ter, James, 1 1
( il emoniaJ i ooms, 1 60
Changes, demographic, 165; sociological,
165
nplcte sot uh uliui.il system, 1 4
put- i . anal) sis by, 1 59
( ore pi epai ation matei ial, 1 2 3
Cultural systems, past, description of, I >8

I >ates, difference in, 1 43
I tomographic changes, 1 65

i l • in. in.iti ilineal, 1<>4

1 n elements, 12<>
1 JifferentiaJ densities, 1 4~

I ation, 1 1

I i onment, \>\i\ sit al, shift in, I

I 1 1 m nt omission, 1 5 3

Explanation ol stability and change, 1 58

1. prepai ation of, 1 60

ll pull. Ml tl.lt. I. l

/ tesl

u tionaJ analysis, 127; infen ni

Great Ki\ .i^. 1 66
i .i i mingen,
Gurlej I 1.11

Hahn, Maxwell, 11

II. lit "t .1 SOt mi IlltUI .il 5) Stem, 1 •

[ntegration, inter-village, 166; intra-

villa
Interpretations ol art haeologit al data

158
Intel -■• illagc integt ation, 1 66
Inti .i-'. illagc integi ation, 1 66
I topes, 1im . I

K ii. hen, mi of, 1 46

K napping activil

1 ..linn .itui \ . .! IVei Rin I
Levt Is, natui al,

\l.iii ilineal dest ent, 1 64

Mi dii inc • ) lindei s, I'M

Model, theoretit al, I

Morphological characteristics "i pottcrv,

National Science Foundation, 11
Natui .il le\ els, 1 55
Natui .il Si iem c I .aboi atoi \
Norm, 1 4(>

i iiliui .il s\ stems, description of,
Patterned distribution, I 24
l'h\ sit .il en\ ii onment, shift in, i
Pollen data, fossil, 1 59; sam| ill
Population, dei i ease in, 1 66
Pot covers, KH

Potter) vessels, restored, 1 33; whole,
Precipitation, 12
Pi mi. n \ shaping debris, 1 2 3
Probability sampling, 148; criticism
150

Random sample, 1 3
Ret iprocal exchange, I
Representati\ e, sample, 1 48
R inaldo's i lassificatii in ol ai tifa p
Rim ii ii l urn uim. 155, 160; types, 160; typ-
ical, 1 55
Rooms, i eremonial, 160; functions of the,
160; large, 160; small, 160

Sat kheim, Judd, 1 1

Sample numbei

Sampling, simple, 148; stratified, 148;

l>i in edui es, 1 4
Set iiini.ii \ shaping debi is, I
Shaping debris, pi imai )

123
Sherd tot il

■^i n iw Hake Blat k-on-v* hite, I
..il units, different es in, 1 4
H ultui al ' hangc, pro
'. ultural system, half, 146; complete,
■

i .il ( han(
Stability and i hange, explanation
Statistit .il tests, signifit an
Mum- knappei 'a i • •< • 1 kn

178

INDEX

Tablets, 101, 104
Tinklers, 104
Trash, amount of, 30
t test, 97
Tucson, Arizona, 139

Universe, to be sampled, 152

Uxorilocal residence, defined, 161; units,
161

Variability, 147

Wenner-Gren Foundation, 1 1
Whole systems, 158

Publication 1027