University of Virginia Library

THE PEOPLE OF PUEBLO BONITO

Pueblo Bonito originated in the ninth or early tenth century as a
cluster of rudely constructed masonry houses. This original house
cluster, occupying the slight elevation where a Pueblo I pit village
formerly stood, expanded crescentically to right and left as new homes
were required. Then, perhaps in the second quarter of the eleventh
century, the local population was suddenly increased by arrival of
the second group, emigrating from some as yet undetermined point
of departure, presumably in the north.

These newcomers, culturally more advanced than their hosts and
perhaps numerically superior, lost little time in assuming leadership
of the community. They encompassed the old village in their first
constructional enthusiasm; later they unhesitatingly razed their own
and neighboring houses to make way for successive alterations; they
increased the impregnability of the pueblo and twice enlarged it, the
last time after having abandoned plans that would have doubled its
ground area.

Differences between the two peoples responsible for Pueblo Bonito
are evident in many ways. The one group was old fashioned, unchanging;
the other, alert and progressive. Each had its preferences
in architecture; each had its favored shapes for kitchen utensils.
Since we do not know the real name of either we shall hereinafter,
merely for convenience, designate the first people the "Old Bonitians"
both because they were the original settlers and because they remained

so stubbornly conservative until the end. We shall refer to the second
group as the "Late Bonitians" since they were, in fact, late comers.
It was these latter who molded Pueblo Bonito to its final form, gave
it the mastery in art and architecture that set the tempo for all Chaco
Canyon, and won for it a fame that echoed down the beaches of
Lower California and through the jungles of Veracruz. Together,
these two peoples naturally become the "Bonitians."

Old Bonitian houses were built usually of sandstone slabs as wide
as the wall was thick. The slabs might vary in length and weight,
but they were always reduced to standard width by breaking away
the sides. Since spalling left the edges thinner than the middle,
quantities of mud were required to bed the slabs evenly. And because
that mud was spread and pressed into place by human hands, fingerprints
invariably appear on the surface. Sometimes a mosaic of stone
chips on outside walls protected the mortar from rain and windblown
sand. Interior walls occasionally were made of upright poles bound
together with willows and packed between with mud and chunks of
sandstone, a practice handed down from Basket Maker III and
Pueblo I times.

In contrast, Late Bonitian masonry consists of a core of rubble
and adobe, faced on both sides by neatly laid stonework. Ignoring
for the time being several nondescript but contemporaneous varieties,
we may recognize three successive styles in Late Bonitian wall construction:
(a) that veneered with blocks of friable sandstone of
unequal size and shape but all pecked or rubbed smooth on the exposed
surface only and chinked with pieces of laminate sandstone
about a quarter inch thick; (b) that with fairly uniform, dressed
blocks of friable sandstone, or laminate sandstone, alternating with
bands composed of laminate tablets an inch in thickness, more or less;
and (c) that faced solely with laminate sandstone. Beginning with
that peculiar to the old people, these four kinds of masonry will be
referred to hereinafter as types 1, 2, 3, and 4 (pl. 5). Their relative
ages may be approximated from the fact that tree-ring dates for 65
beams range from A.D. 919 to 1130, as published by Douglass
(1935, p. 51).

The Old Bonitian part of town (fig. 2) was built earlier, and it was
occupied later than the remainder. Five feet of blown sand had
accumulated against the outer wall of the original settlement before
the Late Bonitians arrived and built their homes upon that accumulation.
More sand had gathered against old and new walls before
extensive alterations introduced the third type of stonework. Still
later, construction on a very considerable northeast addition was

interrupted in favor of plans that brought Pueblo Bonito to its
final form.

I believe the Old Bonitians continued in residence longer than the
Late Bonitians because practically all the cultural material recovered
by the Hyde Expeditions, and by the National Geographic Society,
came from Old Bonitian rooms. Some of these rooms are former
dwellings; others clearly had been designed for storage. Eventually
eight of them, both storerooms and one-time habitations, were requisitioned
for burial purposes. Materials stored at the time, religious
or otherwise, were abandoned when the first body was brought in.

Late Bonitian houses, on the other hand, appear to have been
emptied of their contents and leisurely vacated. Windblown sand
sifted in, and subsequently flooring and stonework from upper stories
collapsed upon that sand. No Late Bonitian room had been used for
burials, but a number came to be recognized, sooner or later, as more
convenient places than the village dump for disposal of household
sweepings.

Dumping household waste in convenient corners was not a Late
Bonitian trait exclusively. The Old Bonitians were equally guilty.
Indeed, there is probably not a ground-floor room in the entire village
that wholly escaped the bearer of trash. Some of the lesser quantities
we encountered might have been brushed through open hatchways by
housewives living on the floor above, but larger accumulations represent
repeated contributions by all families in the vicinity. Room 323
was a neighborhood dump both before and after its ceiling collapsed.
Trash from a single source was often thrown in various places. A
bowl with drilled holes evidencing ancient repairs (U.S.N.M. No.
336297) is one of several vessels we restored from fragments
recovered from two or more separate debris heaps.

Where vegetal matter is lacking it is not always possible to distinguish
between intentional and unintentional rubbish deposits. Fireplace
ashes may be quite inconspicuous where blown sand is predominant.
At first it seemed reasonable to recognize as a trash repository
any room in which we found 1,000 or more fragments of broken
pottery. But potsherds alone are not enough. Room 247, for example,
with 2,732 sherds, was not really a dump, but the southwest corner
of Room 245, with only 329 fragments present, obviously was. Here
floor sweepings had been poured through a side door of Room 246B,
in the second story, until it formed a 5-foot-high pile in the corner
of the ground-floor room below and adjoining. An unusual number
of potsherds plus an unusual number of discarded implements such
as bone awls, hammerstones, and manos, is a more reliable measure

Plate 2

Dead pine at the head of Wirito's Rincon, southeast
of Pueblo Bonito. (Photograph by Karl Ruppert,
1922.)

Decayed remains of a great pine that stood within the West Court
of Pueblo Bonito. (Photograph by O. C. Havens, 1924.)

Plate 3

Old Wello and Padilla, Navaho neighbors and frequent visitors
at the National Geographic Society's Pueblo Bonito camp. (Photograph
by O. C. Havens, 1925.)

Hosteen Beyal (Mister Money) who remembered Chaco Canyon
as it was about 1840. (Photograph by Neil M. Judd, 1927.)

Plate 4

Upper: The Chaco in flood. Wetherill's well, destroyed a few weeks later, stands at the left;
below it, wagon tracks on the crossing used until 1928. (Photograph by O. C. Havens, 1921.)

Lower: Layers of silt deposited by gently flowing floodwaters underlie a small ruin, a contemporary
of Pueblo Bonito. Pueblo del Arroyo appears at the right. (Photograph by Charles
Martin, 1920.)

1. Spalled-sandstone slabs of wall width laid in
abundant quantities of mud and often protected
from the elements by closely placed stone chips.

2. Rubble veneered with casual blocks of friable
sandstone dressed on the face only and chinked all
around with chips of laminate sandstone.

3. Rubble veneered with matched blocks, either of
laminate or dressed friable sandstone or both, alternating
with bands of inch-thick tablets of laminated
sandstone.

Plate 5.—The four principal types of masonry at Pueblo Bonito, each represented by a
2-foot square section.

4. Rubble veneered with laminate sandstone of
fairly uniform thickness laid with a minimum of
mud plaster between.

of the deliberate rubbish pile than potsherds alone. The wattled partition
crossing Room 256 may have been built to retain trash piled
deeper behind the wall than in front of it.

Whenever Transitional, Early Hachure, Solid, and Plain-banded
types of pottery were preponderant in a given accumulation we assumed
that most of it came from Old Bonitian dwellings. By the
same token a Late Bonitian source was indicated if Late Hachure,
Chaco-San Juan, Mesa Verde, and Corrugated-coil culinary wares

predominated. With these yardsticks 70 percent of 24,587 sherds
tabulated from Room 323 were Old Bonitian varieties, while, curiously
enough, 51.3 percent of those from Room 325, next on the south, were
Late Bonitian. Our data do not identify Kiva Q as a communal
dump, and yet of the 4,527 fragments collected there 33.4 percent
were Old Bonitian and 37.2 percent were Late. Of 5,558 sherds from
rubbish in Late Bonitian Room 334, 60.2 percent were Late Bonitian
types, but of 642 from a test pit beneath the floor of that same room
52.4 percent were Old Bonitian.

In figure 3 I have attempted to show the distribution of trash
accumulations within the walls of Pueblo Bonito. Our evidence is
conclusive in some instances but not in all. Of the rooms and kivas

excavated by the National Geographic Society I recognize 34 as
certain or probable dumps, while if 1,000 or more potsherds were our
only criterion I should have to add nine more rooms and seven kivas.
Pepper's text identifies only four of his excavated rooms (24, 25, 67,
105) as rubbish repositories but I have marked 10 others as probable
dumps on the basis of Hyde's tabulation (Pepper, 1920, pp. 359-372)
of specimens recovered. His four burial rooms are included because
our four (320, 326, 329, 330) all contained debris of occupation intentionally
carried in to cover the bodies. Of 628 potsherds among
household trash covering the 10 burials in Room 320, 46.1 percent
were Old Bonitian and 25.0 percent Late Bonitian, while of 622 like
fragments among additional debris above the second-floor level the
percentages were 41.6 and 30.8, respectively. Clearly both Old
Bonitian and Late Bonitian families dwelt hereabout and had contributed
proportionally to the earlier as well as the later part of the
room fill.

Intramural trash heaps, each marking an abandoned room, suggest
either a shifting of families within the pueblo or a gradual reduction
of population. A decrease in population could have been brought
about by an epidemic, by failure of the water supply, or by any one
of various lesser causes. The most likely, however, and one for
which we have supporting evidence, is annual reduction in the amount
of arable land. A dwindling food supply spurs discontent; famine
has repeatedly impelled Pueblo migrations within historic times.
Families uprooted and forced from their homes by dissension would
leave most of their possessions behind, as at Oraibi in 1906. Thus,
voluntary departure of their occupants seems a plausible explanation
for the emptiness of Late Bonitian dwellings. Presence of Late
Bonitian utensils in Old Bonitian houses evidences contemporaneity.

Room 28 is a case in point but with complicating factors. Here
Pepper uncovered an astonishing hoard of earthenware vessels and
other objects. According to Hyde's tables (Pepper, 1920, pp. 359372),
the specimens from this one room included 111 cylindrical vases,
39 bowls, 24 pitchers, 2 effigy vessels, 75 stone jar covers, and various
other items. Some of them had been burned or blackened by fire.
Several coiled baskets and 33 earthenware vessels bearing either Old
Bonitian or Late Bonitian designs, lay in the northeast corner at the
level of, and actually on, the sill of the door connecting with Room
51a. About 7 feet to the west, 110 cylindrical vases, 18 pitchers, and
8 bowls had been piled in five layers on "an area of 20 square feet"
(ibid., pp. 119-120). There can be no question that this remarkable
assemblage had been intentionally placed where Pepper found it.

Only Late Bonitian pottery types, in form and decoration, are
discernible in his illustrations.

As he describes excavation of the room, Pepper enumerates many
objects not listed by Hyde; he repeatedly mentions both shell and
turquoise ornaments; 400 shell beads were "associated directly with
the pottery vessels." (Ibid., p. 125.)

Pepper's published field notes are not always easy to follow, but
careful study of them makes it clear that Rooms 28 and 28a originally
formed a single ground-floor Old Bonitian dwelling about 40 feet
long. An introduced partition later divided this long chamber. Room
28 extends from the partition westward beneath Rooms 55 and 57.
These latter are 2-story Late Bonitian structures whose east walls
rest upon logs inserted at ceiling level of the old room below. The
concave north side of that old room was straightened by the Late
Bonitians, but they left the convex south side undisturbed to serve
as foundation for a new wall they built to enclose a second-story
chamber over the east half of Room 28 and all of 28a. This secondstory
chamber, on a level with the first stories of Rooms 55 and 57,
we shall henceforth refer to as 28B.

Pepper's description of Rooms 28 and 28a leaves no doubt that
they were originally constructed of first-type, or Old Bonitian, stonework.
But his figure 44 (1920, p. 116) and his previously unpublished
prints 103, 104, and 120 (herein pls. 6, upper and lower; 7, lower),
together with our own notes, show that the substitute north wall of
Room 28, westward to its junction with the outer southeast corner of
Old Bonitian Room 33, is constructed of laminate sandstone, chinked
with thin little tablets in the manner of our second-type masonry; that
the west and north walls of 28B, including continuation of the latter
to Room 58 (the second story of Room 33), are of second-type stonework
in which dressed blocks of friable sandstone predominate.

Bonitian architecture is often bewildering. I give it thought in this
place only because I believe architecture helps explain the cultural
complexity in Room 28 and adjacent structures. My deductions are
drawn almost entirely from Pepper's notes and photographs, since
our own efforts hereabout were directed toward leveling piles of
previously turned earth and stone the better to control surface drainage.
In the course of this undertaking we laid bare as many secondstory
walls as seemed wise and made a few observations upon them.

If I interpret Pepper correctly, Room 28 was a 1-story Old Bonitian
house that the Late Bonitians altered without wholly dispossessing
its owners. The newcomers corrected the asymmetry of its north
wall just to provide a straighter foundation for the dwelling they

wished to construct above. This latter, 28B, we know to have been
one of a series of early (second-type masonry) Late Bonitian houses
that overlay Old Bonitian rooms west and south as far as Room 327.
The original ceiling beams in Room 28 were replaced by pine logs
whose north ends were socketed 2 or 3 inches in the old walls. Supporting
posts, if not provided at the time, were inserted later. During
these alterations constructional debris was allowed to lie where it
fell; upon this accumulation, with sand carried in for the purpose,
a new surface was created at, or just below, the sill level of the doors
into Rooms 32, 51, and 51a.

The partition separating 28 and 28a, if not built while these changes
were under way, was introduced shortly after. Pepper's figure 44,
illustrating pottery in the northeast corner of Room 28, shows that the
mud with which the partition was coated had been pressed against the
previously plastered north wall. Here the partition was "about a foot
thick and four feet high" (ibid., p. 117). Described from Room 28a,
this same partition was 6 feet high and supported by a 2½-foot foundation
of large stones which, in turn, rested upon the original floor,
8½ feet below the ceiling (ibid., p. 126). In other words, approximately
4½ feet of constructional debris and sand was already present
in the northeast corner of Room 28 when that pottery was left at the
door into Room 51a. With only a 4-foot headway remaining, Room
28 obviously had little use thereafter except storage. On a 4- by
5-foot space in the middle of its floor, 136 vases, bowls, and pitchers
were carefully piled. Miscellaneous stone slabs, tools, and other
possessions were carried in from time to time and left about the room.
Meanwhile sand drifted in with every windstorm until it half covered
the piled pottery, the tools, and utensils.

Some time later 28B and the room or rooms immediately west of
it were partially destroyed by fire. Their floor timbers and supporting
posts were burned, or partly burned; walls were reddened and the
blown sand "vitrified and formed into a slag" (ibid., p. 125).

Reconstruction soon followed, in 1083 or thereafter.[4] The burned
walls were razed; charred timbers, discarded building stones, and
mud mortar were dumped into the storeroom below. By this time,
however, a third variety of stonework was in vogue among the Late
Bonitians; the substitute walls they built on the south side of Rooms
28B, 55, and 57, and between the latter two, were not of second-type,
but of fourth-type masonry—the kind utilized in the latest addition
to the east and southeast quarter of the pueblo. The west walls of

28B and 57 survived the conflagration in large part, for they stand
today as examples of second-type masonry with fourth-type repairs.
The beams and "cedarbark floor covering" mentioned by Pepper
(ibid., p. 216) in Room 55 are relics of this second reconstruction.[5]

Now when and why were 136 bowls, pitchers, and cylindrical vases
piled in the middle of Room 28? An answer to the first question, at
least, may be deduced from Pepper's notes. Describing Rooms 55
and 57, he remarks that their east and west wall "foundations" were
"simply the debris of the burnt-out portions of the rooms" (ibid.,
pp. 216, 219). Below the cedarbark-and-adobe floor of 55, which had
been crushed down "about 4 feet" (thus evidencing an open space
beneath), the "excavations were carried to a depth of over 4 feet
. . . but nothing but clean sand was discovered." Over 4 feet of clean
sand plus 4 feet of open space above thus approximate the 8½-foot
ceiling height reported for Room 28a.

Again, in Room 28, Pepper observed (ibid., p. 117) that "the lower
portion . . . was filled with sand that had drifted and washed in
before the ceiling fell" and (p. 120) that most of the vessels stacked
on the middle floor "were imbedded in the debris that formed the
foundation of the western wall." Now his print No. 103, first published
herein (pl. 6, lower), identifies this "western wall" as that at
the west end of Room 28B, which was built upon a beam bridging the
Old Bonitian room below. The debris of reconstruction, which
Pepper recognized as such but carelessly misnamed, actually flowed
down from beam height into direct contact with the stored pottery.
Because most of the tabular stones visible in that debris slope down
and away from Room 55 it seems clear this waste was poured
through a hole in the floor on the east side of 55. Beyond this dump,
in the west third of Room 28, the quantity of waste was much less,
for it did not prevent the rebuilt floors of Rooms 55 and 57 from
sagging 4 feet when they, in turn, were later broken by collapse of
upper walls. Striated sand against the west wall of 28B proves that
the second-story floor here, as in Rooms 55 and 57, remained in place
for some time after abandonment.

After he had laid bare all the artifacts in Room 28, Pepper photographed
them from various angles. One of the most illuminating
views is that taken from directly above and reproduced as his figure
42 (ibid., p. 114). If the reader will hold this reproduction in reverse,
thus to orient it with the room, he will recognize in the upper right
corner some of the vessels shown (fig. 44, p. 116) at sill level of the

Room 51a door. To the left of these, ranged along the base of the
north wall, are fragments of a stone tablet, two large stones, one
upon the other, a beam-supporting post with charred top (seen more
clearly in print 103), and then a jumble of stone implements apparently
unearthed elsewhere in the room and temporarily stacked here.
Lastly, in the upper left corner of the photograph, one notes a bulging
section of masonry. That bulge marks a sealed door to Room 32.

Pepper first mentions this sealed door while reporting the excavation
of 32, a burial room. The one body interred here lay about 6
inches above doorsill level, or approximately 18 inches above the floor.
A foot and a half of wind- and water-borne sand had collected before
the body was brought in. That the room continued in use while this
sand was accumulating is evidenced by earthenware vessels and other
objects left at various distances above the floor.

Pepper's four burial rooms, 32, 33, 53, and 56, opened one into
the other, yet their only known connection with the outside was the
door from 32 into 28. Presumably each of the bodies interred in the
three inner chambers had been dragged through this same door. The
Room 32 burial, therefore, must have been last of the series, for it
was left 6 inches above sill level; a number of grave offerings, including
two Mesa Verde mugs, a typical San Juan kiva jar, and a bowl
(ibid., p. 124, figs. 47a, c, 48b; p. 132, fig. 49) were pushed in after
and the door was sealed. Since our stratigraphic tests prove that
Mesa Verde pottery reached Pueblo Bonito quite late, it is obvious
the Old Bonitians remained long in occupancy of their section of the
village. They remained there long enough to augment their own
characteristic tableware with vessels produced at various times by
the Late Bonitians: long enough to welcome, during the final years
of Pueblo Bonito, a few immigrant families from the Mesa Verde
country.

However, the door to Room 32 must have been sealed before
Room 28B was rebuilt with fourth-type masonry, since debris of reconstruction
dumped in at that time not only covered the pottery
piled on the floor of 28 but banked up against the north wall. Had
the door then been open, this waste would have flowed through and
into 32.

But even though the western part of Room 28 was isolated by a
pile of debris rising ceiling high, a corridor at the east end remained
open and in use. Witness, in the northeast corner, an assemblage of
33 pieces of pottery and two or three coiled baskets at sill level of the
open door to Room 51a but under "a heap of sand 3 feet high and 3½
feet in width" (ibid., pp. 117-119). Sand so compressed suggests a

wall paralleling the partition, but one sees no trace of it in note or
photograph.

Open doors connected Rooms 28 and 51a, 51a and 39b, 39b and
37, 37 and 4. But the only means of access to the series, other than the
ceiling hatchway in 39b, was my assumed corridor at the east end of
Room 28 and a door cut through the south wall of the latter directly
opposite the entrance to 51a. Pepper does not mention this south
door, but it may be seen behind the shovel handle in the lower left
corner of his print 103 (herein pl. 6, lower) and in the same relative
position on print 104 (herein pl. 6, upper), beneath a rubbish-filled
door. This latter is the westernmost of the two blocked, T-shaped
doors in the south wall of 28B. By a flight of stone steps this south
door gave access to the terrace overlooking Kivas Q and R; its exit,
at floor level of Room 28B, is Pepper's "bin" in nonexistent Room 40.

That the entire east end of 28 remained open for a long while after
the pile of constructional waste was dumped in from Room 55 is
further indicated by the quantity of blown sand that had gradually
accumulated in and below the stepped doorway in the southeast corner.
Pepper's print No. 115 (herein pl. 7, upper) shows the broken west
jamb of that doorway and the imprint of a decayed post directly
above a pitcher and a cylindrical vase. This latter stands on approximately
8 inches of stratified sand, the strata slanting down to the right
toward the little 4-handled bowl, No. 145. Pepper mentions (ibid.,
p. 118) "a cache of stone jar covers" between the bowl and the other
two pieces; from his unpublished print No. 105 I judge the covers
to be on the same level as the store of 136 vessels and the little bowl
to be perhaps a couple of inches higher.[6]

Thus, from Pepper's data I conclude that Room 28, a one-time
Old Bonitian house, became a storeroom when the Late Bonitians
erected one of their dwellings above it. This Late Bonitian dwelling,
Room 28B, was occupied throughout the period of maximum expansion
and architectural change elsewhere in the village. Sometime
during this period, and most likely during the second or third quarter
of the eleventh century, 136 earthenware vessels, apparently all of
Late Bonitian manufacture, were piled on 20 square feet in the middle
of Room 28. Whether they were placed there by the Late Bonitian
occupants of 28B or by Old Bonitian owners of adjacent structures
is a question I cannot answer. Neither can I guess the motive for

the assemblage or why the vessels were stacked in five layers instead
of being arranged in rows against the wall. They were piled on an
indefinite surface formed by sand carried in to conceal constructional
debris covering the original floor. When fire later gutted Rooms 28B,
55, and 57, the pottery stored below, undamaged by the conflagration,
was abandoned where it lay. A sacrificial offering of shell beads was
scattered over the pile, and then debris of reconstruction was dumped
in upon it. The burned dwellings were rebuilt, in 1083 or later, and
life went on as before. Reoccupancy of these second-story rooms
and those adjoining is established by Pepper's finds in them, finds
that include cylindrical vases and other varieties of Late Bonitian
pottery.

Just as windblown sand had found a way into Room 28, so, too,
sand had collected in other ground-floor rooms throughout the pueblo.
In the eastern section, for example, we repeatedly noted 1 to 18 inches
of clean sand on the floors of Late Bonitian houses with fallen
masonry on top. This fact suggests that the rooms had stood empty
for a time prior to collapse of their upper walls. Contrary evidence
comes from Old Bonitian dwellings.

In Old Bonitian houses sand gathered while the rooms were still
inhabited; blown sand was frequently overlain by occupational debris.
Sand had collected in Room 325 to an average depth of 16 inches
before nearby residents began to use the place as a convenient dump
for floor sweepings and kitchen refuse. Room 323, next on the north,
became a dumping place also, and so too did 328. This latter, a
smallish structure built in front of 325, was filled almost ceiling high
with blown sand and household rubbish. Some of these rooms remained
open and accessible for a time, but the sand deepening in
them year after year eventually invited burials when circumstances
barred access to the accustomed places of interment.

It was probably compelling necessity rather than family preference
that first dictated use of storeroom 320 for burial purposes. The
room was free of blown sand at the time, for most of the skeletons
we found there lay directly upon the flagstones. Altogether, 68 individuals
were buried in the four adjoining rooms, 320, 326, 329, and
330. These four are situated at the extreme southwestern end of
the old, original settlement; solid walls separate them from Late
Bonitian houses abutting on the west and south. Of the 24 bodies in
Room 329, two rested upon the floor and the others in an overlying
14 inches of sand mixed with debris of occupation. Old Bonitian
graves were shallow, hastily dug, and hastily refilled.

These several factors—Late Bonitian houses stripped of their

furnishings and vacated; Old Bonitian families crowded together in
their corner of the village; abandoned utensils and ceremonial equipment;
eight Old Bonitian rooms transformed into sepulchers for a
hundred dead that could not be buried outside the walled town—
appear ample reason for believing that the population of Pueblo Bonito
was first halved by migration of Late Bonitian clans and then further
reduced through piecemeal separations prompted by impoverished
farmlands or enemy attacks or both.

That the inhabitants of Pueblo Bonito were plagued by marauding
bands over a long period is proved by the successive measures they
took to strengthen their defenses. The original settlement had no
door in its convex, or cliffward, wall. When the Late Bonitians built
an abutting tier against that wall they provided each room, even those
in the second and third stories, with external doorways. But these
were soon closed, and permanently. In each subsequent major building
program the Late Bonitians deliberately strove to increase the
impregnability of the pueblo. They never again placed a door in the
outer, rear wall; eventually they closed, or partially closed, all ventilators
in that wall and they barred the only gateway to the village.

The lone town gate, that at the southeast corner of the West Court,
was provided when the Late Bonitians were pressing their second
expansion program. Shortly thereafter they built a transverse wall
across the passage but left an ordinary door through the middle.
When this small opening was subsequently blocked with masonry,
Pueblo Bonito was as unassailable as its occupants could make it.
From that time forward there was no gate, no open door, anywhere
in the outside wall of the town. From that day every man going out
to work in his field, every woman seeking water or fuel, went and
returned by ladders that led up to and across the rooftops of 1-story
houses enclosing the two courts on the south.