40
MAYA SETTLEMENT HIERARCHY IN NORTHERN BELIZE
Norman Hammond
The truism that not all Maya sites were of equal size, complexity or past
political importance was first made explicit by Morley (1937-1938, IV, Table 102;
expressed visually in Morley 1946: Plate 19), who ranked sites with inscribed monu-
ments in a hierarchy of four levels from "metropolis" through "cities" and "large
towns" to "small towns. " The existence of lower-order sites lacking monuments but
still more than simply house-mounds was recognized in the Belize Valley survey of
1953-1956 (Willey, Ballard, Glass and Gifford 1965), where a linear hierarchy was
noted, the more important centers being the more widely spaced and less common.
Bullard, who had carried out the survey work in the Belize Valley, applied this notion
in two dimensions in his work in northeastern Peten in 1958, and proposed a spatial
hierarchy of three levels, the house-ruin, the minor ceremonial center and the major
ceremonial center (BuRlard 1960). House ruins were estimated to occur in clusters
of 5-12 within an area of 4-9 hectares, with the clusters spaced perhaps 200 m apart,
and the occasional presence of a larger structure of apparently specialized purpose.
The "zone" around a minor ceremonial center was estimated to contain 50-100 house
ruins over an area of about 1 km2 where topography was not too irregular. The major
ceremonial center controlled a "district" estimated as 100 km2 of upland,_ together
with bajo. Minor centers were distinguished from house-ruin clusters by the possess-
ion of one or more pyramidal structures, presumably temples, arranged with lower
structures around 1-3 plazas; stelae, altars and ball-courts were absent and vaulted
buildings occurred singly, not in compounds. Major centers were distinguished from
minor both in degree, being larger and with more elaborate masonry vaulted buildings,
and in kind, possessing stelae, altzrs, ball-courts, multi-roomed "palaces" and often
more than one group of structures linked by sacbeob.
This three level hierarchy has remained the accepted model, but it is not
precise enough to be used with the more sophisticated analyses of interactions between
sites in an overall network which are now being carried out with the aid of locational
theory and recent advances in epigraphy: if primary, secondary and tertiary ranks
were recognized by the Maya (as is claimed by Marcus, 1974) or superordinate centers
were extending dynastic control over subordinate, as proposed by Molloy and Rathje
(1974), then we shall need to look more closely about the relative size and complexity
of sites as recorded archaeologically in order to appreciate the distinctions.
It was with the aim of examing the relationship of different levels of a site
hierarchy to each other and also to topography, resources and communications that the
British Museum-Cambridge University Corozal Project began an area survey in north-
ern Belize (formerly British Honduras) in 1973 (Figure 1), and this work has led us to
41
propose a site hierarchy of nine levels within the area. Each level is defined in fairly
general terms so that the system may be more widely applicable, and a type site for
each level has been contour-mapped, with a theodolite (Figure 9 is an example) and
then converted to a conventional plan (Figures 2-8, 10).
The lowest levels, 1 and 2, are the single isolated house-platform and the
house-compound or plazuela; the former scarcely needs definition or illustration,
while the latter is basically what Bullard called a "cluster,?" a 2-6 platforms grouped
around a patio which is often also the upper surface of a common basal platform, with
more than 4 platforms in the compound being rare. Both types of site are found in
isolation, forming these lowest levels of the settlement hierarchy, but also constitute
in agglomeration the majority of residential structures in the settlement areas surround-
ing ceremonial centers; in fact lower-order site types often occur within the purlieus of
the larger sites, as noted below.
Level 3 is termed the informal cluster, and is defined as an aggregation of
mounds, essentially undifferentiated in size and apparent function and usually 6-12 in
number, sited around a central open space lacking obvious limits. The type site for
this level of the hierarchy is the Hipolito Group (Figure 2) between San Estevan and
Chowacol, which consists of 9 mounds below 2 m in height, those numbered 3 and 4
being very low; Structure 1 has a small frontal stair and Structure 5 a small (and
probably Late Postclassic) superimposed platform. The space roughly enclosed by
Structures 1, 5, 2, 8 and 9 is too indefinite to be called a patio or plaza.
Level 4, the formal cluster differs from Level 3 in having mounds clearly
differentiated in size and probably function, with the largest likely to have been shrines
or elite residences, arranged round an explicit plaza with definite limits. The number
of mounds is not necessarily larger than in the informal cluster, but their organization
is coherent. The type site is the Martinez Group, 1 km north of the Hipolito Group
(Figure 3), which has 12 structures, several of them with more than one component,
around a coffin-shaped plaza divided by the low Structure 6. The two highest mounds,
Structure 1 and 4, face each other across the broader northern section, and may repre-
sent the lowest level of ceremonial center elaboration (cf. Bullard 1960: 367; Flannery
1972: 38) or grandiose residences for a microlocal elite. This latter function probably
applies in any case to Structures 8 and 11 which face each other across the narrower
southern end of the plaza, although the multiple levels of Structure 8 are an unusual
feature.
Level 5, the minimal ceremonial center, is by this title acknowledged as
having a ceremonial (i.e. religious/political/economic control) function, marked arch-
aeologically by the presence of one structure too large to be residential, together with
features found in Levels  3, or 4. The major structure is usually over 5 meters in
height and pyramidal in form, facing on to an open area or formally defined plaza. The
type site is Santa Rita (Figure 4), which has a single pyramid some 12 meters high on
the north side of a plaza enclosed by low range structures, but no adjacent plazas and
no concentrated settlement (Hammond, ed., 1973: Figure 73).
42
Level 6 is the minor ceremonial center proper -- as Bullard described it,
the small-scale version of a major center -- distinguished from Level 5 by the possess-
ion of 2 or 3 defined plazas, each containing at least one major structure, and with
evidence of differentiation in plaza function, with one forming the focus of religious
activity and one or two others having elite residences and/or administrative buildings.
The type site is Chowacol (Figure 5), 2 km south of the San Estevan major ceremonial
center, which has two raised and separated plazas on a massive common basal platform,
the southern dominated by a 10 meter high pyramid, Structure 1, the northern by a
"palace" substructure 35 meters long (Structure 5) with an impressive frontal stair.
Sites of Level 6 seem to occur commonly in the vicinity of major ceremonial centers of
Level 8 and above to service the outlying population, and at least three such, San Victor,
San Luis and Platform 143 Group (Figure 6) are known within the 20-25 km2 settle-
ment area of Nolmul, the latter two being at a distance of 1. 75 - 2 km from the main
ceremonial precinct, the former 3 km further out in a detached zone beyond a band of
swamps.
Level 7, the small major ceremonial center, possesses at least one pyramid
over 10 meters in height and at least one ballcourt. Stone monuments would probably
appear at this level, or at Level 6 sites within the ambit of major centers, in other parts
of the lowlands, but northern Belize has so far yielded no definite in situ monuments
(a plain stela was suggested by Bullard (1965: 14) at San Estevan; another, more probably
a roof slab, by Gann and Gann (1939: 3) at Nohmul; part of a carved stela reset in Post-
classic times was found at Chan Chen in 1974 (R. Sidrys, personal communication) and
may have been transported from elsewhere). At this level substantial circumambient
settlement begins to concentrate around the ceremonial precinct, whereas at lower
levels it has been fairly diffuse. The type site, Colha, has a plaza barely large enough
to hold the ballcourt and a ceremonial precinct overshadowed by hugh compounds of
administrative or elite residence structures (Figure 7, groups 35 and 40), but a large
and dense settlement area clearly resulting from the site's idiosyncratic status as a
major flint-implement factory (Hammond, ed., 1973: 55-60, Figures 64-70).
Level 8, the medium major ceremonial center, has two or three pyramids
over 10 meters in height, at least one ballcourt, several ceremonial and elite residence
plazas with a differentiation into religious, other ceremonial and residential plaza func-
tions (cf. Hammond 1972). There is a substantial circumambient settlement perhaps
containing minor centers of Levels 5-6 and formal clusters of Level 4 as low-level
ceremonial foci. Sites of Levels 7 and 8 are widely enough spaced to indicate that each
exercises control over a number of minor centers existing also beyond the settlement
area as lower-order service centers (Figure 1; cf. Hodder 1972: 897-903). The type
site is San Estevan (Bullard 1965), with six plazas, three ceremonial and three residen-
tial, three major and three minor pyramids, a ballcourt and several enclosed elite-
residence courts and plazuela groups adjacent to the ceremonial precinct (Figure 8).
The Hipolito Group, Martinez Group and Chowacol sites all lie within the region of con-
trol of San Estevan, which from the spacing of contiguous equivalent-level centers
probably covered an area of 250-325 km2.
43
Level 9, the highest in this regional settlement hierarchy, is that of the
regional ceremonial center; it is differentiated from Level 8 by having larger and more
numerous ceremonial structures, but particularly by their division into two groups
linked by a sacbe, with either or both groups including an elevated acropolis supporting
further structures. The type site is Nohmul (Figures 9-10), where the eastern group
includes the acropolis and ballcourt, and a second Level 9 site seems to exist on the
southern boundary of our research area at El Posito, only cleared and explored for the
first time in 1974, where both groups consist of acropolises but the sacbe is apparently
discontinuous.
Neither of these sites is of more than medium size when compared with those
of the Peten-Campeche heartland -- a smallish Peten center such as Dos Aguadas would
be a Level 8 site -- and the typology outlined above would need another two or three
higher levels to accommodate large sites such as Seibal and Nakum, Yaxha, and the
"supersites" of Tikal, El Mirador and Calakmul. The range of site types is also of
course a continuum within which any selected "threshold" feature will occur over a sub>
stantial vertical distance, quite apart from any horizontal bias resulting from the con-
sideration of sites within a small area, and this typology merely seems to isolate
apparently salient features of the increasing size and internal differentiation in order
to make characterization more accurate and comparison of site status as based on arch-
aeological data more valid. These features reflect the increasing complexity of social
process operating in higher-order centers: the beginnings of collective effort on per-
haps only an extended-family level documented in the common basal platform underlying
many Level 2 sites (see below), the imposition of conscious planning and structural dif-
ferentiation for secular or religious presitge in Level 4, the larger collective effort in-
volved in the erection of public buildings from Level 5 upwards, with the associated
skills of architectural design, quantity surveying, labor organization, quarrying and
transportation of materials, the erection of masonry structures and their embellish-
ment with stucco, sculpture and fresco, and the overall conceptual direction of such a
programme. The tributary population within the realm of a Level 8 center such as San
Estevan may have been around 20,000 persons, that within the settlement area alone of
a Level 9 center such as- Nohmul in excess of 3, 000, so that an adequate pool of labor
and potential craftsmen certainly existed. A single substantial house-platform such as
that of Structure 139 at Nohmul, excavated in 1973-74 and found to be built up of about
10, 000 locally-quarried soft limestone blocks, could be erected by six men in six weeks,
allowing for two cutting stone and four carrying and placing the blocks, a rate of 6
blocks/man/hour and a 10-hour working day, so that fairly ambitious construction pro-
grams could be executed by a local population over a single season.
Such ergonomic estimates are not however the purpose of this paper, although
they show how easily settlement and construction data may be translated, using explicit
criteria, into an approximation of the degree of organization and labor recruitment and
support required. I have merely tried to show that the settlement hierarchy in at least
one corner of the Maya lowlands is more complex than present schemes allow, partic-
ularly at the lower levels, but that this complexity can be defined and divided in such a
way as to provide valid intersite., and perhaps interarea, comparisons.
44
FIGURE 1:
Maya Sites in Northern Belize, classified on Bullard's hierarchy (list below)
San Estevan
Chowacol
Nohmul
Colha
Louisville
Cuello
Avenutra
Santa Rita
San Victor
Progresso I
Kichpanha
Santa Elena
Buena Vista
Caledonia
Hipolito Group
Martinez Group
San Antonio
Benque Viejo
[Santa CruzI
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
El Posito
Cerros
Chan Chen
San Luis
Patchchacan
Progresso II
Honey Camp
[ Laguna de On]
San Lorenzo
Yo Creek
San Lazaro I
Barklog
Chunox
Pueblo Nuevo
Chan Pine Ridge
San Pedro Ambergris
San Juan Ambergris
Bound to Shine
Sartenej'a
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
Saltillo
Sajomal
Consejo
Orange Walk
Carolina
Spainsh Point
Yakalche
August Pine Ridge
San Roman
San Lazaro II
Chiwa Lagoon
Shipstern
Mile 70
Indian Hill
High Bluff
Condemned Point
Rocky Point North
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
:
0
of
0
I-
t-
LLI
Cr)
-LJ
0
aJ
LL
2
N
C)
0
Q:
0
L)
CID
a:~
a:
0,
I,~~~~~~
u)
E
CDG
s~~~
0D
ow
Figure 2: Plan of Hipolito Group, a Level 3 site.
LL
N
LLJ
m
45
0
C1)
LL
LU
o
CD
t-  b
I 8
8
2! 1
CD
_0
N<
(0a
C.)
O       ll~~J
OD
LJ    CD
CY)
0~~
Z     LLJ U
O)    >-
z     LD   z
0     >-
LU    Z)   c
U)       0
T
Figure 3: Plan of Martinez Group, a Level 4 site.
46
B
m
w
0
0
E
uC
0n
0
0
Figure 4: Plan of Santa Rita, a Level 5 site.
47
a
E
I
I-:
I-
z
< i
0E
I
48
0   LL
z~~~~~~~~~~~~~~
D i
o~ ~
oL   LL  3 z
0
Figure 5: Plan of Chowacol, a Level 6 site.
E
5R
S2
tn
cn
CN
49
C,
w
oz
U:
IL
z
0
P
z
0
LU
o
0
Z N
-  g
z
Ch  d
I
zR
Filgure 6: Plan of Platform 143 Group at Nohmul, a Level 6 site in the settlement
area of a Level 9 site.
E
to
tj
5
50
a
0.
03
Figure 7: Plan of Colha, a Level 7 site.
-j
z
0
Wy N
:-J
00m
m
Z- X
Z O
0
In
w
cr
a.
O
z
Cl
li
C6
t
ce
r-
4m
w
z
x
0
w
w
cr
V)
I.-
w
_a.
c]-
0
9
C3
N
4 C
LU
_
)
Z U
<  0
U) -a
/0  s
/4
ii;
la
S
~-E
m   8l
CD  I l
16n
as
a    f
I   -   -
I
I
I
I
I
I
I
I
/
'4
/n
I D 03 %
I
4%
I
I                                                   -I
>I
I
E~  I o-
( 12
I ._
Plan of San Estevan, a Level 8 site.
51
N
N
N4~
It
I
I
a
l
I
I
_I_
C2
Figure 8:
52
U
I a
2  ! :
Figure 9: Contour plan of.the center of Nohmul, a Level 9 site.
k
,l11
.lx
53
I !
R <,
f , e
_
I 1
Ie
Figure 10: Conventional plan of the center of Nohmul, adapted from Figure 9;
Figures 2-8 were similarly adapted.
54
RE FERENCES
Bullard,
W.R.,
1960
Jr.
Maya Settlement Pattern in Northeast Peten, Guatemala. American
Antiquity 25: 355-372.
1965   Stratigraphic Excavations at San Estevan, Northern British Honduras.
Occasional Papers of the Royal Ontario Museum of Art and Archaeology
No. 9. Toronto.
Flannery, K. V.
1972
The Origins of the Village as a Settlement Type in Mesoamerica and
the Near East: A Comparative Study. In: Man, Settlement and
Urbanism, P. J. Ucko, R. Tringham and G. W. Dimbleby, ed.,
Pp. 23-53. Gerald Duckworth, London.
Gann, Thomas and Mary Gann
1939  Archaeological Investigations in the Corozal District of British Hon-
duras. Bureau of American Ethnology Bulletin 123: 1-56. Washing-
ton, D. C.
Hammond, N.
1972
Locational models and the site of Lubaantun: a Classic Maya cbntre.
In: Models in Archaeology, ed. by D. L. Clarke, pp. 757-800.
Methuen, London.
1973   Editor.  British Museum-Cambridge University Corbzal Project,
1973 Interim Report. Centre of Latin American Studies, Cambridge
University.
Hodder, I.
1972   Locational models and the study of Roman-British settlement. In:
Models in Archaeology, ed. by D. L. Clarke, pp. 886-909. Methuen,
London.
Marcus, J.
1974 On the "Square" Model of Maya Territorial Organization (with M.
Romanov and N. Hammond). Science 183: 876-877.
Molloy, J. P. and W. L. Rathje
1974   Sexploitation Among the Late Classic Maya. In: Mesoamerican
Archaeology: New Approaches, ed. by N. Hammond, pp. 431-444.
University of Texas Press, Austin.
55
Morley, S. G.
1937-8 The Inscriptions of Peten. CIW Publication 437. Washington, D. C.
1946 The Ancient Maya. Stanford University Press, Palo Alto.
Wiley, G.R., W.R. Bullard, Jr., J.B. Glass and J. C. Gifford
1965   Prehistoric Maya Settlements in the Belize Valley. Papers of the
Peabody Museum, Harvard University, 54. Cambridge.
Acknowledgment: The plans used here are the result of surveys carried out in 1973
by the British Museum-Cambridge University Corozal Project, funded by these institu-
tions and the British Academy, and working under a permit granted by the Government
of Belize through the Minister for Trade and Industry, the Hon. A. A. Hunter, and the
Archaeological Commissioner, Mr. Joseph 0. Palacio. The mapping was directed by
Richard M. Bryant and executed by him and Basilio Ah, Graeme Noble and Rob
Thallon, with assistance from other project staff.