Study of sociopolitical structure and process in the prehistoric and early historic middle Ulua drainage required the conduct of an extensive regional survey. Knowledge of changing settlement patterns in the area was essential if interrelated demographic and sociopolitical processes, especially political centralization, were to be reconstructed and understood. Conduct of the survey was, however, hampered by several factors. Efforts to overcome these obstacles resulted in various compromises and accommodations which, in turn, affect the results achieved. One of the principal conditions with which we had to deal was the general lack of archaeological information available for the middle Ulua drainage. Gualjoquito (Site 1), a large center ca. 6km north of the municipal capital of Santa Barbara, was one of the few recorded prehistoric loci in the area (Sheptak 1983a). This center's relatively large size, suggestive of its resident's political importance, made it an appropriate node around which survey would be pursued. We, therefore, set ourselves to examining the realm centered on Gualjoquito and the changes this political unit underwent through time. As of 1983, we were not sure of the time period(s) during which the "Gualjoquito polity" flourished nor where its boundaries might have fallen in different temporal intervals. Site 1, initially at least, served as a good starting point for our studies.
Moving out from Gualjoquito we encountered logistical problems that had significant impacts on the course of survey. First, there was the inevitable question of defining boundaries for the research zone (e.g., papers in Fish and Kowalewski 1990). The middle Ulua drainage is not conveniently delimited by topographic features such as mountains or rivers. Instead, central Santa Barbara is characterized by discontinuous, variably sized segments of level terrain, distributed along the Rio Ulua, its major tributaries, and in a few upland valleys, separated from each other by precipitous hillslopes. Each of these habitable areas could be treated as a distinct investigative unit though it is highly unlikely that the occupants of any one vega or valley ever existed in isolation from the residents of comparable neighboring zones. The very small sizes of these level pockets, never covering more than 8km2, made such isolation highly improbable. We quickly came to the conclusion that any borders we drew would be arbitrary divisions of complex communication networks that stretched well beyond the limits of what we could reasonably cover in the time available. Survey, therefore, began with an intensive examination of the Gualjoquito vega and the surrounding foothills and proceeded north, south, and west along the Rios Ulua and Jicatuyo as far as we could reach by the end of the last field season (1986). Investigations extended, therefore, 16km north, 15km south, and 19km west of Gualjoquito. No survey work was conducted east of the Ulua because of the paucity of vegas and highland valleys in this direction (see below).
The next problem concerns the rugged nature of central Santa Barbara terrain. Survey within this area of steep escarpments was physically difficult. We reasoned that most of this dramatically sloping landscape was inimical to ancient and historic settlement. Consequently, research efforts were concentrated within the largest areas of level terrain on both banks of the Rio Ulua and its major tribuary, the Rio Jicatuyo, on the supposition that these pockets held the most productive agricultural land and were the most attractive to past occupation. The valleys cut by the above watercourses also provided major routes linking central Santa Barbara with other portions of Southeastern Mesoamerica. It was hoped, therefore, that by concentrating survey along these channels we might gain some insights into how demographic and sociopolitical patterns were affected by changes in the nature and intensity of communication passing through the area. We also hoped that following this strategy would yield maximum information on settlement distributions for the amount of time and money available to the investigation.
Survey along the Ulua was further encouraged by the planned expansion of the San Pedro Sula-Santa Barbara road within that river valley. Widening or, in many areas, changing the course of this route posed a potential danger to ancient settlements, especially those situated on the vegas. Anticipated population growth and economic intensification following road improvements also constituted significant threats to prehistoric and early historic remains in areas near the highway. We were, therefore, very concerned that sites in greatest danger from construction efforts, those within the Ulua valley, be recorded before they were lost.
Survey coverage, within these constraints, varied in intensity depending on a range of factors including: ease of access to areas to be examined; the amount and extent of modern construction covering a vega or valley; restrictions on studying specific areas, including dense overgrowth and permission of landowners to enter their properties; proximity to Gualjoquito. Overall, the vegas closest to Gualjoquito were subject to the most intense scrutiny, sometimes extending over several field seasons and involving removal of dense vegetation, to understand in some detail the history of occupation and organization of population in areas surrounding the regional center. North of Gualjoquito, eight vegas within 16km of the center were examined though there was less time to strip away obscuring overgrowth and more modern construction (the towns of Gualala, Chinda, and Ilama) with which to contend. Promising vegas and valleys south of Gualjoquito are frequently sites of extensive recent occupation, including the departmental capital of Santa Barbara, and most survey in this area concentrated on the extensive Tencoa vega roughly 11.6km upstream from Gualjoquito. Westward along the Jicatuyo investigations were conducted exclusively on the Yamala vega ca. 19km distant from Gualjoquito. This choice was based on: the vega's large size, containing the most extensive area of flat land within 24km west of Gualjoquito; the general inaccessibility of the six smaller vegas intervening between the Ulua and Yamala; and the mention of Yamala as a center of indigenous population in 16th century Spanish documents. Upland valleys included within this survey, situated southwest of the Ulua/Jicatuyo confluence, were selected because of their appearance in ethnohistoric accounts of native populations at the time of the Spanish conquest. Further study of other areas remote from the principal rivers was curtailed by lack of time and personnel.
These compromises yielded a biased sample of investigated terrain. We are fairly certain that the vast majority of surviving prehistoric and early historic sites were recorded in each investigated area. It is equally obvious that large segments of the middle Rio Ulua drainage went completely unstudied. Some small vegas (especially along the Rio Jicatuyo), many upland valleys, and extensive areas of steep escarpments were ignored during the 1983-1986 field seasons. Though it still seems likely that most of these areas did not support extensive prehistoric and early historic occupation, especially the hillslopes, we cannot rule out the presence of significant loci within these areas. Further, a convincing case cannot be made that the areas investigated somehow comprise a "natural" unit at all points in time within which social, economic, political, and ideological transactions were concentrated and beyond which such ties were attenuated. What we have gathered is detailed information on settlement patterns from a number of areas within the study zone differentially situated with respect to a wide range of variables that could reasonably have affected past settlement; i.e., agricultural potential, access to raw materials for craft production, and situation on potential communication routes leading outside the middle Rio Ulua drainage. Comparison of settlement forms and histories among these investigated zones provides a basis for advancing hypotheses that attempt to reconstruct demographic and sociopolitical structures and processes and account for changes in them over time. These hypotheses do not provide definitive answers but are steps in the process of learning about a previously unknown, topographically complex, portion of Southeastern Mesoamerica.
The following sections desribe and/or define, in order: the methods employed on the Santa Barbara Archaeological Project (SBAP) survey program; the basic chronological sequence into which located sites were placed and something of the basis for those temporal assignments; terms used in site descriptions; environmental and topographic characteristics of the investigated areas; and descriptions of the sites, themselves.
An effort was made to examine each investigated vega and upland valley in its entirety, coverage extending into the bordering foothills. Survey teams were led by an archaeologist assisted by one to four trained volunteers (Ed Schortman and Pat Urban led surveys of the Gualjoquito vega, seven vegas north of that center, and the Inguaya vega to the south; Julie Benyo directed investigation of the Tencoa vega and conducted initial reconnaissance in one upland valley west of the Ulua and three vegas north of Gualjoquito; John Weeks was responsible for study of the Yamala vega, three upland valleys, and the Gualala vega north of Gualjoquito). Extant fields, where available, served as initial units of investigation. Surveyors walked parallel transects through these units, situated close enough so that their fields of vision overlapped. Spacing between team members, however, varied depending on ground cover and slope. In general, the clearer and more level the terrain, the better the ground visibility, the further apart investigators were spaced.
Densely overgrown areas, comprising small minorities of the investigated zones, were usually not walked. The principal exception was the Gualjoquito vega. Here, four to six local laborers, trained to recognize archaeological sites of varying sizes and types, cut paths through the vegetation. These narrow swaths (ca. 1m across) were expanded to expose fully any evidence of past occupation. The number of paths cut, their spacing, and orientations were determined in the field depending, in part, on the density of vegetation and the size and shape of the overgrown area. Every effort was made to investigate thoroughly these vega segments by cutting numerous swaths. Nevertheless, this was not a systematic survey. Paths were not spaced at regular intervals and oriented consistently to specific compass directions. While the benefits of the latter approach are appreciated (Kowalewski and Fish 1990) there was neither sufficient time to train the laborers in such methods nor enough staff to oversee the work. The method described above, though opportunistic, was seen as the best compromise between the need to study unexamined, and difficult to reach, areas and the practical realities of the research. It is very possible that some sites were missed as a result of our approach, it is unlikely that they were either numerous or sizable.
Once a site was located, all survey team members congregated at the locus to aid in its recording. Sites were defined as any locus that produced evidence of past human occupation in the form of one or more of the following diagnostic criteria: structures; features; terraces; rock concentrations; and/or artifact scatters (definitions of these terms are provided in the Glossary section below). Occupation areas separated by 75m or more of seemingly "empty" terrain and/or a natural feature such as a stream channel were defined as distinct sites and given a unique number in the ongoing project sequence and in the developing site registry of the Instituo Hondureno de Antropologia e Historia (sites are also referred to in this report as loci and settlements). In order to avoid duplicating site designations, surveyors working in one of the three zones in which reconnaissance was being conducted simultaneously drew their site numbers from a different portion of the total sequence (settlements immediately surrounding and north of Gualjoquito were designated as Sites 1-299; those from Tencoa, the upland valleys, and the Yamala vega along the Rio Jicatuyo, Sites 300-516; Honduran site registry designations were assigned after the completion of survey and form a single, continuous sequence). Three exceptions are Sites 347-350 located north of Gualjoquito. All materials recovered from a locus, whether in excavation or surface collections, were keyed into this numerical system.
Sites were located on the 1:50000 topographic maps then available for the area. Notes taken on each locus covered the following points: location vis a vis other sites, major topograhic features, and modern constructions; the extent and nature of disturbance suffered by the settlement; modern land-use practices including, where obtainable, informant reports of soil fertility; the presence, proximity, and nature of resources presumed to be of importance to the site's inhabitants (e.g., perennial and seasonal streams, building stones, raw materials of use in craft activities, and potential communication channels); the nature of the site, including information on the number, sizes, and arrangements of structures, the extent and density of artifact scatters, where relevant, and the sorts of rocks used in construction. Loci with one or two extant constructions were mapped using a hand-held compass with distances paced. Larger settlements and/or sites with distinct artifact clusters were recorded using a compass mounted on a tripod and a tape to record distance measurements. Transit maps were made of the largest centers and this instrument was employed, wherever possible, to tie together maps of distinct, smaller settlements. All maps were plotted at a scale of 1:1000 with structures drawn using the conventions of rectified mapping, modified where necessary to convey the adaptation of buildings to local terrain (Carr and Hazard 1961). Surface collections, where possible, were linked to specific portions of investigated sites. Central Santa Barbara settlements, however, usually yield little material without excavation and most collections were so small that they were grouped at the site level.
The major chronological divisions recognized within the middle Ulua drainage are outlined here along with a brief discussion of the procedures by which specific loci were placed within this sequence. The major temporal units defined by the SBAP are:
Late Preclassic: 200 BC-AD 200
Early Classic: AD 200-600
Late Classic: AD 600-950
Early Postclassic: AD 950-1200
Late Postclassic: AD 1200-1500
Historic: AD 1500-1900
Each of these units is defined by stylistic changes noted in various artifact classes as well as architecture. The dates provided for each span were arrived at through direct dating of central Santa Barbara deposits by C-14 and archaeomagnetic procedures and correlating stylistic shifts recognized in the middle Ulua drainage with those known from neighboring portions of Southeastern Mesoamerica. The latter work relies on published sources supplemented, where possible, by direct comparisons of artifact collections. In some cases it is possible to subdivide a period based on subtle changes, e.g., in the percentages of specific artifact taxa defined using stylistic criteria and modifications in building forms and practices. For the most part, such distinctions can only be made in cases where sites have been extensively excavated. The vast majority of settlements recorded during survey have never been dug and can only be placed very approximately within one of the above broad categories.
A total of 61 sites are dated based on excavations conducted from 1983-1986, 24% of all located settlements (this excludes sites where limited shovel tests alone were conducted). Another 113 loci from the Gualjoquito zone can be assigned tentative chronological placements based on surface data, i.e., temporally diagnostic artifacts (primarily ceramics) found in artifact collections and building arrangements. The chronological use of building organization is based on changes from tightly nucleated, patio-focused groups composed of moderately tall platforms in the Late Classic to dispersed arrangements of low, extensive platforms lacking a patio orientation in the Early Postclassic. Some shifts in building procedures also seem to be temporally sensitive, especially the use of unmodified stones set on end to form building foundations and platform facings during the Early Postclassic. Late Postclassic constructions are also not usually organized around patios and tend to lack substantial stone facings; most platforms dating to this interval are made of earth alone. Historic sites generally consist of only a few scattered buildings, defined by stone foundations that barely protrude above current ground surface, many of which contain an elevated area of earth and stone set against one wall (a hearth?). Changes in site planning and construction practices noted above were consistently attested to at excavated central Santa Barbara sites and we feel fairly secure in extending these patterns to unexcavated loci. It is acknowledged, however, that temporal placements based on surface data tend to emphasize the last major period(s) of construction and habitation at a settlement and generally fail to identify the full history of occupation. Earlier buildings and/or deposits may well lie buried within or beneath later edifices and residence may persist after the period of initial construction. Excavated data from central Santa Barbara confirms both possibilities. All we can claim, and that tentatively, is that dates based on surface materials indicate the final, principal period of a site's use, not its full occupation span.
In order to reduce confusion and facilitate description, terms commonly used in site descriptions are defined below.
• Angular rocks: Unmodified stones with relatively little evidence of natural weathering (synonym: chunks).
• Artifact scatter: A variably dense concentration of objects related to human occupation of an area, usually not associated with surviving architecture.
• Bajareque: Wattle and daub construction.
• Cobbles: River-rounded, unmodified stones.
• Construction: Any artificial modification of the landscape or significant remnants of such modification; includes structures (platforms and 0-elevation edifices), features, terraces, faced contours, saddles, projections, and rock concentrations.
• Faced contour: A low stone wall set into a natural rise, presumably to reinforce the ascent. No evidence of side or back walls.
• Feature: A stone line flush with modern ground level that is not part of an extant, larger structure (abbreviated Feat.).
• Foundations: Stone walls set flush with, or rising slightly above, current ground surface that likely served as supports for perishable upper constructions.
• Lot: A collection unit made within a particular subop., numbered sequentially as defined in the course of fieldwork (see, Operation, below).
• Masonry block: Construction stone that has been artificially modified to create its current shape (synonym: cut stone block).
• Monumental construction: Platforms rising at least 1.5m high.
• Operation/suboperation/lot: Major investigative units are designated operations, usually coterminous with an entire site (Gualjoquito, Site 1, is an exception). Operations, abbreviated Op., are a indicated by numbers. These large-scale research units are divided into suboperations (subop.) that comprise particular foci of investigation. Each suboperation is designated by a letter following the Op. number. Surface collections at any site were always included in the "A" subop. for that settlement (again, Gualjoquito is an exception). For example, Subop. 31A would refer to surface collections made at Site 31. Lots are collection units whose horizontal and vertical positions are specified within an op. and subop. These entities are designated by numbers in a running sequence unique to each subop.
• Patio: A space enclosed by constructions on at least two sides.
• Platform: A raised substructure that may or may not have supported construction on its summit.
• Projection: A construction extending out from a larger structure.
• Quebrada: Seasonal stream channel.
• Rock concentration (RC): A dense to moderately dense aggregate of stones that suggests the former presence of a now-destroyed structure.
• Saddle: Elevated construction linking two platforms.
• Structure (Str): Any evidence, elevated above ground surface or not, for the existence of a prehistoric or early historic building that is sufficiently well preserved to allow reconstruction of its overall form.
• Superstructure construction: Any building elements (e.g., benches, foundations, floors, and so forth) raised on top of a platform. Most constructions raised atop these substructures were made of perishable materials, most likely bajareque.
• Terrace: Similar to a faced contour except side walls extend back from the artificially modified ascent. No evidence of a back wall.
• 0-Elevation structure: A building constructed directly on ground surface and defined by stone foundations (sometimes referred to as a, surface-level construction or structure.).