PalaTial gardens are important green space for ancient states and empires because of the considerable investment in plantings, layouts, embellishments, and upkeep. For Mesoamerica, the existence
Of palatial gardens before the Aztecs is an open question that must be addressed archaeologically because of the scant portrayal of gardens in imagery. Parks, because of their greater size, normally will be located beyond an ancient settlement or on its outskirts, so I focus more on gardens. Note that in modern U. S. usage, park is applied broadly to gardens, parks, forests, shores, and other phenomena with various sizes and functions and a broad range of locations, such as inside cities, on their edges, and far from them. Many former royal gardens in a variety of nations have been converted into public parks or gardens (local terms vary). In the garden and landscape literature, the term garden is applied to more intensively managed green spaces of varying sizes, usually associated wIth residences or public buildings, and usually with well-delimited borders (see Doolittle 2004:398; Hunt 2000:14-29).
To address urban palatial gardens in Mesoamerica, several criteria are relevant. First, I use examples of well-mapped urban settlements in order to identify candidate garden spaces. Certainly, many Aztec royal gardens describeD in ethnohistoric accounts were outside the Aztec Triple Alliance capitals; many gardens and parks were situated in environmentally symbolic and commanding positions or at locations suited to particular purposes, such as hunting, or a lower altitude with a milder climate to accommodate a broader range of species. Nevertheless, some gardens were established as part of elite or royal residences within cities. Elaborate gardens and parks are almost always accompanied by residential accommodations suited to the elevated status of the patron. Therefore, we can begin by examining the availability of space adjacent to palaces, with palaces defined here as elaborate residences of powerful elites or royals, not solely royals (Webster's [1963:605] second definition of "a large stately house," rather than the first definition: "the official residence of a sovereign"). Although at times archaeologists focus on rulers' residences, a broader range of elaborate residences is of interest for my purposes, and, in any case, royals may maintain multiple palaces. Certainly other institutional buIldings, such as temples, may have been accompanied by elaborate gardens - as was the case in ancient Egypt and Rome, for instance - but ethnohistoric data from Mesoamerica mainly address royal gardens and parks attached to palaces.
Identification of adequate open space is a first step in my investigation of urban gardens. Regrettably, comparative literature does not always present scaled drawings of palaces and their grounds to
Guide us concerning how much open space to expect - and garden space surely varied. We might expect that gardens within densely built-up central districts often will be more cramped and might not be readily recognizable from surface indications. For my purposes, assessment will concentrate on evidence for sizable gardens, that is, garden space at least one or two times the area of the palace architecture; often the space will be far greater. Assessment focuses on spaces adjacent to palaces (or reached through bridges across water features) because of the characteristic integration of gardens with palatial architecture. Cross-culturally, palatial gardens were used for a variety of social entertainments, sports, strolls, contemplation, rituals, and social competition, making spatial contiguity with palaces a priority. This proximity does not preclude additional gardens and parks, also with palatial accommodations. Elites and royals often enjoyed multiple residences. Recognition of appropriate space is less straightforward than it might first appear. Palatial grounds often include kiosks, pavilions, shrines, and other structures. From a map of an abandoned city, such architectural remnants could mask the extent or presence of sufficient garden space because such structures may appear similar to surface traces of residences. A further concern is the contemporaneity of different structures. Garden space for noble families may be abandoned and reoccupied later by other city dwellers. Despite these caveats, it is of interest to know if adequate open space existed adjacenT to Mesoamerican palaces.
ANCIENT OPEN SPACE, GARDENS, AND PARKS
A second consideration is waterworks or water-control devices. Waterworks (e. g., fountains, pools, streams, cascades) are one of the most common components of elaborate gardens comparatively. However, waterworks may vary enormously in scope, and only certain kinds are likely to be visible in surface mapping. In arid or semiarid environments, waterworks may have been part of the effort to sustain garden plants. Nevertheless, substantial labor for tending the garden may have been available to supply water as needed in different seasons, or parts of the garden may have been "hardscaped" with paths or other surfacing, reducing the need for water. In some cases, such as Monte Alban in Oaxaca, water supplies are not indicated with obvious canals. How would gardens have been watered? EgypTian art shows servants performing hand watering of gardens (Wilkinson 1998:20-31), pointing to the possibility at Monte Alban of a combination of rainfall, labor investment in hand watering, and use of plants reasonably well adapted to the rainfall regime. Just as
In the case of monumental architecture, gardens can be examples of conspicuous use of labor. Paralleling Trigger's (1990:119) definition of monumental arcHitecture as exceeding in scale and elaborateness what is functionally necessary, ostentatious gardens may present sizable installations and demand constant labor to water and tend plants. Thus, although water is an issue to be considered, arid or semiarid environments may rely on social solutions to the water (and nutrient) demands of plants through command of labor.
Delimitation of garden space is an additional criterion to consider, as walls may have surrounded gardens. Many gardens in ancient states are enclosed because they are associated with "private" residences, not public buildings. Without garden protection, plant delicacies might be subject to depredations by people and animals, such as deer. However, walls can be constructed in various ways; puddled mud or mud bricks may, like houses, undergo decay and erosion, leaving only a slightly elevated residue to represent the wall, or perhaps nothing visible. A more significant archaeological problem is the use of living fences that formed a blockade of trees, cactuses, or thorny plants to effectively screen off the space (Gutierrez 2005). Thus, lack of remnant enclosing walls visible on the surface is not a secure basis to dismiss the possibility of palatial gardens. In the examples considered, walls are not evident - although caution is appropriate because faint features could be overlooked by archaeologists when not expected.
For Mesoamerica, Aztec royal gardens and parks are reported ethnohistorically (anD documented archaeologically at Tetzcotzinco [Medina 1997; Parsons 1971:94-95, 122-125]). Through historical analogy and the abundant comparative cases of elaborate gardens in ancient states, we might suspect their presence earlier than the Aztecs. However, other expectations can be advanced. Perhaps suf-icient space for palatial gardens will be detected in the tropical lowlands where low-density urbanism prevails, but not in compact highland cities where space is less abundant. Or, perhaps the abundance oF open green space in low-density urbanism renders special gardens overly redundant and less likely to be important for elite distinction (yet content may distinguish them) (see Magnoni et al., Chapter 5 In this volume). Thus, perhaps neither highland nor lowland circumstances favor elaborate gardens. Ultimately, the issue of adequate space is an empirical question. Maps allow assessment as long as we bear in mind that urban places are remodeled with some frequency, which may obscure or blur past open spaces.
Figure 11.1 Locations in Mexico of Cerro de las Mesas and Nopiloa, Veracruz, and Monte Alban, Oaxaca (drawn By author).
ANCIENT OPEN SPACE, GARDENS, AND PARKS
As examples, I consider Cerro de las Mesas and Nopiloa, two successive capitals in south-central Veracruz, and Monte Alban, in highland Oaxaca (Figure ii. i). The former two represent low-density urbanism and the latter represents a compact capital on a series of terraceD hills. At Monte Alban, the natural constriction of the upper hills might be thought to prohibit sufficient space for palatial gardens. As I demonstrate for all three cases, however, there is space for gardens adjacent to palaces. For each case, I assess open space (not occupied by architecture) in immediate proximity to palaces that is of sufficient size to have accommodated an elaborate garden (at least equal to the footprint of the palace or larger). The issue of water supply for plants and boundary marking is discussed for each situation. In the designation of possible palatial garden space on figures, I arbitrarily draw rectangles to indicate open space; in some instances, even more space might be marked, but the point of the exercise is simply to indicate the presence of adequate space, not all potential space nor its exact shape. Archaeological excavations are required to assess space, in any case.
This settlement is located within my Veracruz survey, which comprises severaL blocks along the Blanco and Guerengo Rivers in south-central Veracruz. Cerro de las Mesas is an Early Classic (AD
Figure 11.2 Cerro de las Mesas, possible palatial garden spaces. Contour-mapped areas of monumental construction are surrounded by a line; features outside that area were recorded during pedestrian survey (open circles). Monumental palatial platforms are 98, 59, 45, and 823; monumental palatial platforms with temple mounds atop are 712, 35, and 924 (drawn by author).
BARBARA L. STARK
300-600) capital in the Blanco Delta, with an earlier founding during the Late Preclassic period (600-300 BC) (Figure 11.2). It was the largest ancient center in the region, unusual in its agglomerative construction with repetition of monumental plaza groups in the core (Drucker 1943; Stark 1999, 2003; Stark and Ossa 2007; Stirling 1943). Cerro de las Mesas and other Classic-period centers in the region had waterworks - artIficial ponds - integrated into the design of the centers in various ways (Daneels 2002; Stark 1999; Stark and Ossa 2007). The central pond at Cerro de las Mesas lies amid the concentration of temple mounds, but not proximate to palaces. These ponds took advantage of the proximity of the water table, and
Their excavation undoubtedly contributed fill to the predominantly earthen monumental architecture. The ponds appear to be part of the civic layout and only in a few cases incidentally close to palaces. Any palatial gardens could count on hand-dug wells down to the water table for "pot irrigation" of plants, that is, drawing water up by hand from a level likely only 6-8 m below (a practice employed today with gasoline pumps).
ANCIENT OPEN SPACE, GARDENS, AND PARKS
Massive rectangular platforms are the most obvious candidates for palatial residences (Stark 1999). Such functions for similar structures are confirmed by excavations at La Joya in the next drainage west (Daneels 2010). At Cerro de las Mesas, similar platforms sometimes have conical mounds added on top of the platform. Possibly these represent a conversion of a palace platform to support a funerary or commemorative temple when a royal or other prominent individual died. A commemorative temple atop the massive East Platform is reported from La Joya (Daneels, Guerrero, and Liberotti 2013), and funerary temples are known from the Maya lowlands (Coe 1956). Thus, for Cerro de las Mesas, both kinds of massive platforms can be examined, but the ones wIth temple superstructures are less certain to have had a palatial function without excavations. Other elite residences are likely present at the center, involving mounds with attached lower aprons (terraces), but their functions are not yet studied through excavation, and they are not analyzed here.
All five massive palatial platforms at the core of Cerro de las Mesas have adjacent space(s) sufficient for a garden, as do the two comparable platforms with a pyramid on top (Figure 11.2). None of the open spaces marked as possible gardens has obvious delimitation by a wall or other construction, but perishable or living fences may have been employed. Living fences are used in the region today around houses, and wooden posts cut from certain local species take root in the fence line arounD fields and form living fences.
When paramount power waned at Cerro de las Mesas, several centers succeeded it, carving up the previously unified territory during the Late Classic period (AD 600-900). Nopiloa was one of the successors, located to the south along the Guerengo River (Medellin 1987; Stark 1999, 2003) (Figure 11.3). The monumental core nestles in a bend oF the Guerengo River, with some peripheral artificial ponds further
BARBARA L. STARK
Figure 11.3 Nopiloa, possible palatial garden spaces. Contour-mapped areas of monumental construction are surrounded by a line; features outside that area were recorded during pedestrian survey (open circles). Monumental palatial platforms are 6234, 6409, 6404, 6382, and 6309 (drawn by author).
Restricting the central space. The artificial ponds associated with the core are not located adjacent to the central palace platform, and thus, as with Cerro de las Mesas, they seem to have been designed as part of the civic layout, not the private purview of palaces.
At Nopiloa, the earlier tradition of massive rectangular palatial platforms continued. During the Late Classic period here (and at other capitals), some palatial platforms occurred at varying distances from the monumental core. The largest palatial platform in the region is at the south end of the main plaza at Nopiloa (feature 6382 in Figure 11.3), but possibly it is underlain by a natural hill. The river wraps around the eastern end of the monumental core where palatial platform 6382 is located. Today, a low, flooded area northeast of the large platform is partly a function of modern irrigation that drains from northern pastures to the Guerengo River. With less induced flooding, this area may have provided space for a garden, as marked on Figure 11.3. Four other palatial platforms are scattered around the monumental core, within a distance of approximately 1 km. Each has possible garden space adjacent. One platform (6309) has an artificial pond adjacent, the sole instance of a waterwork that
May have been integrated into garden space. As with Cerro de las Mesas, none of the possible garden spaces is delimited with walls or other structures recognizable from the surface.
ANCIENT OPEN SPACE, GARDENS, AND PARKS
Monte Alban
Monte Alban was the capital of the Valley of Oaxaca for hundreds of years (Blanton 1978). Founded around 500 BC, it gained control of the Valley in subsequent centuries and maintained its paramount position until approximately AD 700, with some later occupation and use of the site. Monte Alban was founded on a hill located at the juncture of two branches of the Valley of Oaxaca, along the Atoyac River. Occupation eventually spread to adjacenT hilltops. The settlement was mapped, and surface collections were used to date gridded segments of the site. Blanton defined fifteen districts of the settlement according to a combination of hilltops and the distribution of civic-ceremonial and elite-residential buildings away from the Main Plaza, where the preponderance of monumental and governmental structures is located. Residential occupation is located on terraces built on the hilltops and their upper sides. Blanton defined elaborate residences on the basis of mound groups (i. e., not just a terrace, but also mounD structures) arranged enclosing a shared plaza or patio; outlying civic-ceremonial buildings consist of a single mound in an open area or a double-mound group. The elaborate residences are considered palaces for my analyses.
Based on Blanton's (1978) survey, Figure 11.4 Locates elaborate residential mounds and instances where open space (not a terrace) was located in immediate proximity.1 Of the thirty-five palatial residences, twenty (57 percent) have possible garden open space adjacent. More examples are found farther from the congested Main Plaza vicinity, as might be expected.
One unresolved issue concerns whether some terraces themselves may have supported elite gardens. Six additional elite residences have an adjacent terrace that is unusually long or the residence itself sits on an unusually large terrace that may have allowed garden space (elite groups 1461, 165, 174, 1306, 278, 1453 in Blanton's numbering system). In general, terraces were considered residential by the survey crews, but the possibility that some might, instead, have been devoted to ostentatious gardens should be kept in mind. Were these six to have had garden functions, 74 percent of elite residences
Figure 11.4 Monte Alban, possible elite garden spaces (adapted from figure 1.3 in Blanton 1978:4). Possible elite terrace gardens are unusually large terraces associated with elite residences. Possible open space gardens are not terraces, but open ground.
Had adjacent open space for a garden. Aside from the six cases with adjacent terraces, which have physical delimitations, the other open space appears to have no surrounding wall.
The Valley of Oaxaca is semiarid, with a pattern of seasonal summer rains. Permanent surface water is not available on the hilltops, but the river and irrigation canals for farming provide water near the base of the main hill (Blanton 1978:54-55). Water features were noted only at Atzompa, El Gallo, and at the edge of Monte Alban proper (to the northwest). Only at Atzompa and El Gallo are
Water impoundments part of the possible garden spaces marked on Figure 11.4. Thus, if species intolerant of local conditions were planted, extra labor by servants or slaves or through labor tax would be required to haul water and decant it to plants.
ANCIENT OPEN SPACE, GARDENS, AND PARKS
Settlement growth seems to have played a role in the possibility for adjacent open space for elite residents (Blanton 1978: figures 2.1, 3.1, 3.2, 4.1, 4.2). Palatial residences on outlying knolls are located in parts of the city (Atzompa, El Gallo, and Monte Alban Chico) that generally have little or no pottery from the Early I period (500-200 BC) and little from the Monte Alban II period (200 BC-AD 200). By Monte Alban III-IV (AD 200-950) these areas have moderate or marked amounts of diagnostic pottery, suggesting outward growth of the settlement and a possibility that outlying elaborate residences were constructed after much of the central space was too congested to permit additional extensive gardens.
Discussion
The important conclusion of this exercise concerning open space adjacent to palatial residences is that such space is available for many of the elite residences at Monte Alban as well as those in the two Veracruz capitals. Therefore, the potential for elaborate gardens is not restricted to the two low-density urban capitals. Monte Alban displays a congestion of residential terraces on and around the tops of the set of hills, yet elite residences typically have space adjacent that could have accommodated gardens.
As Monica Smith (2008) notes, archaeologists have not been preoccupied with open space in settlements, an observation applicable to palatial gardens (see Wynne-Jones and Fleisher, Chapter 4 In this volume, who include attention to open space). Unless we flag and investigate possible elaborate garden spaces, we will be immensely limited in assessing investments in open spaces. As indicated by the tendency of elaborate residences to locate in newer, outer areas of Monte Alban where more open space existed, green space in seTtlement peripheries may be more characteristic than in the central core, which can affect our determination of settlement boundaries, as discussed next.
World History
