Creating a Typology of Cave Function and Ideology in the American Southwest (2013)


In a typology, material remains are classified for use as the basis of inferences whose evidence are originally unrepresented in the materials (Gardin 1980). The purpose of this project is to create a functional typology, which can be used to analyze data for specified research questions. Focusing in various caves across regions in the American Southwest, this typology is specifically aimed for interpretations of cave usage and ideologies. The benefit to a project such as this is to provide a theoretical basis for cave archaeology projects, through the creation of a searchable database. This database has the potential of being used in conjunction with Geospatial Information Systems (GIS) for predictive modeling or display of analysis.

The questions I aim to answer in creating this typology relate to cave functions and ideology in the southwest. These questions span the theoretical realm, and include orientations toward cognitive, environmental, and ethnographic archaeologies. Is there a difference in use based on location? Do caves maintain a single function across occupations, or does it change through time? What role does the environment play in cave selection?

The American Southwest Region

The American Southwest stretches over many diverse environments, including mesas, canyons, gorges, mountain ranges, basins, and volcanic deposits (Fagan 2005), however The entire southwest is within the Dry Domain which is a land characterized by deficits of water (Cordell 1997). The climate on average is arid to semi-arid, but Fagan (2005) notes this generalization is skewed, as the Southwestern climate is “highly localized” (Fagan 2005).

The flora types of the Southwest are adapted to extensive sun and sporadic rainfall. This includes various Grasses, juniper, pinion pine, agave, evening primrose, wild onion, potato tubers, Indian rice grass, and sunflowers. The same adaptive consideration applies to Fauna. The intense southwestern climate provides sparse nutrient plant food, but game such as mule deer, pronghorn antelope, and bighorn sheep fare well because they find preference in plants such as juniper and pinion pine (Cordell 1997).

Cave and Rockshelter Formation

Rockshelters differ from caves in that they are typically more wide than deep, with an extensive overhang. Subaerial exposure makes them more prone to weathering than caves. Rockshelters are formed by weathering, and often by streams which undercut cliffs. Caves are generally formed by the dissolution of carbonate rock surfaces, as with limestone caves; or weathering of friable sandstone. Cave networks are larger, and the interior atmosphere is generally more humid than in rockshelters (Goldberg, Paul, and  Macphail 2006).

Sample Cave Sites

Ventana Cave

Ventana cave is a weathered sandstone cave located in southern Arizona. This cave is heavy in faunal data including bison, horse, sloth, rabbit, and deer. Shell and feathers were also found. Haury found artifacts that indicate hunting such as projectile points and scrapers, as well as flintknapping hammerstones and flakes. Perishable materials recovered in Ventana cave include sandals and netting (Haury 1975).

Cowboy Cave(s)

Cowboy Cave(s) are two adjacent caves[1]located in Wayne County, Utah along the Barrier Creek drainage (Jennings and Holmer 1980). Artifacts were recovered including sandals, basketry, and projectile points. A cache of maize was also found. This cave also has split twig figurines. The rock art is portable incised pebbles (Jennings and Holmer 1980).

Cliff Palace

Cliff Palace is an extensive architectural village built in Mesa Verde, Colorado. It boasts 220 rooms, many of them multi-level, and 23 kivas (Cordell 1997). The extensive architecture was built using sandstone blocks, held up with wood beams, and chinking for added resilience (Fewkes 1911). Artifacts include groundstone, ceramics, textiles and sandals.

Balcony House

Balcony House is another architectural cliff dwelling of Mesa Verde that was built into a sandstone alcove (Fiero 1998). There was no report published for the first 1910 excavations by Jesse L. Nusbaum. And much of the information has been compiled from field notes read 80 years after. The Balcony House site consists of two kivas and 8 rooms, 9 meter entrance ladder. Retaining wall built along the alcove (Fiero 1998). Recovered artifacts include Ceramic vessels, baskets, stone and bone scrapers, turkey bone awls, multiple axe heads, and maize cobs (Fiero 1998).

Bat Cave

            Small Cobs, very similar to wild maize, but its speculated to have been cultivated (Mangelsdorf 1954).

Feather Cave

            Feather cave, located in Lincoln County, New Mexico was ignored at first due to little being recovered save for numerous reeds, 114 sandals, and a few ceramic sherds (Ellis and Hammack 1968). Later expeditions found rock art delineating a passageway into another cave (Ellis and Hammack 1968).

Arrow Grotto

            Arrow Grotto is located just to the north of Feather Cave, and has been referred to as the “inner sanctum of feather cave,” (Ellis and Hammack 1968). It is accessible only through a narrow passageway near the floor on the north side of Feather Cave. Its name derives from the abundance of miniature bow and arrow artifacts recovered within (Ellis and Hammack 1968). Other data include rock art, numerous reeds, feathers lodged into a crevice, and one bead-thought to be from a bow string (Ellis and Hammack 1968).

Fresnal Shelter

            Fresnal Shelter is located in New Mexico, in the Sacramento Mountains (Tagg 1996). Over 30 small cylindrical pits were found, along with one human burial and many small stone circle hearths (Tagg 1996). Preservation was ideal, and approximately 30 fragments of basketry and 400 sandals were recovered along with animal hides and shell (Tagg 1996). Cultigen samples revealed the presence of maize and beans. In fact, it appears the only artifact type not seen in Fresnal Shelter is pottery (Tagg 1996).

White Dog Cave

            White Dog Cave was named for one of its recovered dog burial.  An adult male was buried with the dog, who appears to have had long white hair with brown spots, and another dog burial had a shorter coat (Cordell 1997).   


Creating the Typology

The first step in formulating a typology for cave use was to compile a descriptive spreadsheet derived from preexisting data obtained from various sites across the American Southwest. Specific categories included on the spreadsheet for data collection include: cave material and formation process, cardinal facing of the cave opening, associated site features, whether there is rock art present or not, artifact types, floral data recovered, and faunal data recovered. I also included a function category, which reflects interpretations fromprevious research, as well as my own interpretations.

Cave selection was intended to be evenly dispersed across all regions of the southwest, with at least one sample site from each of Kidder’s defined cultural regions (Figure 1)[2]. I soon realized these were tight parameters, as cave sites do not follow an even site to region distribution. I was able to select a sample of 9 caves from various locations within the southwest region.

Fig. 1   Kidder’s Regional distinctions (Kidder 1962)

With a basic spreadsheet in place[3], I began literary research to compile data from various caves into the spreadsheet. After each cave had been described, I designated one or more functions. These possible functional designations are: Areas designated for storage, dwelling, camping, and ceremonial use.

In order to increase the efficiency of my queries, and enhance the display of the final spreadsheet, I established a coding system. These codes apply to the artifacts, floral, faunal, features, architecture, morphology, and function designation categories. Table 1 provides a reference of all codes utilized in the final spreadsheet for this project.

Table 1. Codes for Cave Database

Using the Typology

Because my sample size was small (9), I was able to calculate statistics and ratios by hand to answer sample research questions. These questions include: How likely is it researchers will encounter textiles or perishables; How many caves have a ceremonial function designation and rock art; How likely that you will encounter a burial?


            The results of my typology indicate that textiles and perishables can be found in 78% of caves in the American southwest. Caves which have been interpreted as ceremonial, and have rock art present are 33%, and the likelihood of encountering any sort of burial is 45%. I also looked at the frequency of sandals occurring in cave sites, as they represent a large sample of the perishable data. Sandals occur at a whopping 78%.

Table 2. Frequency of Floral Data in Southwestern Cave sites


The sample used for this project was miniscule, and therefore the statistics derived may not necessarily reflect the regional cave population. In my literary research, the primary issue was looking for specific categories of information, when many caves yield unique individual data. An issue for query is in the organization of the database. Issues such as multiple occupation levels, and numerical counts for artifact types are not reflected. This data is important for interpretation. What I was able to show was a presence or absence at a particular site. There are too many factors for this to be a really functional typology.


In conclusion, I found that this specific typology project is not ideal, as there are too many factors for organization to be efficient. A similar project could be useful with different organization and different data criteria.

Possibilities of future research, after a working database has been created, are to import into a Geospatial Information System. This would make it possible to conduct analysis on data which may not be available by literature research. For example, after plotting the coordinates on a base map, A hydrology layer could be added to assess the proximity of a site to a sustainable water source. With more data and more organization, the prospects of functional database typologies such as these could greatly impact archaeological project planning and non-invasive research.


Cordell, Linda

1997    Archaeology in the Southwest. Academic Press, San Diego.

Ellis, Florence Hawley, and Laurens Hammack

1968    The Inner Sanctum of Feather Cave, a Mogollon Sun and Earth Shrine Linking Mexico and the Southwest.  American Antiquity 33(1):25-44.

Fagan, Brian M.

1995    Ancient North America: The Archaeology of a Continent. New York, NY: Thames and Hudson.

Fewkes, Jesse W.

1911    Antiquities of the Mesa Verde National Park. Smithsonian Institution Bureau of American Ethnology, Bulletin 51. Washington Government Printing Office.

Gardin, Jean-Claude

1980    Archaeological Constructs: An Aspect of Theoretical Archaeology. Cambridge University Press.   

Goldberg, Paul, and Richard Macphail

2006    Practical and Theoretical Geoarchaeology. Malden, MA: Blackwell Publishing.

Haury, Emil W.

1975 The Stratigraphy and Archaeology of Ventana Cave. University of Arizona.

Jennings, Jesse and Richard N. Holmer

1980 Cowboy Cave. Salt Lake City, UT: University of Utah.

Kidder, Alfred Vincent

1962 An Introduction to the Study of Southwestern Archaeology. New Haven, CT: Yale University Press

Mangelsdorf, Paul

1954    New Evidence on the Origin and Ancestry of Maize. American Antiquity 19(4):109-410.

Simmons, Alan

1986    New Evidence for the Early Use of Cultigens in the American Southwest. American Antiquity  51(1):73-89.

Tagg, Martyn

1996    Early Cultigens from Fresnal Shelter Southeastern New Mexico. American Antiquity 61(2):311-324.

Waters, Jennifer

            2008    Early Dog Burials in the Southern Southwest. Archaeology Southwest 22(3).

[1] Cowboy cave and Walters cave

[2] I chose Kidder’s regional designation, because he organized them around major river drainages. One of the criteria I had intended to analyze was the proximity of the site to a water source.

[3] A first draft of the spreadsheet, I developed in Microsoft Excel. This was a table with the categories for data collection entered as the fields, or columns; and the site names as the records, or rows. The rest was blank and intended for manual data collection through literary research.

Published by Nikki M

Applied Anthropologist and Digital Dance Specialist

Leave a comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: