Tag Archives: Caribbean Archaeology

A Look at the Archaeology of the Caribbean with an emphasis on Puerto Rico

A Look at the Archaeology of the Caribbean with an emphasis on Puerto Rico


Delving into the archaeology of Puerto Rico has immersed me into what seems like a dismaying profusion of cultures, sequences, styles, names, and a sleuth of other complexities. I could not in good conscious create a paper on Puerto Rico itself without understanding the Caribbean sphere itself. I discovered that this is a region blanketed with multiple islands that are connected by water and in a sense, exist as giant ethno-organism. Throughout my research, the history of the Caribbean revealed a bit of itself, yet, provoked one to seek its mysteries.

To condense the archaeology of the Caribbean in the following small paragraphs would be an injustice. I believe there is no definitive source that can capture its complexity in one volume. Therefore, this paper will attempt to do three things, 1) provide a look at the archaeology of the Caribbean in order to understand the archaeological background and theories, 2) take a glimpse into the alternative approaches that can provide insights, and 3) a critical look at the long established model that governs the Caribbean, how it affects Puerto Rico archaeology, and view an emerging approach that many Caribbean archaeologists are favoring. A summary and final thought will follow.

A Look at the History of Caribbean Archaeology

In the early 1900s, American archaeologist Jesse Fewkes analyzed artifacts from many Caribbean collections and stated they indicated a sequence in time or distinct cultural stages. The stages he offered were cave dwellers, agriculturalists, and the Carib. This became known as three cultural epochs. Fewkes suggested looking for geographical areas of similar culture instead of creating a chronology. Archaeologist Gudmund Hatt entered the scene in the 1920s and performed archeological research on the island of St. Croix, followed by about 30 excavations on the Virgin Islands. Unlike others before him, he meticulously analyzed the stratigraphy of the sites he excavated and proposed a three-phase sequence for the Island’s prehistory. The first phase was characterized by flaked stone tools, followed by painted pottery in the second, and pottery with no paint in the last phase. His three-period sequence was refined and expanded on in the years after his work. The dawn of the 1950s saw a wealth of new and systematically collected data came from the work of archaeologist Irving Rouse in Haiti and Cuba, and other works supported by the New York Academy of Science in Puerto Rico. With this data, the late 50s saw Irving Rouse propose a relative time series with periods, marked by Roman numerals I through IV; from earliest inhabitants till after the Europeans arrival.  Alternative classification systems  were developed through the years and were introduced by numerous Caribbean scholars that included French, Dutch and Spanish archaeologists. They ranged from dividing pottery into domestic and ritual use, classifying pottery by decorative/size/shape factors, grouping pottery by functions using ethnographic data, up to the Southeastern System that identified types, local sequences, and specimens. Some archaeologists used the new systems around the Caribbean and the Southeastern System was notable, with the approach employed at Grenada’s Windward Islands, Antigua, St. Thomas, and the Bahamas.

These early years of Caribbean archaeology through the 1960s were largely based on classification, description, and time-ordering of archaeological artifacts, assemblages and their sites in order to make sense of the seemingly complex culture variations. The goals were to establish cultural historical sequences throughout the Caribbean. Archaeologists spent significant efforts on developing culture areas solely based on excavations and its seriation of artifacts, artifact collections, and the increasing overall comparison of artifact styles and compositions of assemblage through space. There were no absolute-dating techniques in early years.

As the alternative new approaches were randomly used from the 1950s through the 1960s, Rouse’s proposed time series model from late 1950s had considerably bloomed all along. I think it is vital to understand a bit of Irving Rouse’s background which will help understand this character before we go on.

Irving Rouse began his career at Yale in 1930. His undergraduate work was in plant science with the intention to go into forestry. At the time, his savings for school was $500, but that was swallowed by bank in the stock market crash. Working through odd jobs to survive, he ended up raking leaves on the Yale campus. This led to a chance encounter  with Dr. Cornelius Osgood, the curator at the Yale Peabody Museum. Osgood took a liking to Rouse, and gave him a job cataloging anthropology collections at the museum. It was at this museum, that Rouse first laid his hands on collections of artifacts, sorting and cataloging them during his time at Yale. Osgood’s confidence in Rouse led him to start developing an interest in his field. Once he had completed his undergrad degree in plant science in 1934, he considered that field to be matured, and with the fascination he developed working through collections at the museum, Osgood had urged him to pursue graduate studies in anthropology. This led to Rouse completing his Ph.D in 1938. Osgood already had an interest in the Caribbean and was working on developing a comprehensive and systematic program of regional study. With private funding, he created the Caribbean Anthropological Program (CAP) in 1933. Yale graduate Froelich Rainey was sent sailing through the Bahamas to find archaeological sites, but found nothing of value. Osgood then sent Rainey and Rouse to Haiti in 1934 under CAP to undergo an archaeological investigation, which later served the basis for his dissertation. Rouse next headed to Puerto Rico in 1936, 1937, and 1938 that was sponsored by the New York Academy of Sciences that entailed the Scientific Survey of Puerto Rico and the Virgin Islands—which was a notable program that examined and combined investigations from many scientists regarding the archaeology and environment of Puerto Rico. In 1939, he examined Caribbean collections in European museums followed by fieldwork in Cuba with Osgood in 1941. He investigated prehistory in Venezuela with frequent visits from 1950 to 1957. The National Science Foundation gave a grant to Rouse in 1963 to obtain the first radiocarbon dates from the Caribbean. 31 samples were analyzed at Yale from Guadalupe, Venezuela, and Puerto Rico sites. The preceding years brought more obtained dates which Rouse used to refine the region’s chronology. Rouse was an instructor at Yale in 1934 and evolved to full professor by 1954. Not surprisingly, he worked at the Yale Peabody Museum as assistant curator after his Ph.D, becoming associate curator in 1947, and finally research associate in 1954.

Throughout Rouse’s anthropology journey, starting from sorting collections at the Peabody through fieldwork and examining Caribbean collections, he was obsessed with the power of classification. This thinking originated his undergrad years in plant science. His sound background in taxonomy and being influenced by linguistics, he believed that knowledge was obtained through classification. His belief was that if you could identify cultures and throw them in the correct containers of time and space, you would ultimately create a total culture history. Rouse’s lifelong questions were the identification and meaning of human migrations and others included how to separate events and processes that occurred in Puerto Rico’s past by its neighbors. His taxonomic approaches created the theoretical framework for his questions.

That chance encounter with Osgood led Irving Rouse to be later coined the “father of Caribbean archaeology.” There is no doubt that Rouse was heavily influenced by Osgood who nourished and paved his archaeological explorations.

The proposed time series model that bloomed in the late 1950s was derived from Rouse’s earlier fieldwork in Haiti 1934, which was published in 1939 in two monographs—the first was entitled Prehistory in Haiti: A Study in Method. This became a defining landmark in Caribbean archaeology. Rouse outlined the question what is the archaeologist is trying to do, why, and what method best suits his need. His work presented a way to systematically characterize, classify, and compare assemblages of artifacts across and within islands and to the accordant mainlands. This landmark work also introduced his method of “modal analysis.”

The time series model, or “modal” approach, that many increasingly adopted through the 60s was refined and subdivided. As I noted above, the approach was marked by periods, identified from I through IV, which then became subdivided, for example, period II-a and period II-b. He argued that pottery styles are characterized by an exclusive set of material, shape, and/or attributes which ultimately leads to the identification of the period and people behind it. Essentially, that each style represents the entire pottery collection of the people in one cultural period at a specific location.

He noted that how specific will vary by each case, but typically tends to be local or island-bound. Assemblages that unearth artifacts containing a unique set of characteristics in material, shape, and/or decoration defines the style. A group of styles is based on characteristics that are related throughout time and space, therefore established as a “series.” Series became dubbed by the suffix “-oid.” Archaeologist Gary Vescelius who used Rouse’s approach, suggested the addition of subseries, that consisted of smaller geographical, cultural units that share a common ancestor. Rouse refined and implemented the idea of subseries, that was marked by the suffix “-an.” For example, the first described preceramic assemblages at the site Casimira was called Casimiroid, hence the Casimiroid series. Within this series, a further distinction was proposed between the sites Cayo Redondo (Cuba) and Couri (Haiti). Redondo is the first representative and Couri the first good case. Therefore, under the system of series and subseries, the two subseries are designated as Redondan Casimiroid and Courian Casimiroid. Likewise, Saladoid refers to the site Saladero, Venezuela (a 500BCE seafaring culture that later populated the Caribbean and introduced ceramics). This naming scheme is crucial to understand, as it is heavily embedded in Caribbean archaeology.

Rouse in a sense devised an evolutionary, hierarchical system created in such a way that styles are classified within subseries and subseries within subseries. The logic is that artifacts can be fed through this approach, creating sense out of cultural sequences, and ultimately tracing back to its common ancestor. Rouse strongly believed that cultural assemblages were part of a single line of development.

A Look at Some New Methods

Researchers Christy de Mille and Tamary Varney had questions of the manufacturing methods, production organization, and variability across sites of the Saladoid lapidary industry. That industry is often cited as evidence for interisland movement and for its economic stone goods, notably stone beads, that were produced widely in the Saladoid period.  Since this a crucial item that was significant in the socioeconomic world of the Saladoid, and led others to hypothesize trade networks and regional production centers, they created a method to examine the bead creation process. First they used a scanning electron microcopy (SEM) on molds of stone beads from Saladoid sites on Antigua. What they discovered was fine concentric rings in the mold, evidence for drilling technology. Next, they attempted to create a bead mold, by using a modern day drill tool, with a wooden drill bit. They placed the drill on top of a calcite tablet with  quartz sand (to act as an ancient abrasive), and  found it took six hours to drill a 5mm hole with a modern day tool whizzing at 10,000 rpm. A similar experiment was done, but with small wooden drill tip, that failed. I think that while this provided some answers to their questions, it raised more: if there is evidence of concentric rings in Saladoid stone objects—that we cannot replicate with a power tool—how was it done? They are continuing to refine their methods to understand more.

Archaeologist Jeff Walker employed the experimental archaeology approach to study the manufacturing techniques and functions of lithic artifacts discovered at the Sugar Factory Pier site on St. Kitts. Armed with ethnographic data, he replicated stone tools that he used to peel and grate tubers, sawing, carving and planting wood and inscribing shell material. Walker also used the combined ethnographic and ethnohistorical data that brought forth new perspectives on the site’s activities, subsistence patterns, and possible trade networks. Establishing the origins of trade and material culture can assist the archaeologist in unlocking data that can better understand context, such as the significance of a piece of temper outlined in a fantastic read, “What Good is a Broken Pot?”

In order to understand the source of origin for clay and temper, the first petrographic analysis in Caribbean archaeology were carried out in the late 20th century. Archaeologist Charles Hoffman set out to answer ceramic inclusion questions and discovered that ceramic products on the islands of Anguilla and Barbuda was derived from resources that originated on the volcanic Leeward Islands of the Lesser Antilles. Likewise, it was also discovered that the Cayo pottery assemblage from St. Vincent contained caraipe, the bark of the kwepi tree, which is not found in the island system but are endemic in Trinidad and South America. It is now hypothesized there was contact between the islands and the mainland of South America.

These are some of the techniques that will increasingly add data that perhaps can be used cumulatively in search of establishing the forgotten past.

A Critical Look at Rouse’s Model

In the book “Rethinking Puerto Rican Precolonial History,” it feverishly forges chapter after chapter with two primary goals in mind – discredit Rouse’s long standing model and slowly revealing a new model. Rouse’s assumptions have always been that the Caribbean’s period were all descended from one line. Ramos argues the opposite and also argues that Rouse’s model is one that keeps Caribbean archaeologists in a strait jacket, unable to make progress.

A site in Puerto Rico known as La Hueca-Sorce has unveiled numerous artifacts that aren’t with within the same established context. Again, Ramos questions Rouse’s model and assumption – that all inhabitants were in ethnic isolation, giving rise to the succeeding cultural sequence. But the archaeology at this site has harvested avian pendants, adzes, hafts, and unique three point cores/flakes. Lithic analyses provided evidence that the limestone from some of the artifacts was not a product from the island but was matched to St. Martin. Ramos asked why would this complex society simply leave and not fit into Rouse’s model. Further examination, showed that the sculpted bird artifacts were not native but were found in South America, Central America, and Trinidad.

Perhaps the most compelling pieces of evidence are the jade artifacts. These jade artifacts were not only found at La Hueca-Sorce, but also at another site Punta Candelero. Jade is not found in the area. What is profoundly unique about these artifacts, is the discovery of similar artifacts in Costa Rica. The designs of the bird and frog are remarkably similar. Ramos examines the artifacts at the Costa Rica museum tells of macroscopic analysis that provided evidence of coming from a common origin.

Ramos also argued that part of reason why many followed Rouse’s modal approach, is due to the initial radiocarbon dates that he collected and analyzed half a century ago. Ramos offers a calibrated means of 20 radiocarbon dates for each of the pottery styles on the island and matched them up to Rouse’s calibration.  Clear distinguishable differences are apparent on the chart comparison concerning time scales.

Ramos questions Rouse’s assumption of a monocultural landscape of the island. A line of evidence comes from the archaeologist Pagan Jimenez, who has done research at the sites Puerto Ferro and Maruca. Jimenez has uncovered evidence of agricultural; sweet potatoes, tannia, beans, and maize. These were recovered from pre-Arawak stone tools by using a starch analysis approach. Evidence of domestication and shaping of the landscape was further detailed by Ramos, arguing that there was multicultural presence in the past.

He challenges Rouse’s long standing assumption that there was a single migration of pottery creating agriculturalists (the Cedrosan Saladoid) to the Antilles. Ramos noted the incredible amount pottery that did not fit Rouse’s series and subseries model. And because they did not fit, and the archaeologist restricted by the model, they were labeled anomalous. He critically argues that these “anomalous” artifacts and sites should be examined as they may be showing the development of the initial occupants. Going back to the new calibrated radiocarbon means, he uses it as substantiating evidence to depict that Rouse’s very own “Ostinoid” styles goes farther back to a different time, thus explaining that why the anomalous artifacts did not fit.

Ramos stated that the approach is  “…not Rouse’s fault. The fault lies with those of us who have rested comfortably on the pillowed squares that form the basis of his model and have not made concerted attempts to critically evaluate its primitive axioms.” I agree with Ramos. Some models/theory when established, are simply not questioned, and when new perspectives shakes the foundation of a new theory, some are not ready to depart—either ego takes over or the need of security from following tradition.

Ramos concludes by offering a model, the “reticulate model” for Antillean archaeology in contrast to the long prevailing phylogenetic one that has dominated Caribbean archaeology for nearly a century. The idea is simple, to mark the Caribbeanscape as a giant vein system and match corresponding vein systems. He explains that his model is “most adept for addressing situations that feature plural scenarios where multiple developments that constitute each other over the long term are registered.”  The wide region littered with islands that are infused with abundant cultural diversity result from the “reticulate comings and goings of people from island to island.” Rather than assume a single descent as previously thought, the reticulate model will open doors to resolve anthropological processes that unfold across a cultural vista composed of multiple networks of people with distinct ancestries that interact with each other for prolonged periods of time.

Thoughts and Conclusion

We have visited the key figures in the development of Caribbean archaeology and the growth of the modal that has been, and still is, a major player. An understanding of Rouse background provided a deeper understanding of his arrival at his modal, the influences behind him, and most importantly the range of fieldwork he accomplished. We have explored but a miniscule of techniques that are outside of conventional theory, but nevertheless attribute to our cumulative knowledge. Finally, we have seen a handful of evidence that begs to collapse the long standing modal in search for an encompassing and flexible approach.

All one has to do is glimpse at a map of the Caribbean to be enthralled by its simplicity yet complexity. After all the readings in an archaeological theory course, I could not believe that dominant Caribbean theory was based on the assumption that all lines of pottery had a common ancestor, and if it didn’t fit the modal, it was brushed aside. It seemed almost infantile. Ramos seems to be building a following with the reticulate model approach. I think it would fit perfectly, and wonder how long before it is adopted as mainstream. But would that also mean revisiting every site and every artifact stored in collections to correctly attribute its cultural association?

The emerging new techniques in archaeology are simply fascinating. How we can analyze ancient starch and reveal a dictionary of data from a site is science at its best. Archaeological theory is a must, but science should be merged in its tenets—they must coexist in harmony.

Rouse seems to have built an empire around his modal with many that adore him. In fact, one book used in my research, Ancient Borinquen, is dedicated to him. In all the side literature I’ve read on Caribbean archaeology, it seems that none go without his name adorning their citations. It would be ignorant to say that his modal didn’t provide a tremendous amount of use to Caribbean archaeology, but at what cost?

Ramos almost seems unforgiving against Rouse, yet remained a scholar and elegantly plucked away at Rouse’s modal. I once read a quote in a dusty L’Amour book, “Never judge a man against the canvas of his time.” We cannot judge Rouse, but the canvas is now in our time. What we paint it with is up to us.

-antonio kuilan


Hoffman, Corrine L., et al. Crossing the Borders: New Methods and Techniques in the Study of Archaeological Materials from the Caribbean. Alabama: University Alabama Press, 2008. Print.

Siegel, Peter E. Ancient Borinquen: Archaeology and Ethnohistory of Native Puerto Rico. Alabama: University Alabama Press, 2005. Print.

Ramos, Reniel Rodriguez. Rethinking Puerto Rican Precolonial History. Alabama:University Alabama Press. 2010. Print.

Wilson, Samuel M. The Archaeology of the Caribbean. New York: Cambridge University Press. 2007. Print.

“Benjamin Irving Rouse.” Biographical Memoirs: V.90. Washington, DC: The National Academies Press, 2009. Web. http://www.nap.edu/openbook.php?record_id=12562&page=306 Accessed: April 25th, 2013.