A Bayesian Approach to Dating Agricultural Terraces: A Case from the Philippines

Article excerpt


Practices of relative dating techniques in archaeology have been modified by newer and more easily accessible radiometric dating methods (i.e. AMS, TL, etc.) and the realisation that some archaeological strata are not formed by straightforward top-to-bottom, younger-to-older relationships. The diminishing reliance on 'stratigraphic superposition' in archaeological reconstruction and greater emphasis on radiometric dating (especially [sup.14]C dating), have created a pitfall for arbitrary interpretation of the calibrated information provided by laboratory results, which might not relate to the archaeological event being dated.

Following Dye's (in press) call for a standard methodology for calibrating [sup.14]C results and incorporation of stratigraphic information in the calibration, this article proposes the use of Bayesian modelling (Buck et al. 1996) to date agricultural terraces, which by nature have layers with a chaotic mixture of materials. I use the Ifugao rice terraces of the northern Philippines as a case study to illustrate the suitability of Bayesian calibration and modelling in establishing archaeological chronology. Anywhere in the world, dating agricultural terraces presents methodological difficulties because of their construction technology and use. However, as this article illustrates, a Bayesian approach addresses the problem by incorporating stratigraphy, ethnographic information and [sup.14]C dates in the calibration process.

In addition, the Ifugao case study offers a significant contribution to anthropological studies of agricultural intensification. Foremost of these is the illumination of relationships between landscape, water management and social organisation. However, to understand these relationships in a diachronic perspective, it is necessary to determine when the Ifugao constructed the earliest rice terraces.


The Ifugao rice terraces

The origins and age of the Ifugao rice terraces in the Philippine Cordillera (Figure 1) continue to provoke interest and imagination in academic and popular debates (Table 1). For Southeast Asian scholars, dating these terraces is critical for understanding Philippine prehistory and Southeast Asian patterns more generally. Beyond the scholarly community, the terraced Ifugao landscape has captured the world's imagination as an important cultural landscape (UNESCO 1995). To date however, insufficient work has been undertaken to determine either when the terraces were first constructed, or the period of time involved in building this tiered landscape.

The Ifugao are one of several minority ethnolinguistic groups in the northern Philippines, and one of the best documented by ethnohistoric and anthropological scholars. At the turn of the twentieth century two prominent figures in Philippine anthropology began an intensive investigation of the Ifugao (Barton 1919, 1930; Beyer 1926, 1955). Both scholars proposed a 2000-3000 year old origin for the Ifugao rice terraces, using observations and qualitative speculations on how long it would have taken the Ifugao to modify the rugged topography of the area (Figure 1). This 'long history' has become a kind of received wisdom that finds its way into textbooks and national histories (UNESCO 1995; Jocano 2001).

At the other end of the spectrum, several scholars have proposed a more recent origin of the Ifugao rice terraces. Using evidence from lexical information and ethnohistoric documents, these studies suggest that the terraced landscapes of the Ifugao are the end-result of population expansion into the Cordillera highlands in response to Spanish colonisation. Lowland-mountain contacts are known even before the Spanish arrival. These contacts might have facilitated the movement of lowland peoples to the highlands when the Spanish established bases in their locales (Keesing 1962).

Today, the Ifugao practise a combination of wet-rice terraced farming and swiddening. …