Chapter III: Methodology

In this dissertation, a realist framework has been adopted to address the question of the relationship between urbanization and deforestation in Rondônia, Brazil. Within this framework, urbanization has been seen as arising from the set of processes, operating at a macro scale, that were described in Chapter II. As suggested by Figure 3, urbanization itself is a kind of conceptual bridge between these processes and at least some aspects of the phenomenon of deforestation in Rondônia. A s is also suggested by that diagram, urbanization need not be the only such connection, but it is the one chosen for analysis here.

This chapter describes the methodology used for an iterative research project examining the specific linkages between urbanization and deforestation in Rondônia. It includes a review of the methodological opportunities and constraints imposed by the realist framework and a description of the research methodologies employed.

One of the key features of crit ical realism, as described in Chapter I above, is that an effort is made to prevent the framing of questions and the design of methodologies from constraining the results of the research. This requires flexibility so that the researcher can benefit from ongoing learning as the research unfolds. Another key aspect of realist research is the recognition that social systems are open, and therefore not subject to controlled experiments. The isolation of causes from patterns is therefore problematic, and the foc us of research must be on both causal links themselves and formal groups or patterns.

1. Research Design
1. Study Sites: Urban Places in Rondôn ia

 

 
 
 
 
 

The município seats are the focus of rural-to-urban migration in the Amazon region (Browder and Godfrey 1990). In Rondônia, they vary in population from a few thousand to well over 100,000 in the case of Porto Velho. Some of the cities are located on transportation routes that were developed in the rubber-boom period of the late nineteenth century, such as the Madeira, Mamoré and Guaporé rivers, while others are found on the Cuiabá-Porto Velho Highway (BR-364), which was built in the 1960s as part of the regional economic development programs described in Chapter II above. Most of the cities along the highway are in a zone of widespread deforestation, whereas several of the riverside towns are in areas that have not experienced extensive deforestation to date (based on 1995 LANDSAT imagery obtained via the Internet from the National Institute for Amazonian Research (INPA)).

1. Data Acquisition

 

 
 
 
 
 

A rolling interview technique was adopted, by which a starting point could be selected, however arbitrary, that would be likely to lead to further, increasingly productive interviews. I chose to make Porto Velho the center of my activities for a variety of reasons. First, I found that my initial contacts in Porto Velho with professors at UNIR, including one English p rofessor from whom I rented a room for the duration of the fieldwork, led me to a steadily widening circle of useful contacts both in Porto Velho and throughout the state.

It could be argued that this technique would introduce bias into the data-collection process, because the selection of subjects for interiews could be limited in some way by selection of initial contacts. Sayer (1992) addresses this concern in his defense of intensive research techniques described above. In the present case, it sh ould be noted that the intensive phase was used to draw hypotheses, not conclusions, so that whatever bias might have been introduced by the rolling-interview technique would be eliminated during the extensive, hypothesis-testing phase of the project.

To supplement the information I acquired through interviews and direct observation, I sought published data related to the growth of urban populations in the state and the economic activities undertaken in them. I discovered that in addition to the net work of knowledgeable contacts in Porto Velho, many relevant published data sources unavailable to me in the United States were to be found within the federal, state, municipal, and non-governmental institutions in the city. Gaining access to these data sources was more time-consuming than I would have expected, as was their interpretation. I comment further on this aspect of the data acquisition process in Chapter V below.

Even given the advantages of basing the research effort in Porto Velho, dire ct experience in other cities and rural areas of the state was clearly required in order to understand more fully the data that I was able to collect. For this reason I arranged visits to rural areas of Porto Velho as well as to rural areas and urban centers in other municípios, including Jamari, Candeias do Jamari, Rolim de Moura, and Ouro Preto do Oeste. In each case (except for my first visit to Candeias), my initial contact was with somebody I had met in Porto Velho. In each place, however, I qui ckly moved beyond my initial host to interview a variety of knowledgeable local people. I asked not only about the development of the town but also about how my informants had come to live there and from where they had migrated. I also walked extensively, taking informal transects of each place -- usually with my camera -- in order to learn as much as possible about economic activities, land use, and the cultural landscape of the place.

1. Analysis

The second phase of the research built upon the preliminary results of the first. Because the intensive phase of the research was undertaken in a variety of cities and towns throughout the state of Rondônia, it raised important questions that were best addressed at the state level, using mostly the published sources that were themselves gathered as part of the rolling interview process.