Fieldnotes: Coding using CSAC Context Codes

Stephen Lyon, Dept. of Anthropology, Durham University. 2006

This is copied from a section of my doctoral thesis (U. of Kent 2002 - available to download). An variation of it is also available in Lyon. S. 2004. Anthropological Analysis of Local Politics and Patronage in a Pakistani Village. Lampeter: Edwin Mellen Press. pp: 47-49.

To see edited fieldnotes of selected fieldnotes go to my site.  An older version of the CSAC Context Codes is in Lyon's doctoral thesis.

Fieldnotes

I used the CSAC Content Codes (Fischer et al. 1996) to create meta categories in my field notes based on the content. I did this on a daily basis. I wrote up my field notes in the evening and sometime the fol lowing morning I entered the content codes and wrote an abstract for every note. At the end of every month I made an edited copy of the previous month's field notes available on the website (Lyon 1999). While content coding is of obvious value for the pos t fieldwork phase of research, it is equally valuable during the data collection phase. After allowing myself the liberty of 'stream of consciousness'field note writing, I forced myself to focus on what my notes were about. What was my field note entry actually saying? Who was involved? W ere issues of time, language or location important? All questions I was able to answer easily in the fie ld, but for which I would probably have no answer now. Content coding allows cursory formal description without complicating the research in sophisticated analysis before all the information is produced. Apar t from other considerations, if I did not understand what my field notes were saying the day after I wro te them, then I was doing something seriously wrong and the content codes served to keep me aware of tha t. The CSAC Content Codes extend what the Outline of Cultural Material (OCM) codes) do for the Human Rel ations Area Files (HRAF). Where the OCM are hierarchically organised and non-propositional, however, the CSAC CC provides a way of producing meta tags which can be meaningfully li nked (in ways more sophisticated than simply boolean searches). That is, tags are not hierarchically org anised and may be combined in any combination to construct contextual propositions about data. The follo wing abstract of a field note offers an example of how coding for social and cultural context may enhance the usability of qualitative material (this is a modified version of the Appendix in Lyon 1998):

• Abstract:
• T:Thread DocProj:
• K:EthnoInt:GoodEx:
• L:MetaCon:Behav:intervu
• M:Agent:Grp:
• N:Prep:down:
• O:Role:Care:low
• H:Jur:Prot:
• D:Soc:Status:

(Down at the hotel and barbershop conducting some semi-formal interviews about when and why people go to zamindars for help. Most people do go to zamindars (sometimes indirectly through the elder members of t heir family) for everything from food to broken tractors to ill children to enemies who want to beat the m up. Not all zamindars help people. Heads of households seem to take this role more seriously. People f rom outside Bhalot reported that they went to Bhaloti Maliks before their own village zamindars. There w ere logical reasons for this (neighbouring land, they do most of their work for Bhaloti Maliks, their dh ok is closer to Bhalot than the official village it is attached to))

{T:Thread DocProj:

I had three major strands to my research. The first was my thesis. I created a Thread called DocProj (Doctoral Project) which could label all notes directly concerned with my thesis. I also had a Thread to id entify notes related to Ethnicity, and a Thread to identify notes on Development. This allowed me to glo bally extract data for inclusion in different simultansously run projects.

{K:EthnoInt:GoodEx:

The Ethnographic Intent (EthnoInt) term is a meta-meta tag. It helps to isolate ethnographic incidents by their use-value to me. In this case the incidents provide me with good examples of particular types of behaviour.

Fields