The sheer impenetrability of local organisation to short-term observers was brought home to me during a visit to Tonga. A team of agricultural developmentalists had made visits to two outer islands in order to gather people’s reactions to proposed innovations. In particular, they wished to monitor women’s reactions to the proposed changes and how the agricultural innovations might affect women or alter their ways of carrying out certain tasks. They had gathered together local people to discuss these matters but found that no-one would speak to them, let alone offer a frank opinion.

One of these consultants practically fell at my feet when he found I was an anthropologist: ‘Thank God,’ he cried, ‘I have finally met someone who can explain what was going on!’ During my years in the field, this reaction to my chosen profession remains unique. It is more usual for anthropologists to be greeted by developmentalists with hoots of derision or cool indifference. We are usually accused of either ‘taking the people’s part’ in arguments over why projects failed to work, or of being fusspots who ‘provide far too much detail’ on issues about which the development people wish to be informed but not know too much. This is especially true of topics such as gender relations or the role of women in agriculture, which are included in their terms of reference—being, together with ‘environmental issues’, current aid donor buzzwords—but which do not interest, or overly concern the consultants.

In this case, however, the social scientist, namely myself, was meant to provide the key to the research dilemma immediately. ‘Tell me why they wouldn’t talk,’ he continued desperately. ‘Why, we even drew diagrams in the sand for them to illustrate the objectives of the project, but the people all sat there solemnly and would not utter a word in response.’ I answered that I could not possibly say, because I had not been to those particular villages nor did I know the people who lived in them. That reply appeared to disappoint. ‘But you’ve been in Tonga, you know the people and the culture!’ Yes, to be sure, to a certain extent, I do, but not those people and not that local situation. I might have added that, by 1996, I am not sure that any people anywhere should respond to diagrams drawn for them in the sand.

To hazard a guess, however, the islanders’ silence was probably because the people in question were waiting for their elders to speak first. If the elders had chosen not to speak—either because they did not understand the issues, or because they understood the issues and did not agree with the proposition but were not prepared to be so ungracious as to say so, or because they understood the issues, agreed with them and had nothing to say—or any one of a number of other possibilities; then, their junior relatives and others who were similarly inferior to them in the village structure would not speak before they did, without being asked by them to speak.

The silence did not necessarily mean that the local community did not have ideas about the proposed scheme, but it did suggest that the consultant team did not know how to tap into that local opinion. Tonga is, after all, a highly structured society where deference is customarily shown to village and family elders, the heads of extended families, ulumotu’a, the skilled and competent people, ivilahi, and the ‘keepers of the land,’ tauhi fonua who are the ones who should be looked up to in a local setting. ‘Oh,’ sighed the expert, ‘if only we had had you with us on the team!’ I only smiled modestly at this novel reaction.

Although it is currently the case that more provision for ‘the social components’ of a project and for social researchers is being made available on consultant teams, equally, I was aware that I might not have had any more success than the team in getting the village elders to speak if they did not want to. My only advantage would have lain in knowing where the difficulty most likely was located. In such a case, it might have been more advisable to have had someone originally from the village with the team, to explain the project to the islanders. But, then, in that case, the local response probably would have depended on how the erstwhile islander was regarded in village terms, and the leaders’ estimation of the person and of the family that he or she represented. If there is any lesson to be learned from the consultants’ discomfiture, it is that the total social situation within which the proposed development is to take place needs to be considered as part of the initial project proposal rather than added later as an afterthought, when it has presented itself as an obstacle to the achievement of the intended goals.