I arrived at community A on the morning of Monday 6 August in advance of the census party, and went to the school, where the other enumerators were gathering. The CFO, the CC and E1 arrived shortly thereafter by vehicle. Everyone was there except E2, who came about half an hour later.

The session proceeded. It quickly transpired that the enumerators were going directly to filling out the SIHFs, apportioning them between themselves as they went. The CFO and I noticed this separately, more or less simultaneously. He decided to allow further training to proceed 'on the job'. He continued to observe what the enumerators were doing, and intervened with instructions if he saw something not being done correctly, or if he was asked directly for advice. The CC and E1 were also advising the other enumerators.

Household members and visitors were listed from memory on each SIHF in turn, and also absent members of households (who were marked 'no' for a SIPF). These lists were not inviolate: later on, as the enumerators went round the houses and discovered in more detail who was there and who was away, the yes/no column was changed, as people were added or subtracted.

The CFO instructed explicitly that the family name should go on the top of the SIHF, then the head of the household should be listed as person 1. This person, he suggested, would usually be the father, then the mother and rest of family should follow. He reminded the enumerators that for every person the second name had to be filled in as well as the first.

The CFO also instructed explicitly about where the RNO and dwelling number should go (the enumerators were not filling those in until he told them to). The enumerators went to the map to find the dwelling number (the lot number), and then cross-referenced to the Dwelling Check List to find the assigned RNO. I was having trouble with this system, but it appeared to cause no difficulty to the enumerators.

The SIHF also asks for the age of each person and their relationship to person 1. The enumerators used the school roll for the ages of the children (and the teachers), but everybody else's age was a guesstimate. E1 had compiled a list of everyone's ages derived from local clinic records, but he had left it behind at the HA office in community B. He commented that this did not matter, because the ages could be fixed up at the office later.

There was some discussion about how to translate local kin terms into their English 'equivalents'; this topic was to recur throughout the enumeration. 'Wife', 'husband', 'son',- 35 - and 'daughter' were unproblematic: everyone was familiar with how these are applied, even though they do not correspond to local terms. E1 reiterated that English terms had to be used, not local terms. The question of kinship and how it was reflected on the SIHFs and SIPFs is discussed in some detail below.

E1 explained that a SIPF was needed for everyone on the SIHF who was present at the time of the enumeration. 'Even the little children?' he was asked, to which he replied in the affirmative.

There was considerable discussion about how to treat visitors and absent residents: both the enumerators and the CFO were aware of the complex possibilities. E4 asked what to do about visitors from other communities. The CFO said visitors should be included on the SIHF of the dwelling at which they were staying, and marked as 'V'. They should be asked if they had been counted anywhere else, and that if not they should have a SIPF filled out. In cases of doubt it was better to be on the safe side and do a SIPF, otherwise they might be missed altogether. One of the women expressed shyness about asking visitors for their details.

The CFO was also asked: 'How can you ask questions if someone is not here?' The questioner had in mind someone who was away on a trip, and who would be staying in a hotel. The CFO said that anyone like that would be counted at the hotel on census night, so they should go on the SIHF, but should not have a SIPF done for them. More generally, anyone who was away in a place where they would be counted should not get a SIPF. He was asked: What about people who are away at community C or D? The CFO suggested doing SIPFs just in case they got missed there. What about people who would be coming in tomorrow for the funeral of the baby? The CFO suggested doing a SIPF if they had not been counted anywhere else and if the dwelling was still in the process of being enumerated. But if the enumeration of the house had been completed they should not be added.

E1 suddenly remembered the tent in the front yard of one dwelling. It was normally occupied by the mother-in-law and father-in-law of 'person 1' at that dwelling, who were on an extended visit. They were away at community B, but had left initially to attend a funeral ceremony elsewhere. The CFO's view was that since the tent was not a permanent dwelling, and its usual residents were in any case visitors to the community, it should not be registered, and so it had no SIHF, and no SIPFs were attached to it. Its absent residents were not included as visitors on the SIHF for the dwelling.

After all the household forms had been completed, the CFO suggested counting the number of SIPFs needed for each dwelling and putting them in an envelope with the SIHF. E1 and the CFO both stressed several times that the SIPFs had to be filled in at the house wherever possible.

During this process at least one extra person was remembered for one house. There was some further discussion about people who were currently down as 'yes' but who turned out not to be there. The CFO suggested that if they did not come back during the process of enumeration they should be changed to 'no', and that no SIPFs should be done for them. At this point, one dwelling was listed as a vacant dwelling because all the occupants were away. However they returned during the process of enumeration, so the status of the dwelling was changed and SIPFs were done for the returned occupants.

- 36 -

The households were initially divided up between the enumerators as follows:

Thus the workload was distributed fairly evenly, and everyone except E2 and E3 was assigned to their own 'side' of the community. E2, the most senior ZC to the apical ancestors of the two lineages, was assigned dwellings from both sides. One of the dwellings assigned to him turned out to be the most problematic in terms of the enumeration, because it contained some visitors from community E who were not all that concerned with—indeed who actively evaded—being counted.^[2]

As far as I am aware, the final page of the SIHF (number of bedrooms, number of vehicles, etc.) was not filled in at this session, and I do not know when these details were added. I assume that the characteristics of the dwellings could have been added later from HA housing records, and the enumerators would have had a good idea of how many vehicles there were in the community and who owned them.

E1, in his decision to get everyone to complete the SIHFs at this session, was operating on the assumption that the training of the enumerators had been completed at B earlier. In a sense this assumption rested on a minor failure of communication between him and the CFO. The CFO later told me that the 'training' session described above was a unique occurrence: the enumeration had not proceeded at any other community in quite this way. When they set off to begin the enumeration, the enumerators had still not gone through the SIPF in detail, or familiarised themselves with the questions.

However, E1 was operating on another, correct, assumption: that he and his fellow enumerators knew their community well enough to fill in the SIHFs before proceeding to the enumeration proper. In doing so, they spared the interviewees a lengthy part of the process entailed in following the 'proper procedure' (see Will Sanders, this volume). The filling in of the SIHFs at this point turned out to be an advantage rather than a disadvantage: this point will be revisited below.

The session was indeed a training session in some respects. At the end of it, the CFO could be confident that the SIPFs would be correctly identified, and properly cross-referenced to their SIHFs. Between them, the CFO and the enumerators established a clear set of procedures for dealing with absent residents and visitors: the enumerators were alert to the possible permutations and the CFO held to a clear set of principles, based on the standard enumeration procedure, which he applied consistently to these permutations. They also had discussions that were as useful as they could be (which is to say not very) on the translation of kin terms.

- 37 -