On Friday 3 August, I accompanied one of the two fully-trained and now practically-experienced interviewers to another town camp to begin the count there. This was a more developed camp with a block and street layout and houses numbered from one up into the twenties, but no street names. We approached one of the houses which had people visibly in and around it that morning and the collector asked for the person whose name was listed against this house on the Dwelling Check List derived from the Tangentyere housing list. It was quickly ascertained that this person no longer lived there and that others had taken over. A middle-aged man offered himself as household spokesperson and, through that process, became nominated as 'Person 1' for the SIHF. Once more, there followed a 15-minute process of adding household members' names, genders, ages and relationships to a SIHF. What emerged was an extended family household of middle-aged brothers and sisters and their partners, and a more elderly father of those brothers and sisters. There were no children physically evident at the household and an inquiry about this elicited the response that the children were back in communities out bush. This was largely an adults' place for town-based living, including a fair bit of drinking, although clearly the links with communities and children out bush were quite strong and frequent. Indeed while we were there a taxi load of five more people arrived at the house, complete with blankets and including two children. Also one of the adults present offered the information that he and some others were that afternoon heading out to Yuendumu for the sports weekend and that they might not be back for a while.

Once the initial process of filling out the SIHF had been completed, attention turned once again to filling out SIPFs. Four were completed within an hour and before people started wandering off and doing other things for the day. By 10.15 a.m. the collector was having trouble getting those listed on the SIHF to sit with him and do a SIPF. They had contributed to the filling in of the SIHF, watched two or three people do SIPFs and were now losing interest in the process. The interviewer too was getting tired and did not want to have to push people. We adjourned from the camp for a break and for the interviewer to buy some more cigarettes. He had just given away half a packet as a part of the process of interesting people in the census collection process!- 84 - When I adjourned to my office I wrote the following in my field journal:

I wondered whether SIPFs would ever be obtained for the six other people in that morning's dwelling who had been listed on the SIHF, since some were now heading off to Yuendumu, and whether the family that had arrived in the taxi with their blankets would ever be counted. They were now staying at a house where, like the small camp of the day before, the census had for all intents and purposes largely been done, though there were still SIPFs for that household which needed to be filled out.

The central Australian census manager was, at this stage of proceedings, very much aware that he needed to get more interviewers working in the town camps. He had initially been aiming for between eight and ten interviewers, on the assumption that they could each complete between 100 and 200 forms over a two or three week period and thus potentially together count somewhere between 1000 and 1500 people in the Alice Springs town camps. Tangentyere had over recent years attempted some very basic counts of its own which had indicated a town camp population of between 800 and 1300 and this had formed the basis for the census manager's approach. However the initial recruitment and training drive had not yielded anything like this number of interviewers. Another attempt would have to be made.

At this point, the attention of the central Australian census manager was fairly fully taken up with the upcoming conduct of the larger, more general census only four days away. He did, however, manage to recruit three more potential interviewers through the Tangentyere Job Shop and conduct a training session for them on census day itself, Tuesday 7 August. Although I did not attend this training session, I did on that day have further contact with the census interviewer that I had observed in action the previous Thursday and Friday. He reported that he had 'not done all that much' as he had had a few personal issues to deal with. He had also clearly been somewhat discouraged by the experience of working on his own, as he suggested to me that it had been better working in a team, like he and the other interviewer and the CC had done on the Thursday morning. He intended checking in with the central Australian census manager in the next day or so, but was clearly finding the work pretty tough going.

The day after census day I met again with the central Australian census manager and a team of people from Tangentyere whom he had put together to do the homeless count in Alice Springs on census night. Although the team was drawn from Tangentyere, that night's work focused on places outside the 19 formally recognised town camps. It was intended to enumerate people who were sleeping rough that night, in places like the Todd River bed, which runs through the middle of Alice Springs and provides a convenient- 85 - sitting place for Aboriginal people during the day and sometimes a sleeping place at night. This homeless enumeration appeared to have gone quite smoothly. It involved only a single short personal form with nine fairly straightforward questions (see Appendix D). So maintaining the interest and attention of numerous individuals over an extended period of time through both SIHFs and SIPFs was not an issue. For the interviewee, this was a quick, individualised one or two minute process, while for the interviewer it was simply a process of dealing quickly with one individual at a time and then moving on to find the next. The homeless persons census process enumerated 115 people in the Alice Springs area that night, all of whom identified themselves as Aboriginal.

While the central Australian census manager was clearly very pleased with the way the homeless persons census procedure had worked out, he was equally clearly quite worried about how the town camp enumeration procedure was going. He was aware that the original two interviewers had not done all that much and that the three he had trained on 7 August were only just getting going. He set about recruiting some more interviewers, including one who had helped with the homeless census and found it easy, and organising yet another training session, while also trying to provide some more support for the five interviewers and two CCs who were actually out there working.

On Thursday 9 August another training session was held for three new potential interviewers. Two were retained to the end of the training session and then taken into the camps under the supervision of the two CCs on Friday 10 August. Collection in the camps, with seven interviewers, continued over the weekend of 11–12 August and on 13 August a meeting was held at Tangentyere to see how they were all going. Some were clearly finding the work hard and were in danger of losing interest. A quick survey of the work that had been done produced the estimate that somewhat less than 100 people had actually been enumerated in the town camps in the 12 days since 2 August. There was still a long way to go. The collection effort in the town camps had been built up to the sort of level that the central Australian census manager had initially envisaged. But the census manager was still clearly worried that he was not going to be able to sustain the interest and effort of this group of interviewers and CCs for the two or three weeks that were still necessary for completing the collection in all the town camps. A decision point was looming.