In some guest rooms the regular group is more sharply defined than in others. The D guest room for example was attended by three related lineage segments almost to a man. Any visitor would have been an outsider. In Haci Osman's guest room, or Nureddin's, people dropped in and out and a visitor was much less conspicuous. But all the guest room groups except Zeynel's were collections of people who were close kin or neighbours, or both. The Zeynel group puzzled me a good deal. Here the explanation seems to be negative. Some who were neither kin nor close neighbours went to this guest room because they felt even less welcome and less comfortable in any other, and the alternative, sitting evening after evening with the family, was grim. One guest room had to rank lowest and to collect the left-overs. The only guest room whose owner ranked lower in the village hierarchy was Nureddin, a less pretentious and more jovial character, whose guest room was organised in response to popular demand, and not on his private initiative for his own prestige or political purposes.
In the autumn of 1951 I attempted a similar enquiry in Elbashï, but the number of guest rooms and room groups was greater, the time shorter, and the season not properly begun. I myself was still an object of curiosity, so that my arrival in a guest room drew in neighbours who would not otherwise have come; and I was still unable to identify everybody readily by sight.
Nevertheless, it is plain that a similar system operated, though with a larger margin of non-attenders, and a proportionately larger number of guest rooms in use, making for smaller groups. Five guest rooms appeared to draw mainly on five main lineages, but with varying degrees of completeness and varying numbers of other kin and neighbours. Several of the better-off villagers used their own guest rooms, but seldom