Who makes the assembly rules

I asked whether there was any evidence that plant communities exist, in the sense of assemblages in which the co-occurrence of plant species is restricted by generalized assembly rules (Wilson 1991). I gave examples of possible assembly rules; it was not intended to be an exhaustive list. However, I would not like in advance to "agree on a list of n properties a system must have", as Keddy (1993) advocates. It's like hunting for the Yeti (Wilson 1991). We can have various guesses what it would be like, and look for that kind of beast, but we have to keep an open mind until/unless we find it. Similarly with plant communities, we can have various guesses what the assembly rules might be, but the real question is which rules we can find evidence for in nature. The plants make the assembly rules, we have to find them. The question of whether plant assemblages exist, in which the co-occurrence of species is limited by generalized assembly rules, is a basic question of plant community ecology. In fact, the question. It must be answered by careful examination of plant communities. We have to: formulate hypotheses on what generalized assembly rules might exist, find ecological situations where we can test those hypotheses, see whether there is any evidence for the rules, subject to the usual statistical tests to ensure that the patterns are not due to chance. M.B. Dale (1994) seems to be advocating something very similar to this. However, I have always sought simple, summary rules. As Dale implies, there is a limit to the depth of insight we can obtain about a community by examining its static structure. Obtaining a complete set of assembly rules, as Dale suggests, would be difficult in practice. The question I asked (Wilson 1991) can be phrased in various ways: "Do plant communities exist in any more meaningful sense, as integrated, discrete entities?"; "Are there phytosociological limitations on the occurrence or abundance of plant species?"; "Are there 'Assembly' rules ... ?" . Of course, it is possible to change the definitions of terms in these questions, so the intended meaning is removed and the questions become tautologies. Palmer & White (1994) suggest doing this with the term 'community'. This would indeed "put the debate about their existence behind us" (Palmer & White 1994), but by sweeping it under the carpet. I fail to see how not answering the question advances the science. There is a fatalistic thread running through some of these Forum contributions. Keddy (1993) implies that such questions about assembly rules are 'untractable'. Palmer & White (1994) believe that: 'Vegetation science has no hope of discovering such a force'. Dale (1994) questions whether: "the effort expended on finding them is worth the gain". I do not believe we should give up. Assembly rules can be found. I cited (Wilson 1991) examples of apparent assembly rules. Since then, I have found evidence for limitation to species coexistence at a fine scale (Watkins & Wilson 1992; Wilson et al. 1992a), and for community structuring by guilds (Wilson & Roxburgh 1994; Wilson & Lee in press). Not only can it be done, it has been done. The problem, as I see it, is that if there are no assembly rules there to discover, then vegetation science is reduced to stamp-collecting. Take, for example, the British National Vegetation Classification (Rodwell 1991, etc.). This is a monumental work, widely admired