Making progress in understanding global and long-run problems raised b
y human perturbations of ecological systems, and in making that unders
tanding relevant to policy formulation, will require intense collabora
tion between, at least, ecologists and economists. There is far too li
ttle of this now and far too much fairly fruitless argument and lectur
ing-in part because each side seems embarrassed by its lack of ability
to do what it claims needs to be done. Ecological models cannot predi
ct the long-run changes in the multiple functions of natural systems t
hat might follow from human actions such as land clearance or waste st
orage. Economists' valuation techniques cannot convincingly deal with
multiple dimensions of change in the environment or with the non-use v
alues of any changes. An ecological-economic modeling project of 20 ye
ars ago reveals a few lessons for those who would like to see progress
on collaborative research ultimately aimed at correcting these defici
encies. For example, real collaboration requires investment of substan
tial up-front time, investment best made in the context of a mathemati
cal model of the problem to be investigated. The model forces particip
ants to fit their ideas together rather than to string them out indepe
ndently. It also will likely force compromise from disciplinary purity
. There is much to be said for using such projects to train a small gr
oup of Ph.D.s who will have observed and participated in the collabora
tion. The existence of such a group will do more for ''eco-system valu
ation'' than ad hoc and episodic forums involving old dogs, even thoug
h some of these would be happy to learn new tricks.