Social Visions of Future Sustainable Societies 415in Time. The probabilistic approach assumes the future to be a ‘stationary stochas-
tic process’ in which a pre-image of the range of possible outcomes exists and vari-
ous possible futures become rival hypotheses to be tested (Perrings, 1987, p113).
But true uncertainty assumes an incomplete set of images of the future and denies
that the future is knowable from the past.
The Swedish work on ‘surprising futures’ (Svedin and Aniansson, 1987) and
the IIASA project to develop ‘surprise-rich scenarios’ (Toth, 1995) are both
attempts to address these concerns and to face head-on the major shortcomings of
projections. Schneider and Turner (1995) hosted an Aspen seminar on surprises in
global environmental change in 1994, aimed at identifying a rich panoply of types
of surprise, an endeavour to which Kates and Clark (1996) have also contributed.
Despite these helpful advances, the primary concern is still with population, eco-
nomic growth, technology, energy use and agricultural production. Although
much more detailed and sophisticated than earlier modelling efforts, and vigor-
ously developed during the 1990s, the current second generation of models and
their scenarios do not include social, cultural and institutional developments and
they lack the requisite variety needed to capture possible worlds reflecting the
dynamic interplay and different social and cultural groups (Thompson and Rayner,
1998). The models and scenarios still rely heavily on conventional socioeconomic
indicators, such as population, energy production, SO 2 and CO 2 emissions, labour
force, production capacity, industrial emissions, agricultural production, forest
production, water demand – plus a long list of environmental indicators, although
more social attributes are being worked into the impacts. Future histories, as in
Beyond Hunger in Africa (Achebe et al, 1990), are richer and explore a wider range
of concerns, including cultural ones. Meanwhile, the process for developing inte-
grated assessments also remains disputed; none of the second generation of models
is being developed in close conjunction with the decision makers and others who
will use them. As a result, they remain politically simplistic and naive and lack the
flexibility needed to represent the interests of various interests and social groups
over space and time (Rotmans and Dowlatabadi, 1998, p341).
Several promising developments in the current global modelling efforts deserve
mention. The 2050 Project undertaken by the World Resources Institute during
the 1990s intentionally sought to broaden the social visions involved in alternative
future worlds. So, in 1993, project members solicited nominations from a range of
colleagues and the World Council of Indigenous Peoples and other NGO leaders
to prepare essays envisioning positive futures for their regions or locales. Some 52
‘envisionaries’ from 34 countries eventually submitted essays in five languages,
with developing countries substantially over-represented. The results provide a
highly suggestive tapestry of the potential richness of global visions (Nagpal and
Foltz, 1995). Hammond (1998) also subsequently examined three scenarios, or
worldviews, of the future which he entitled the market world – a world of rapid
economic growth and technological innovation; a fortress world, in which future
market failures create a future in which enclaves of wealth and prosperity coexist
with widening misery, desperation and conflict; and a transformed world, where