Alternative worlds reasonable worlds? plausible worlds?

While philosophers think in terms of 'possible worlds', it remains unclear what criteria would be good to distinguish such worlds from 'impossible worlds'. In contrast, scientists need to make these judgements as a regular part of their ongoing work, yet such judgements are rarely straightforward. Scientists habitually develop accounts of their worlds in ways which enable them to explore and understand their phenomena by constructing and using models of those phenomena, by developing counterfactual alternatives to them, and even by designing and engineering versions of them. Each of these different modes of work develops alternative accounts of the world, and uses them to explore the implications of those accounts. By analysing a set of modelling and counterfactual cases from economic science, it becomes clear that these 'alternative world' constructions can be judged as plausible or well-reasoned according to the scientists' knowledge (both empirical and theoretical) about the phenomena of their world. There are no simple rules of rejection or acceptance of the alternative world account, but the analysis here suggests ways to expand the domains of judgement on possible world accounts.