| Popis: |
According to Richard Levins, there is no single, best all-purpose model in ecology, nor can useful models simultaneously maximize theoretical virtues of generality, realism and precision (Levins, 1966, 1968a,b). A 'virtue trade-off' must be made if models are to be useful for understanding, explanation, prediction or control. Recently, large-scale data synthesis projects have aimed to systematically review many empirical studies to evaluate major hypotheses. A specific version of systematic review called the 'hierarchy-of-hypotheses' approach has been proposed (Jeschke et al., 2012; Heger and Jeschke, 2014; and Chapter 2, this volume). I argue that investigation of the patchwork of models organized by Levins' theoretical 'virtue space' and discovery of robust theorems across a theoretical landscape could be enhanced by recent efforts to manage evidential complexity in a patchwork of empirical studies by the hierarchy-of-hypotheses approach. |
