Declarative and Procedural Paradigms - Do they Really Compete? (Panel)

Some may be tempted to rephrase the title: are the paradigms still competing or competing again? have they stopped competing? And in the latter case we may add: are they beginning to cooperate? The former procedural/declarative controversy [Win75] has reappeared under various guises, e.g., in discussions about object-oriented/logic programming or knowledge representation. Both research paradigms have developed into large user communities, while the 'logicism ~ debate following Drew McDermott's "A critique of pure reason" [Lev87] has not yet influenced developers of knowledge-based systems. The present workshop is concerned more with technical issues, including the compilation of declarative knowledge into procedural forms. But now that efficient knowledge-base compilers are being developed, the declarativist position may become relevant for practice ("declarative knowledge can work") and perhaps, reconciled with the proceduralist position ("declarative and procedural knowledge can be combined"). Hopefully the panel had something to say to AI developers puzzled both by epistemological sophistries and expert-system shells.