A Human-Participatory Social Turing Machine without External Oracles

This work introduces a theoretical computation model termed the Human-Participatory Social Turing Machine (HSTM). Unlike classical Turing machines or blockchain consensus mechanisms, HSTM does not rely on any external oracle or objective ground truth. Instead, computation proceeds through anonymous, parallel, and blind submissions by participants, where majority convergence determines the system state transition. The model prioritizes procedural fairness, openness of participation, and resistance to centralized control, while explicitly not guaranteeing objective correctness of outcomes. We show that, under these assumptions, HSTM can simulate a classical Turing machine and is therefore Turing-complete in a formal sense. This paper is intended as a theoretical exploration of social computation models rather than a practical protocol specification.