This paper introduces a Linguistic Dynamic Multi-Criteria Decision Making (LDMCDM) framework for evolving decision scenarios with time-varying alternatives, criteria, and experts. The proposed solution combines three key innovations: (1) a feedback-enabled architecture for temporal adaptation, (2) the novel 2T \(\prod \) associative aggregation operator that ensures computational efficiency through its reinforcement properties, and (3) 2-tuple linguistic modeling that maintains human-interpretable outputs. The 2T \(\prod \) operator eliminates full-history storage requirements while dynamically merging linguistic evaluations across periods. Validated through a subcontractor evaluation case study, our approach demonstrates consistent performance in handling dynamic system changes. The framework bridges a fundamental gap in LDMCDM by simultaneously addressing linguistic uncertainty and temporal data evolution, offering practical advantages in both computational performance and decision quality.