“…Regarding the stress dependence, the Raj-Chyung model predicts n ¼ 1 for the aforementioned two rate-controlling processes, which is clearly inconsistent with numerous studies reporting stress exponents different from unity. In silicon nitride ceramics, for instance, stress exponents different from unity are commonly found in conditions where solution-precipitation is reported to operate (Backhaus-Ricoult et al 1995, Wilkinson 1998, Schneider and Mukherjee 1999, Boling-Risser et al 2000, Xie et al 2000, Yoon et al 2000, Zhan et al 2000a, Mele´ndez-Martı´nez et al 2002. Recently, Wakai (1994) has developed an alternative, and significantly improved, model for solution-precipitation creep: the so-called 'step model'.…”