Although high-temperature material response is known to be history-dependent, many models of creep crack growth assume the history-independent Norton constitutive law. Even so, these models capture the experimentally observed creep crack growth by adjusting only the damage model. This is explained presently by showing that the damage evolution ahead of a stationary crack in a material obeying a history-dependent unified creep-plasticity constitutive law due to Robinson can be 'fit' by simply adjusting the damage parameters in a model implementing Norton's law. The implication of this result to the case of propagating cracks is discussed. © 2015 Elsevier Ltd.