Models of word and object processing have been informed by experimental data relating to the effects of important marker variables, particularly word frequency. In recent times, however, the role of word frequency has come to be questioned due to the demonstration of the impact of a correlated variable called "age of acquisition" (AoA), which is an index of when a word was first learned. Early acquired words tend to be of high frequency while late-acquired words tend to be of low frequency. Currently, there is a lively debate in the literature about the operation of AoA and word frequency in word and object processing, and how these effects might be modelled in connectionist architectures. Issues relate to: the relationship between AoA and word frequency; the locus of impact of AoA; and the types of identification task (e.g., recognition, comprehension or naming) that are influenced by the AoA of the target. Originally, AoA was thought to be a factor which operated at a name retrieval stage and which consequently only affected naming responses. However, more recent conceptualizations suggest that AoA effects arise out of the order in which items are learned and consequently may operate on many types of recognition task. Recent demonstrations of AoA effects in non-naming tasks may be seen to favour the latter explanation, but the debate is far from over.