Magnitude of negative priming varies with conceptual task difficulty: attentional resources are involved in episodic retrieval processes.

In a variant of the negative priming (NP) procedure, the larger of two presented animals is to be named in each trial. Eight animals of different sizes are used, which allows a manipulation of conceptual task difficulty in terms of pair distance (difficult: one step, versus easy: three steps) on the series. Distances are varied for prime pairs and probe pairs orthogonally. NP effects were found for easy (wide) probe distances (Experiments 1 and 2) and, additionally, for easy (wide) prime distances (Experiment 2). This pattern is interpreted in terms of different theories of NP, which emphasize either forward-acting (prime to probe) or backward-acting (probe to prime) processes. The present results are most compatible with a backward-acting mechanism defined by the episodic retrieval perspective; they are less compatible with a forward-acting inhibition perspective. The results have implications for resource requirements of retrieval-based accounts of NP.