Age-of-acquisition effects in picture naming: Are they structural and/or semantic in nature?

Abstract

Pictures having early acquired names are named faster than pictures having late acquired names. Age-of-acquisition (AoA) effects in picture naming are generally ascribed to lexical phonological representations, but alternative hypotheses state that they are located at the levels of semantic and/or object recognition. In Experiment 1, a semantic locus of AoA effects was tested. Participants performed both a picture naming task and a name–object verification task on the same items in two different sessions. AoA effects were reliable in picture naming latencies but not in name–object verification times. In Experiment 2, an object recognition task was used with the same items as employed in Experiment 1. Late acquired items were responded to faster than early acquired items. The findings do not support a semantic or a structural locus of AoA effects in picture naming.

Acknowledgments

The authors wish to thank Chris Barry, Bob Johnston, Catriona Morrison, and an anonymous reviewer for helpful comments and Delphine Martin for running Experiment 1.

Notes

¹In the present paper, we chose the term of AoA to remain compatible with the AoA literature. However, we and others have argued elsewhere that AoA effects are better described and operationalized as frequency trajectory effects (Bonin, Barry, Méot, & Chalard, 2004; Zevin & Seidenberg, 2002, 2004). Frequency trajectory refers to changes in frequency over ages and is naturally correlated with AoA (Zevin & Seidenberg, 2004). We have shown that frequency trajectory effects are reliable in picture naming but not in word reading or spelling-to-dictation (Bonin et al., 2004). It should be noted that, for the purposes of the present paper, the distinction between frequency trajectory and AoA effects is not crucial.

²Imageability, entered as a covariate in the by-item analysis, was significant when error rates were taken as the dependent variable. Nevertheless, this variable did not interact with any of the other independent variables, i.e., AoA, task, order.

³Imageability, entered as a covariate in the by-item analysis, was significant when spoken latencies were taken as the dependent variable. Nevertheless, this variable did not interact with the independent variables (i.e., AoA, task, order).

⁴Imageability, introduced as a covariate in the by-item analyses, was not significant in the error rates analysis as in the latencies analysis.