Recognising a zebra from its stripes and the stripes from “zebra”: the role of verbal labels in selecting category relevant information

Distinguishing members from non-members of some categories can be accomplished by identifying one or several diagnostic features (e.g. zebra-stripes are diagnostic of zebras). Other categories lack diagnostic features (e.g. dogs). Consequently, distinguishing members from non-members requires attending to many correlated dimensions. Interestingly, children and non-human animals are less adept at using diagnostic features compared to adults – possibly due to adults’ more developed verbal labelling abilities. We examined whether recognition of categories with diagnostic features (“sparse” categories) is (1) linked to better abilities to selectively attend to relevant information and (2) aided by labelling. In Experiments 1–2, we quantify and validate a measure of category sparsity. Experiment 3 demonstrates that sparse categorisation, assessed by an implicit naming task, correlates with performance in the flanker task, a measure of selective attention. Experiment 4 demonstrates up-regulating activity over Wernicke's area via transcranial direct current stimulation – hypothesised to enhance labelling – selectively improves sparse categorisation.