Behavioral experiment

Twelve healthy, right-handed volunteers (6 females; aged 20–24 years, mean ± SD = 22.8 ± 0.8) participated in behavioral experiments. These participants were distinct from the volunteers who participated in the semantic similarities investigation. All had normal or corrected-to-normal visual acuity. In addition to 40 animals, 20 artificial objects were selected for the sematic judgment task. The experiment involved ten separate sessions; each consisted of 30 experimental and 30 filler pairs. For experimental trials, the target words were marine mammals, which were primed by fish (Fish-MM), terrestrial mammals (TM-MM), or birds (Bird-MM). For filler trials, 5 pairs each of Bird-TM, Fish-TM, TM-Fish, Bird-Fish, TM-Bird, and Fish-Bird were made. Furthermore, 20 TM-artificial object, 20 Fish-artificial object, and 20 Bird-artificial object pairs were used for nonliving trials. Each prime and target word was visually presented for 300 ms. The stimulus-onset asynchrony between the prime and target word was 1000 ms. The intertrial interval randomly varied between 1200 and 1500 ms. Before the experiment, participants were instructed to perform a semantic decision task. The participants were told that two names of animals in four categories (mammals, birds, fish including shellfish, and artifacts) were presented sequentially in a trial and to judge whether the target word was living or nonliving as quickly as possible. To direct participants’ attention to the categories used in this study, we provided the stimulus categories in the oral instructions for the task (e.g., the animals were mammals or birds or fish). We used Presentation software (Neurobehavioral Systems Inc. Albany, CA, USA, https://www.neurobs.com/) for stimulus presentation and recording of participants’ responses and reaction times. Task scheme of experiments are displayed in Fig. 1.

Differences in reaction time were assessed by repeated measured one-way ANOVA among the Fish-MM, TM-MM, and Bird-MM conditions. For planned comparisons, t-tests were performed, and alpha levels were controlled by Holm method for multiple comparison.