On the acquisition of scalar and additive inferences: Evidence from Spanish incluso ‘even’ and ni siquiera ‘not even’

Studies on the acquisition of English even have focused exclusively on comprehension of the scalar inference it induces, leaving aside the additive inference it also induces. This study fills this gap by exploring the acquisition of both the scalar and additive inferences, and extends the investigation of scalar focus particles to Spanish. To our knowledge, this is the first experimental study to investigate comprehension of both the scalar and additive inferences triggered by incluso/even and ni siquiera/not even in tandem. We show that the acquisition of these inferences increases with age and that there is a systematic relation between the two inferences. (i) Children who showed target-like behavior on the scalar inference also behaved target-like on the additive inference, but not conversely. (ii) Children who behaved non-target-like on the scalar inference, by volunteering so-called ‘opposite’ answers (Kim 2011, Gowda et al. 2020), do not behave randomly on the additive inference: children who systematically reversed their response on the scalar inference, consistently reversed their response on the additive inference. These findings support Rullmann’s (1997) claim that the additive inference does not have an independent status but arises indirectly as a pragmatic entailment due to the interaction of the scalar inference with the assertion. We conclude that children’s target as well as non-target answer patterns for both inferences combined show they have scalar knowledge: they know there is a scale, its directionality, correctly draw least/most likely inferences and on the basis of these inferences correctly infer the others’ likelihood to have succeeded.