This article examines the performance of heritage speakers (HSs) on two types of acceptability judgment tasks (AJTs) as well as their corresponding confidence ratings. Data were collected from 75 native speakers of Spanish who were subsequently divided into three groups: Spanish monolinguals, Spanish-dominant HSs, and English-dominant HSs. An AJT targeting morphosyntax (Task 1) showed that all groups were affected by task stimulus, that is, they were significantly less accurate on the ungrammatical items. A lexically focused AJT (Task 2) revealed the tendency of HSs to accept complex words that are possible yet not attested. Confidence ratings on both tasks were very high overall, and there was a significant relationship between accuracy and confidence. Thus, the data question the assumption made in a number of previous studies that HSs suffer from a lack of certainty in their linguistic intuitions. Finally, between-group effects show that the Spanish-dominant HS group was not different from the monolingual group on Task 1, yet there were significant differences between all three groups on Task 2. By comparing morphosyntactic and lexical phenomena on two tasks administered in the same modality, this research contributes to the discussion of vulnerable domains in heritage language bilingualism.

Despite the growing interest in the phenomenon of learning without intention, the incidental learning of phonological features, especially prosodic features, has received relatively little attention. This paper reports an experiment on incidental learning of lexical stress rules, and investigates whether the resultant knowledge can be unconscious, abstract, and rule based. Participants were incidentally exposed to a lexical stress system where stress location of a word is mainly determined by the final phoneme, syllable type, and syllable weight. Learning was assessed by a pronunciation judgment task. Results indicate that participants were able to transfer their knowledge of stress patterns to novel words whose final phoneme was not previously encountered, suggesting that participants had acquired abstract and potentially rule-based knowledge. The combined use of subjective and objective measures of awareness in the present study provides a strong piece of evidence of the acquisition of implicit knowledge.