Negation is a unique and fascinating property of human language which has been given extensive theoretical and typological treatment. One categorization divides negation use into metalinguistic negation and descriptive negation (Horn, 1985). Descriptive negation (DN) is the truth-functional semantic operator which has received the most attention in the literature. Metalinguistic negation (MN) is a device for registering objection to some aspect of a previous utterance, see (1) for examples of grounds for objection and abbreviations.;(1) A. Descriptive Negation: Chris doesn't have 5 sisters; she has 2 sisters. B. MN Objection to Morphological Realization (OTM): Chris doesn't have 3 childs; he has 3 children. C. MN Objection to Register (OTR): It's not 'later dude'; It's 'goodbye grandma.' D. MN Objection to (Scalar) Implicature (OSI): Chris doesn't have 2 brothers; he has 4 brothers. E. MN Objection to Presupposition (OTP): The King of the U.S. isn't bald; the U.S. doesn't have a king.;Cross-linguistic studies of MN suggest that it is ubiquitous in human language. However, there is a dearth of research on the acquisition of MN. This study fills a gap in our empirical knowledge of language acquisition by charting the development of MN comprehension. To that end, a non-experimental study consisting of three tasks was developed and conducted with two different groups of participants.;This research specifically seeks to answer the following questions: At what stage in linguistic development is MN comprehension acquired? Is the course of development language specific? And, is MN a unitary phenomenon or dependent upon the objection type and associated processing requirements?;Each child completed a theory of mind (ToM) task probing false belief understanding, a sentence repetition (SR) task ensuring sufficient phonological memory and grammatical representation capabilities, and an MN task. The MN task consisted of a training section and four conditions: OTM, OTP, OSI, and OTR (see 1).;This study builds on and adds to existing acquisition findings. For example, children acquire irregular plural morphology by age 6 cross-linguistically (Slobin 1985). Studies on scalar implicature (SI; Horn, 1972) show that adult-like comprehension isn't acquired until 7-9 years (c.f., Chierchia et al., 2001; Katsos, 2008; Huang & Snedeker, 2009). Drozd (2002) suggests that children actually use MN first and acquire DN later (around age 3 (Slobin 1985)). So, MN comprehension might follow a pattern similar to acquisition of the sources of objection. However, MN may also be a unitary phenomenon acquired separately from any of the potential sources of objection.;The first group of participants included 43 monolingual English speakers (aged 3-14, mean 8.26, sd 3.45, 51.2% female, 48.8% male). As expected, significant positive association was found between age and performance on all tasks and conditions, (ToM task tau (43) = .558, SR tau (43) = .661, OTM tau (43) = .715, OTP tau (43) = .754, OSI tau (43) = .799, OTR tau (43) = .727; for all p<.001). A series of linear regressions showed statistically significant results between age and all tasks and conditions. Age contributed the smallest percentage of explained variance in ToM task results (F(1,41) = 54.333, p<.001, Adjusted R2 = .487) and the largest in OSI (F(1,41) = 136.974, p<.001, Adjusted R2 = .764). Age explained between 69% and 76% of the variance for the four MN conditions. The results show that MN is a unitary phenomenon, acquired around age seven.;The second group of participants included ten bilingual Iraqi-Arabic and English speakers (aged 5-11, mean 7.6, sd 2.12). These participants completed an adapted version of the study in their native language. The results show positive correlations between age and performance across conditions (OTM tau (10) = .501, OTP tau (10) = .535*, OSI tau (10) = .282, OTR tau (10) = .645*; for all p<.05). All conditions showed a marked comprehension increase between 5 and 8 years. This finding suggests that MN acquisition follows the same pattern cross-linguistically. However, the small sample size makes this data suggestive due to individual difference. Interestingly, bilingual participants had more adult-like responses at a slightly earlier age than reported for monolingual English speakers. These findings support future cross-linguistic, multilingual, and special population studies on MN, pragmatic processing, negation, metalinguistic awareness and lexical structure.;With the advent of clinical pragmatics (e.g., Cummings 2007) and the rise of experimental pragmatics (Noveck & Reboul, 2008), it is important to have a linguistically informed empirical basis for technicians in other fields to draw upon. This research provides a baseline for comparison and a foundation for future researchers and clinicians.
展开▼
