Constructionist research on L2 learning has focused on the degrees to which skewed frequency (Goldberg, Casenhiser & White, 2007; Casenhiser & Goldberg, 2005; Goldberg, Casenhiser, & Sethuraman 2004) in a person's linguistic environment can facilitate "entrenchment," "schematization," and "contingency learning" (Ellis, Romer, & O'Donnell, 2016; Ellis & Ferreira-Junior, 2009a; Ellis, 2002). Usage-based learner corpus studies by Eskildsen (2009, 2011, 2012, 2014, 2015, 2017), focusing on just one or two L2 learners in an ESL classroom, found evidence for (1) learning in the forms of entrenchment and schematization as evidence of developmental sequences (e.g. Bardovi-Harlig, 2002) within individual grammatical constructions, and (2) the learners' experiences with talk-in-interaction helped to provides some of the exemplars that drive fixed multi-word expressions (MWEs) toward schematic, end-state constructions. Meanwhile, Ellis & Ferreira-Junior (2009a) provide an account of contingency learning among adult immigrants to the UK by comparing their distributions of words across three grammatical constructions in both the learners' speech and the speech of native speakers. This study found similar distributions between native and non-native speakers. Gaps remain for Constructionist/Usage-based research to account for contingency learning in connection with observable experience in an L2 that is distant from English and during early stages. Addressing these gaps, this dissertation study investigates contingency learning under conditions of heavily skewed input in L2 classrooms, i.e. institutional forms of social interaction (Heritage & Clayman, 2010). A learner corpus was created to follow ten beginning learners from the Mainland United States during an intensive Mandarin Chinese language camp in Hawai'i. The learners had minimal or no experience with Chinese learning prior to the start of camp. Instruction was organized around several types of pedagogy: the comprehension-in-interaction oriented "Teaching Proficiency through Reading and Storytelling" (TPRS; Ray & Seely, [1997] 2015; Cahnmann-Taylor & Coda, 2018; Lichtman, 2013), peer-talk-in-interaction oriented "Task-Based Language Teaching" (TBLT; Long, 2015, 1985; Ellis, 2009), Cold Character Reading (CCR; Neubauer, 2018; Waltz, 2015), "Extensive Reading" (ER; Ro, 2017; Jeon & Day, 2016; Nation, 2015; Hitosugi & Day, 2004), and Chinese "scaffolded writing" (Waltz, 2015). Collostructional Analysis (Stefanowitsch, 2013; Stefanowitsch & Gries, 2003) is used to compare frequency distribution, collexeme strength, and contingency (measured with bi-directional Delta P) in five main corpora (capturing language that was heard, said, read, and written) with corresponding test corpora (freely written and spoken stories) across five recording periods. "Concreteness" (one form of "salience"; e.g. Crossley, Kyle, & Salsbury, 2016; Brysbaert, Warriner, & Kuperman, 2014) is considered as a factor that may complicate effects from frequency distribution. Finally, institutional interaction (Heritage & Clayman, 2010) is investigated in regards to how teachers and students use and re-use limited language for "talking their institution into being" (p. 20). Findings reveal how the participants used a single Chinese pattern as a resource to (a) acquire that Chinese pattern, and (b) co-construct institutional practices around story-building. These analyses illustrate how this institution-specific interaction resulted in highly skewed frequency. The collexeme analyses reveal a close match between frequency distribution in classroom experience and the learners' freely written and freely spoken stories in test corpora. These findings highlight an active role for contingency learning during early construction learning and language development, given the environments these particular learners experienced. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]