In this dissertation, I investigate the formal semantics and pragmatics of alternative focus in Kwak'wala, a critically endangered Northern Wakashan language of British Columbia, Canada. I show that several notable phenomena and outstanding mysteries of Kwak'wala grammar involve focus expression, and by making their discourse contexts explicit we can observe how changes in discourse-relevant alternatives correspond to changes in morphosyntactic expression. These observations invite reappraisals of classic claims about Kwak'wala and Wakashan grammar, such as the claims that Kwak'wala lacks a noun/verb/adjective distinction (Boas et al., 1947, p. 280) and also lacks a copula (Boas et al., 1947, p. 205). Instead, I argue that Kwak'wala does indeed have a noun/verb/adjective distinction as well as equative (but not predicative) copulas, and show that these are tied up closely with the expression of focus. I argue, contra Koch's (2008) proposal for Nɬeʔkepmxcín focus, that Kwak'wala focus is not based on alignment to the edges of prosodic phrases, but based on the use of marked predication structures in which speakers choose non-optimal predicates like NPs and DPs over unmarked predicates like VPs. I also examine Kwak'wala additive and exclusive focus operators, and in particular investigate their distinctive association patterns, in which different exclusive operators associate with different types of “focus phrase” (Drubig 1994), while additive operators exhibit free association. I propose a hybrid focus model, a combination of the models in Wold (1996), Roberts (1996/2012) and Krifka (2006), among others, in which Kwak'wala focus operators associate with focus phrases, but derive their specific alternatives indirectly, through constraints on a contextual “question under discussion” variable. Finally, I examine the ubiquitous “discourse” enclitic =ʔm, which I propose expresses a discourse-relevant bipolar (e.g., {P,¬P}) contrast, and thereby distinguishes bipolar from monopolar (e.g., {P}) questions and answers (cf. Krifka 2013). The appearance of =ʔm in all additive and exclusive sentences provides morphological evidence that such sentences respond to complex alternative sets consisting of both constituent-type and polar-type contrasts (Krifka 1998, Rullmann 2003).