In this dissertation, I use a bilingual epistolary corpus to examine the interaction of language contact, language- and genre-specific conventions, and speakers' individual communicative strategies. Although multilingualism was a common historical condition, many traditional language histories and studies of historical speech only consider monolinguals, or bilingualism as it affects a particular language, rather than seeing multilingualism as part of an individual and community repertoire. In contrast, this study uses both quantitative and qualitative methods to examine the linguistic practices of a bilingual network of speakers in both of their languages. The corpus analyzed here is a collection of private letters which I selected and transcribed from the family correspondence of Francisco Bouligny, a soldier and military governor in colonial New Orleans. These letters were written in both French and Spanish as the family and acquaintances corresponded between New Orleans, France, and Spain for a period spanning from the mid-18th to the mid-19th century. I address the following research questions through case studies at three levels: phonological/orthographic, morphosyntactic, and pragmatic. (1) In what way(s) does bilingualism affect language usage? How do bilingual and monolingual usage differ? (2) How does bilingualism interact with other factors such as convention, genre, audience, and stance, and how do bilingual writers use the resources of their two languages in their production? In Chapter 3, I find widespread orthographic variation conditioned by factors such as age and geographic location. However, I argue for the inclusion of language-specific education and literacy as an additional factor in variation, as I find that the written standard can obscure variation and contact effects while also serving as a resource for bilinguals when writing in a second language. Chapter 4 addresses the contrast between the complex and simple past tenses in both Spanish and French. As the monolingual patterns of use in each language diverge, I find that bilinguals as a group do not follow monolinguals in showing an increased use of the complex past in French but not Spanish. Although individual patterns vary greatly, bilinguals use the French complex past overall less frequently than monolinguals, arguably because of the restraining influence of Spanish contact. This contact influence can be seen in discourse-pragmatic uses of the two tenses, used to shape the narrative or create temporal contrast. In Chapter 5, I consider the construction of identity and interpersonal relationships in bilingual correspondence through choice of language, forms of address, and expressions of sincerity. I find that speakers choose and continue to use only one language with their addressees, even when both speakers are bilingual, and that this choice is motivated largely by the characteristics of the addressee. The choice of second person pronouns (T or V) patterns fairly rigidly according to language and familial relationship, but I argue that speakers vary opening formulae to express more subtle distinctions in the intimacy of the relationship. Some speakers similarly appeal to existing formulae to convince the addressee of their sincerity, but this can be shown to be particularly true of less literate or second-language speakers, while other speakers eschew explicit mentions of sincerity for other strategies. Overall, although speakers show awarenes of epistolary norms in each language and in many ways adhere to language-specific practices, they also manipulate these conventions in meaningful ways. This study is the first to delve into issues of historical bilingualism through a balanced bilingual epistolary corpus and one of few to explore language use in Spanish Louisiana. I find that individual speakers show evidence of using resources from their bilingual repertoire to aid in composition and construction of meaning in various ways. This is particularly true of loci of variation in the individual language systems or structures that share typological similarities across the systems. However, at the community level the variety of individual patterns and the force of the monolingual norms and written formulae appear to inhibit the spread of change. I underscore the importance in future studies of considering a speaker's (and community's) language use in its entire context, including other languages, literacies, and considerations of genre, as we explore how speakers use the available resources for interpersonal communication and how that translates to the community level.