This paper develops a speaker-centered view of language as an alternative to the monolithic decontextualized abstractions favored by modern linguistics, and suggests the application of the speaker-centered view to indigenous language renewal. The paper contends that the modern notion of languages as homogeneous stable "things" that are taught, learned, and used--a concept deeply embedded in linguistic theory--is fatal to the goal of revitalizing indigenous languages. Language, like an ecosystem, is by its very nature dynamic and ever-changing. In contrast, the speaker-centered perspective asserts that people and their actions are inseparable, recognizes that language is an integral part of people's interactions with each other and their environment, and distinguishes between primary and secondary discourse. Primary discourse is dialogic and immediate, has rules that are tightly integrated with nonverbal experiential knowledge, and is not standardized. Secondary discourse aspires to authority and permanency and need not bear any relationship to things that people actually do and know. Seen from this perspective, indigenous languages and the standardized dominant languages that threaten to engulf them are not as similar as linguists would have us believe. The point of language renewal is not to make indigenous languages resemble standardized ones, but rather to restore the balance between primary and secondary discourse, and with it the balance between people and nature that indigenous communities had once perfected. Successful language renewal requires the interweaving of critical literacy in the dominant language with local knowledge and living relationships expressed through the local indigenous language. Contains 43 references. (Author/SV)