The paper argues that Deacon's co-evolutionary theory provides a basis for changing how we think about language and brains. Instead of ascribing language to either nature or nurture, it is seen as intrinsic to both: biological principles ensure the brain can only function by attuning to its body's worlds. For humans, this means not only that our brains are biosocial organs permeated by history but also that our infants' bodies tightly constrain the nature of our languages.While endorsing the thought that language is insinuated into brains, I also identify what I believe is the theory's Achilles heel. Deacon pictures brains as able to process words qua symbolic 'tokens'. Unlike morphosyntactic patterns, these belong to a private domain where referential interpretation detaches from experience. Opposing this split between symbolic and nonsymbolic aspects of language, I claim it is both unnecessary, implausible, and damaging to co-evolutionary theory.