This article explores some major elements in the human capacity to create meaning, or signify. Motivated by a problematic emerging from artificial intelligence and cognitive science research, the article discusses the specific features of natural language that distinguish it from other forms of communication and underlie the difficulties found in designing electronic forms of language processing. In particular, the article proposes a model of human communication based on a quadripartite model composed of language, emotion, embodiment, and community. It discusses some areas of the representational aspects of language, which have been identified by literary approaches to texts, and which any attempt at emulating human intelligence and communication should take into account. The article supports the claim that the human propensity to create meaning lies largely in representational ambiguity, which underlies most forms of symbolism.