Language and the Evolution of Cognition

The main purpose of this article is to discuss the kinds of mental representations that are required for language to evolve. Firstly, I distinguish between cued and detached representations. A cued representation stands for something that is present in the current external situation of the representing organism, while a detached representation may stand for objects or events that are neither present in the current situation nor triggered by some recent situation. The inner environment of an agent is defined as the collection of all detached representations of the agent. The fundamental difference between signal and a symbol is that the reference of a symbol is a detached representation, while a signal refers to a cued representation. Icons also refer to detached representations, but unlike symbols, the choice of representation is not arbitrary, since an icon in some aspects resembles the thing it represents. A Gricean analysis shows that human linguistic communication presumes an advanced kind of inner environment in order to represent higher order intentions. Not only must we be able to represent the inner environment of other people (i.e., have a “theory of mind”), but we must also represent the other individual's representation of our inner environments. In relation to the evolution of grammar, I distinguish between three levels of grammaticality in a communication system: Systems with no grammar, compositional systems, and systems with grammatical structure. Combining this tripartition with the distinction between cued and detached representations, one obtains six kinds of communication systems. Most animal signaling systems use cued representations and no grammar, but bee's dances have a compositional grammar, while still exploiting cued representations. The evolutionarily first communication systems using detached representations were one-word languages. Gestural communication using icons preceded vocal language using symbols. Donald's proposal of a mimetic stage is an important step in the evolution of language. Gesturing and primitive speech then developed into a protolanguage, i.e., a communication system with a compositional structure, but without grammatical items. Requirements of fast and efficient communication finally resulted in the development of arbitrary symbols and syntactic rules, at the cost of iconicity, to arrive at a language with a full grammar. 1. CUED AND DETACHED REPRESENTATIONS When we communicate by language, our utterances have meaning. The meaning of what we say is represented in our minds. But how do these representations function? It has turned out to be extremely difficult to teach other animals to communicate linguistically. Is there any difference between their mental representations and ours that can explain why they can't learn a complete language? How could such a difference have evolved? This article addresses these questions. The main purpose is to disclose the kinds of mental representations that are required for language to evolve. The first thing to notice is that there are different kinds of mental representations. In Gärdenfors (to appear b), I distinguish c u e d and d e t a c h e d representations, which will turn out to be crucial for the arguments of the present article. A cued representation stands for something that is present in the current external situation of the representing organism. When, for example, a particular object is categorized as food, the animal will act differently than if the same object had been