The Instruction of Imagination

Dor's theory of language is rooted in the assumption that the experiences of animals (including humans) are significantly different from one another—that they are essentially private. The differences between private experiences generate what Dor calls experiential gaps. He defines intentional communication as an attempt to overcome that gap.

Dor argues that apart from language, all intentional communication systems are experiential: the communicator translates her communicative intent into a perceptible output which the interlocutor experiences directly. In other words, these systems rely on shared experience, and experiential communication remains limited to the here-and-now.

Language uses a different strategy: the communicator provides the receiver with a coded set of instructions for imagination, which the receiver is meant to use as a scaffold for imagining the communicator's experience. The receiver, by analyzing the code and recombining his own past experiences, creates a new, imagined experience. Thus, speakers can share experiences outside the here-and-now - things that happened, that could conceivably happen, inner thoughts and emotions, and outright lies. It is this unique strategy, Dor argues, which sets language apart, and makes it an extremely powerful communication technology.

Language is constructed through an iterative social process called experiential mutual identification. At its core, this communicative process can be used to reduce experiential gaps. If, for example, A perceives an object in his environment which he thinks B should notice as well, he can point to that object until B notices - now that are both aware of it, and the experiential gap is smaller.

The same process can be used with the tentative use of symbols. An elementary example of mutual-identification for language is when a child points at something, wishing for his parents to notice it, and calling it by a name. The child is in fact attempting to align his experience with that of his parents, and to verify that they are translating that experience into language in the same way. Humans engage in acts of mutual-identification throughout their lives, constantly testing and affirming the way in which their experiences are translated into language.

Dor defines the symbolic landscape as a semantic web formed out of past language events, modelling in its inner dynamics the world of private experience, only in a more simplified, normative way. Every lexical item in the symbolic landscape is a discrete instructor of the imagination, and is semantically connected to other lexical items in the landscape.

Private experience, Dor argues, is never fully aligned with the symbolic landscape, but instead the two are constantly affecting one another. Although private experience is to some extent primed towards linguistic expression, some aspects of it inevitably remain inexpressible. This inherent discrepancy between the two means that the work of mutual identification for language is never finished, and partially explains humans’ compulsive need to share their experiences.

Dor builds on Merlin Donald’s theory of mimetic communication, an hypothesized precursor of language which included manual, facial, vocal and bodily gestures that were both intentional and representational, but were not compositional or arbitrary. This form of communication, Dor argues, evolved along with a growing codependency of humans, who became increasingly dependent on social learning and cooperative activities (tool-making, hunting and foraging, as well as alloparenting). Critically, mimetic communication and increased prosociality would have allowed for more frequent and elaborate mutual identification. Dor suggests that these developments would eventually initiate an evolutionary spiral: new cooperative behaviors would continuously require upgrades to the communication technology, these upgrades would enable new cooperative behaviors and increase codependence, which would require further communicative upgrades, and so on. This background is hypothesized to have provided the scaffold for a new communication technology, based on an instructive rather than an experiential strategy, and allowing humans to communicate about things that are beyond their shared experiential environment.

Dor, along with Eva Jablonka, and based on evolutionary developmental biology, developed a theory of language evolution in which cultural learning and invention are hypothesized to have preceded and to have guided changes in the genome and epigenome. As described by West-Eberhard, the adaptive evolution of novelties can be initiated with a plastic, phenotypic response by the organism to a new environmental input. If the environmental input is recurring, a subpopulation capable of responding to it can be selected - that is, adaptive variation in the regulation, form, or side effects of the plastic trait is selected, in a process otherwise known as genetic accommodation. As West-Eberhard writes, this ‘phenotype first’ mode of evolution differs from the standard Neo-Darwinian approach in that genes are followers, not leaders, in evolution. In the case of language evolution, Dor and Jablonka suggest it might have originated with collective explorations of a new communicative strategy (an instructive, imaginative one). Its gradual proliferation in human societies would have created a selective environment in which the plastic traits allowing language learning would evolve. Dor suggests a process in which humans were constantly trying to catch up with their new communication technology, and argues that the innate capacities humans currently possess are derivative, emergent, variable and partial. Dor sums up the argument thus: “First we invented language. Then language changed us.”