Project Objectives

IN this research project, we will systematically explore aspects of the elementary gestures involved in human speech articulation and in human tool production and use. Our purpose will be to determine parameters that constrain the evolution of goal-directed action sequences made up of elementary gestures of this kind, and to define markers which can be used to assign values to these parameters in fossil and archaeological material. In this way, we can build a scientific tool-kit to track the emergence of human speech and tool-using capacities in a rigorous manner.

We also intend to re-evaluate the possibility that goal-directed action in the manipulative domain may have evolved with complexity of articulatory manoeuvres in vocal communication, reflecting shared features of neural architecture. Our main hypothesis will be that syntactic language derived evolutionarily from a neural substrate for action recognition and action generation, and that the human dependence on social learning of complex tool-assisted behaviours provided the linking mechanism. Human tool use involves an ‘action grammar’ that is still poorly understood, but that involves hierarchical sub-assemblies, serial order effects, and other properties of open and generative systems more commonly seen as characteristic of human language. Social learning of tool-use by observation, scaffolding, and active teaching presumes the existence of mechanisms of true imitation which entail the capacity to parse the structure and recover the meaning of observed action sequences. I f there is an evolutionary sequence from visual communication of complex gestural sequences to vocal-auditory syntactic language, then this is the most plausible route. This leads us to the hypothesis that the transition to increasingly skill-intensive lifeways during hominin evolution, and in particular the development of increasingly sophisticated tool technologies, was in and of itself a sufficient vehicle for language evolution without the need for a discrete gestural stage.

This is the first-order assumption that informs the HANDTOMOUTH project. Our second-order assumption is that selection for enhanced capability in social learning of complex tool use and in complex vocal communication will have promoted mechanisms for increased gestural precision and for higher-frequency recombination of gestural elements. We therefore propose also to investigate the downstream mechanisms of regulation of output, notably in knapping and in articulatory gestures. Our purpose is specifically to focus on parameters which can potentially be assessed using fossil and archaeological evidence.

We recognize that our project may not contain all the elements required for a new synthesis of the evolutionary relationship between spoken language and tool use in humans, but we believe that all the aspects examined in HANDTOMOUTH address necessary building blocks for such a synthesis, and that in each case our work will identify currently unknown areas of convergence and/or divergence between the two systems.

Our consortium is inter-disciplinary, involving cognitive neuroscience, human movement science, physical anthropology, archaeology, and control systems engineering. Our project is organized into two sub-themes (speech motor control, and action generation and action understanding in tool use), within an overall unifying framework that cuts across all workpackages. They have been designed both to guarantee complementarity and convergence across the whole project, and to ensure a high potential for scientific yield independently within the individual sub-fields which each work package will address.