For more information on the Cognition and Affect Project see
Some related video recordings are available here:
I gave these two lectures on 23 and 24 Feb 2016, introducing ideas about the importance of internal symbolic structures, in the minds of intelligent animals or intelligent machines.
The first lecture focused mainly on language, grammars, parsing, and semantic interpretations of linguistic structures.
The second lecture generalised this requirement for internal symbolic structures to visual perception, showing how perceiving 3-D scenes required internal construction of descriptions of contents of the scenes and the affordances (possibilities for change and constraints on change) provided by those scenes.
This led to a conjecture that in intelligent animals internal languages evolved for various non-communicative, purposes before external languages. The internal uses included representing information acquired through perception, and also intentions, plans, questions, explanations, and theories about the environment. Such internal languages (in simpler forms) must also be used by other species capable of interacting in creative, intelligent, ways with things in their environments, including other animals.
A natural conjecture based on these ideas is that requirements for collaborative actions (and other requirements for interacting with conspecifics) might have led gradually to the evolution of gesture-based languages for expressing intentions, questions, instructions, explanations, warnings, shared goals, etc.
After sign languages had reached a significant level of sophistication, they may
later have been supplemented by sound-based, vocal, communication, with the
advantage of communicating in the dark, around corners, or while hands are busy
with complex tasks. Because of this wider applicability evolution of sound-based
languages (including physical and computational mechanisms of production and
comprehension) seems to have taken over, and become dominant in most human
cultures. But the previously evolved potential to develop equally rich and
expressive sign languages remains in the brains of normal new-born humans, as
shown by the use of such languages in deaf communities.
The same ideas were presented in a different way in this online PDF presentation to a similar group of students in 2015: