Diane Proudfoot

This paper explores the relevance of Wittgenstein’s philosophi- cal psychology for the two major contemporary approaches to the relation between language and cognition. As Pinker describes it, on the ‘Standard Social Science Model’ language is ‘an insidious shaper of thought’. According to Pinker’s own widely–shared alternative view, ‘Language is the magnificent faculty that we use to get thoughts from one head to another’. I investigate Wittgenstein’s powerful challenges to the hypothe- sis that language is a device for communicating independently constituted (or individuated) thoughts. I argue that Wittgenstein offers instead a subtle version of the thesis that language determines thought.

Anthropomorphism is a phenomenon that describes the human tendency to see human-like shapes in the environment. It has considerable consequences for people’s choices and beliefs. With the increased presence of robots, it is important to investigate the optimal design for this tech- nology. In this paper we discuss the potential benefits and challenges of building anthropomorphic robots, from both a philosophical perspective and from the viewpoint of empir- ical research in the fields of human–robot interaction and social psychology. We believe that this broad investigation of anthropomorphism will not only help us to understand the phenomenon better, but can also indicate solutions for facil- itating the integration of human-like machines in the real world.

The imitation game, the recent biopic about Alan Turing's efforts to decipher Nazi naval codes, was showered with award nominations. It even won the 2015 Academy Award for Best Adapted Screenplay. One thing it won't win any awards for, though, is its portrayal of the "imitation game" itself-Turing's proposed test of machine thinking, which hinges on whether a computer can convincingly imitate a person. The Turing test, as it is now called, doesn't really feature in the file. (Given that the movie gets so much of the history wrong, perhaps that's a good thing).

In 1948 Turing claimed that the concept of intelligence is an “emotional concept”. An emotional concept is a response-dependent concept and Turing’s remarks in his 1948 and 1952 papers suggest a response-dependence approach to the concept of intelligence. On this view, whether or not an object is intelligent is determined, as Turing said, “as much by our own state of mind and training as by the properties of the object”. His discussion of free will suggests a similar approach. Turing said, for example, that if a machine’s program “results in its doing something interesting which we had not anticipated I should be inclined to say that the machine had originated something”. This points to a new form of free will compatibilism, which I call response-dependence compatibilism and explore here.

Turing was probably the first person to advocate the pursuit of robotics as a route to Artificial Intelligence and Wittgenstein the first to argue that, without the appropriate history, no machine could be intelligent. Wittgenstein anticipated much recent theorizing about the mind, including aspects of connectionist theo- ries of mind and the situated cognition approach in AI. Turing and Wittgenstein had a wary respect for each other and there is significant overlap in their work, in both the philosophy of mathematics and the philosophy of AI. Both took (what would now be called) an externalist stance with respect to machine intelligence. But whereas Turing was concerned only with behaviour, Wittgenstein emphasized in addition history and environment. I show that Wittgenstein's externalist analysis of psychological capacities entails that most, even all, future "artificially intelligent" computers and robots will not use language, possess concepts, or reason. The ar- gument tells, not against AI, but only against AI's traditional and romantic goal of building an artificial "res cogitans" - as first embraced by Turing and now ex- emplified in the work of Brooks and others on cognitive robotics. This argument supports the stance of the growing number of AI researchers whose aim is to pro- duce, not thinking and understanding machines, but high-performance "advanced information processing systems."

Traditional accounts hold that reference consists in a relation between the mind and an object; the relation is effected by a mental act and mediated by internal mental contents (internal representations). Contemporary theories as diverse as Fodor’s [Fodor, J.A., 1987. Psychosemantics: The Problem of Meaning in the Philosophy of Mind. MIT Press, Cambridge, MA] language of thought hypothesis, Dretske’s [Dretske, F., 1988. Explaining Behaviour: Reasons in a World of Causes. MIT Press, Cambridge, MA] informational semantics and Millikan’s [Millikan, R.G., 1984. Language, Thought and Other Biological Categories: New Foundations for Realism. MIT Press, Cambridge, MA] teleosemantics share this act– content–object picture (which was also held by several early modern philosophers, in particular Locke). The core of the traditional view is the thesis that reference and intentionality are relational (‘thesis RR’). Although deeply problematic, RR is entrenched also in psychology, linguistics, cognitive science and Artificial Intelligence. Using for the most part arguments employed by Wittgenstein, we mount a case against RR and advance a deflationary account of reference and intentionality according to which neither is relational.

This is a detective story. The starting-point is a philosophical discussion in 1949, where Alan Turing mentioned a machine whose program, he said, would in practice be “impossible to find.” Turing used his unbreakable machine example to defeat an argument against the possibility of artificial intelligence. Yet he gave few clues as to how the program worked. What was its structure such that it could defy analysis for (he said) “a thousand years”? Our suggestion is that the program simulated a type of cipher device, and was perhaps connected to Turing’s postwar work for GCHQ (the UK equivalent of the NSA). We also investigate the machine’s implications for current brain simulation projects.