The type of sociolinguistic variation described in Chapter 19 is sometimes attributed to cultural differences. It is not unusual to find aspects of language identified as characteristic features of African American culture or European culture or Japanese culture or multicultural communities. This approach to the study of language originates in the work of anthropologists who have used language as a source of information in the general study of “culture.”

This is a barely recognizable version of the Lord’s Prayer from over a thousand years ago. A later version is included for comparison at the end of the chapter, on page 289. The different versions provide a rather clear indication that the language of the “Englisc” has gone through substantial changes to become the English we use today. Investigating the features of older languages, and the ways in which they developed into modern languages, involves us in the study of language history and change, also known as philology.

In the preceding chapter, we focused on variation in language use found in different geographical areas. However, not everyone in a single geographical area speaks in the same way in every situation. As Captain Marryat learned, in the quotation above (cited in Mohr, 2013: 192), some individuals can have very specific views on socially appropriate language. We are also aware of the fact that people who live in the same region, but who differ in terms of education and economic status, often speak in quite different ways.

The chapter considers the overall organization of the mind. In the first section, we look at agent architectures in artificial intelligence, looking at three different architectures for intelligent agents in artificial intelligence to see what distinguishes cognitive agents from simple reflex agents. The second section looks at Fodor's modularity thesis for cognitive architecture. Fodor proposed domain-specific modules for solving particular cognitive problems, while central processing controls higher-level mental operations. In contrast, the massive modularity hypothesis argues that there is no central processing. This hypothesis is based on evolutionary arguments, which we look at in the third section. In the last section, we look at an example of a hybrid architecture developed by John Anderson and his team - ACT-R, which combines the physical symbol system and the connectionist neural networks.

This chapter introduces challenges posed by the study of consciousness in contemporary cognitive science. The first section looks at conscious and nonconscious information processing in priming experiments and in brain-damaged patients. Building on these findings, the second section explores the functions of consciousness and potential roles the conscious (visual) information can play in the cognitive system. The third section discusses the argument that functional theories could not explain the subjective experience of consciousness but only tackle the "easy" problem of consciousness. The fourth section reviews two of the best-established theories of consciousness – the higher-order thought theory and the global workspace theory.

So far, this book has focused on cognitive aspects of the mind. This chapter introduces the emerging area of affective science and emotion studies, which has gathered impetus in recent years. The first section reviews early theories of human emotions from Herbert Simon and Paul Ekman. The second section looks at how affective scientists explore emotions from a multidimensional perspective, introducing some new techniques in affective neuroscience. The last section draws the different strands together through the case study of fear, looking in particular at the neurobiology and neuroscience of fear.

This chapter considers how connectionist neural networks offer a contrast to the symbolic view of representation discussed in previous chapters. We start by reviewing the structure of neural networks inspired by neurobiology, comparing a single unit in a neural network to a biological neuron. The second section looks at the simplest form of neural network -- a single-layer neural network using the perceptron convergence rule for learning. The third section introduces multilayer neural networks and the development of the backpropagation algorithm. Next, we look at how the multilayer neural network can be trained, and its biological plausibility. The last section summarizes three critical features of information processing in neural networks, as opposed to physical symbol systems: distributed representations, the lack of a clear distinction between storing and processing information, and the ability to learn.

Semantics is the study of the meaning of words, phrases and sentences. In semantic analysis, there is always an attempt to focus on what the words conventionally mean, rather than on what an individual speaker might think they mean, or want them to mean, on a particular occasion. This approach is concerned with linguistic meaning that is shared by all competent users of the language. Doing semantics is attempting to spell out what it is we all know when we behave as if we share knowledge of the meaning of a word, a phrase, or a sentence in a language.

Throughout this book, we have been talking about languages such as English, Spanish or Swahili as if there was a single variety of each in everyday use. That is, we have largely ignored the fact that every language has a lot of variation, especially in the way it is spoken. If we just look at English, we find widespread variation in the way it is spoken in different countries such as Australia, Britain and the USA. We can also find a range of varieties in different parts of those countries, with Lee Tonouchi’s account of “Trick-O-Treat” in Hawai‘i as just one example. In this chapter, we investigate aspects of language variation based on where that language is used, as a way of doing linguistic geography.