Sproat's book traces the history of language technology from writing to digital speech and modern language systems. It describes the social impact of technological developments over five millennia. Ch. 1 introduces two human characteristics: language and the creation and use of complex tools. Ch. 2 presents writing as the first case of a technology specifically designed for representing language, contrasting writing with other symbolic systems. In Ch. 3, the author focuses on how writing works as a technology for encoding language and representing information, particularly at the sound level. He examines the Sumerian, Egyptian, Chinese, and Mayan writing systems. Ch. 4 addresses the methodology of decipherment: the process of learning to read ancient writing systems by mapping unknown symbols to a language that may only partially be known. Two initial questions must be asked about an archeological find that appears to include writing: whether it is genuine, and whether or not the symbols in question are really a script. The author examines how decipherment proceeds and how it must be verified as a scientific hypothesis by testing on additional data that have been held back, as in modern computer speech and language research.

In Ch. 5, the author discusses some common assumptions about literacy and its social impact. He argues that script complexity has almost no correlation with literacy and that illiteracy is a direct result of economic and social conditions, not the difficulty of the writing systems involved. Ch. 6 gives a brief history of the mechanical treatment of language. It begins with a description of early attempts to produce speech and interpret text mechanically, and then discusses the influence of telegraphy, the computer encoding of text, and the Braille system. Ch. 7 starts with an overview of technologies related to two principal components of a dialog system, namely the speech recognition and speech synthesis components. The author argues that, at the language-modeling end, research should achieve a more complete understanding of natural language in order to propose new techniques, rather than just continuing to try to improve the current state of the art. Ch. 8 traces the development of machine translation systems, which require all the natural langue technologies including parsers, semantic analyzers, and morphological analyzers, generators, and alignment systems. It gives an introduction to statistical machine translation and discusses the uses and limitations of this technology.

Finally, Sproat discusses the future of language technology for augmentative and alternative communication devices, stressing areas inadequately explored by computer science. He highlights the importance of investment in developing language technologies, talks about their social implications, and focuses on the need to consider their consequences so that language technologies are not abused. The author's expertise is wide reaching, his style is clear, and his examples are varied and accessible.