I recommend this book as a must-read for anyone who has thought about whether computing is a science and how it relates to the more mature sciences. Computing and computer science as a discipline are rather young. Alan Turing’s theoretical work and Konrad Zuse’s original computer were not created that long ago. Some may say that computing is a form of engineering. Others may not even go that far, but claim it is merely a tool to be used to solve problems elsewhere. The author of this book makes the case for computing as the fourth great scientific domain, along with physical science, life science, and social science. This is because computing has the characteristics of a great scientific domain, which, according to the author, is distinguished by “understanding and shaping of the interactions among a coherent, distinctive, and extensive body of structures and processes.”
The author takes a strange path to make his case. It is important that readers suspend their personal feelings when reading this book. For example, the first chapter on the computing sciences is primarily descriptive, and it claims that there are three great scientific domains. Some readers may disagree with the disciplines included, or excluded, in these domains and may challenge the inclusion of social sciences as a genuine scientific domain. I know I have my doubts, but I decided to let him make his case, that computing is a great scientific domain. He proceeds to show how this is so.
The second chapter describes a relational approach to the interactions and overlaps among the great domains, computing included. He calls his approach relational because it is built on domains and relationships; he invented a symbolic language, metascience expression (ME) language, to describe these relationships. ME language is a shorthand for succinctly describing the essence of the domains and relationships present in a real example. There are ambiguities, and different expressions can be written for the same phenomenon or example depending on how one thinks about it. The symbols emphasize two main classes of relationships: implementation and interaction. Initially, it seemed to me that the ME language was superfluous. As I continued reading, I learned how well the language captured the relationships in the examples and illuminated them.
Chapters 3 and 4 are long because they are full of examples. They cover implementation and interaction, respectively. Computing interacts with the other three great domains via these two relationships. Computing can be implemented by devices or agents in the other three great domains, and in a reciprocal manner, simulations in these domains can be implemented by computers. Likewise, interactions can be uni-directional (in either direction) or bi-directional between computers and agents or devices in the other domains. In all of these cases, the implementation or interaction is discussed thoroughly and the ME language shorthand representation of the relationship (as well as alternatives to account for ambiguity) is given and discussed.
The fifth chapter, on relational macrostructures, elaborates on the foundation in chapters 3 and 4 by discussing four large application areas: mixed worlds (the real world and virtual worlds together), computing in the sciences, research institutes (emphasizing the author’s role in the institutes at the University of Southern California), and academic computing.
In chapter 6, he returns to his original thesis that computing is the fourth great scientific domain. This chapter is the heart of the book. He reviews the earlier chapters to support his claim. He also tackles the issue of where the humanities (literature, art, and so on) may fit in (he suggests as part of the social sciences), whether mathematics is a science (he argues that it is part of computing), and whether philosophy and theology may be part of any scientific domain (he is not sure that they can be). The last chapter is a brief summary of the book with concluding remarks.
I was struck by Rosenbloom’s forthrightness in reporting his own journey in developing his ideas, including failures and frustrations along the way. He owns up to ambiguities when he finds them. He is candid in observing that some readers may find his arguments circular. Overall, this book is a splendid contribution to the philosophy of science.