Many of the technical advances accelerated by World War II emphasized systems of interacting components. Radar engineers learned to think about the systematic interactions of transmitters, receivers, and targets. Developers of digital computers focused on coordination of processors, memory, and input/output devices. The burgeoning field of control theory occupied itself with feedback loops and system response.
A few visionaries, including Ludwig von Bertalanffy and William Ross Ashby, sought to extend systems thinking beyond the limits of a single technology. They proposed to identify general principles applicable to any system, including natural systems (such as social structures), as well as engineered systems—thus, “general systems theory.” As originally conceived, the movement enjoyed only limited success, perhaps because it became too generalized to yield predictions that were useful in specific problems. Its core themes, however, persist in the notions of cybernetics and complex adaptive systems. These themes center on the tension between the whole and the parts of which it is composed, the need to understand each part in the context of its peers, and the potential of the whole to exhibit behaviors that are greater than the simple sum of the parts.
Skyttner seeks to revive the enterprise under its original title. This book has two parts. The first part reviews the basic ideas of holistic thinking, and surveys a range of specific systems theories. The second part considers applications of these ideas to artificial intelligence and artificial life, organizational theory, decision making, and informatics.
One would expect a book about systems to be itself a system of interacting and inter-defining parts. This volume falls far short of such an ideal. It is more of a scrapbook, a collection of bulleted lists and disconnected notes from various systems theorists, without any attempt to relate them to one another or present an overarching structure. Where the text does try to make a point, it is often strangely inconsistent. A prime example is in its discussion of artificial intelligence and artificial life, addressed sequentially in a single chapter. The discussion of artificial intelligence concludes that the enterprise is hopeless. The author’s despair rests largely on the inability of a computer to assimilate the lifetime of experiences on which cognition draws, and is demonstrated by the challenges of natural language processing. Then, the discussion shifts to artificial life, and, on the basis of the game of life and computer viruses, concludes that humans might as well give up, because electronic creatures are the next level of evolution, and are destined to take over the world. One wonders how machines that have been declared constitutionally incapable of understanding human speech can come to dominate human society.
Even a scrapbook can be useful as a summary and pointer to the work of others. Unfortunately, many works cited in the text are not documented at all in the bibliography, vitiating even this use of the book.
In light of these shortcomings, it is not surprising that the volume has numerous minor flaws. It is marred throughout by typographical errors. Many sentences cannot be parsed as English text. Most readers would be surprised to hear the computer scientist Marvin Minsky characterized, first of all, as a “linguist,” or the journalist Kevin Kelly as a “scientist.” And so on, and so on.
The original general systems theorists glimpsed a vision that they were unable to realize, a vision of general principles applicable to any collection of interacting and interdependent parts in a domain-independent way. Others have explored bits and pieces of their vision, and a volume pulling this work together to offer a general synthesis would be a great contribution to the literature. This volume’s title and table of contents promise to make such a contribution, but the book does not deliver.
