Tuesday, January 06, 2004

Review of Alex Rosenberg's Philosophy of Science: A Contemporary Introduction

Alex Rosenberg clearly has a good grasp on the philosophy of science. His book Philosophy of Science: A Contemporary Introduction is a good demonstration of that. This book is in the same series of a book I read a bit over a year ago by Lycan.

Unfortunately there is much about the book that needs to be said. I have few quabbles about the substance of the book, but the form leaves much to be desired. Rosenberg’s book suffers from a number of defects, most of them pedagogical and organizational. The book covers many of the standard issues in the philosophy of science: explanation, theory formation, underdetermination, and historicism. It also has a discussion of Bayes’ theorem and interpretations of probability.

Normally the first question in an introductory book is: what is science? This question is dispersed throughout the book. Instead there is a nice discussion the relationship that exists between science and philosophy.

Chapters two and three discuss explanation. These chapters address the main theories of explanation, motivating the discussion using notions of laws, causation, pragmatics, statistics, and teleology. These two chapters need to be completely reworked. It should have been one long organized chapter instead of the two poorly organized one. There was also not sufficient space given to clarifying the various accounts of explanation itself.

A number of times in the book there are topics that are put together because they are connected in a significant way. But the connections are not clear. Few of the concepts of explanation are spelled out clearly enough such that someone coming in to the topic for the first time, even with a background in philosophy could understand it. The book is not self-explanatory. The same happens with Bayes’ theorem and probability. There were also not many theories of probability discussed. Moreover he dismisses some of them way too flippantly.

The Section on theories and their problems is only discussed vis-a-vis their role in explanation. This may be a good strategy but there are other virtues to theories besides explanation.

The relationship between Kuhn, Popper, and Feyerabend is not spelled out as clearly as could be, and Lakatos is not even mentioned.

Finally, the book ends with a discussion of historicism, and the sociological accounts of science. To say the last chapter is polemical is an understatement. Here Rosenberg has a Paul Bunion-sized ax to grind. While I wholeheartedly agree with his position, the opposition is treated with ridicule. Latour and Woolgar are not so much refuted as bullied and ridiculed.

This book is not the place to start if you have had no exposure to the philosophy. I believe there is a companion anthology of papers that are addressed in the book. I assume that the function of the book is to help connect the papers in the anthology with each other. If that is done then perhaps the book is not a total write-off. But the book can not be read alone.


The author is clearly bright, but needs to work on being able to patiently and slowly explain each concept fully. The connections between each issue ought to be clear as well. Definitions need to be made as early as possible, and as precise as can be. The examples need to be more spelled out too.