You are here

Galileo in Pittsburgh

Clark Glymour
Publisher: 
Harvard University Press
Publication Date: 
2010
Number of Pages: 
150
Format: 
Hardcover
Price: 
27.95
ISBN: 
9780674051034
Category: 
General
[Reviewed by
Tom Schulte
, on
08/16/2010
]

Philosopher Clark Glymour has experience in practical application of research and analysis in the areas of education, climate change, public health, military research, and more. Drawing on this experience, he recalls in several brief essays the problem of misapplied science, the gross errors of military and public bureaucracies, and more. Glymour sees a sausage factory at work in the application of research and prediction science. As a software implementer of mathematical models for an industry that seems largely ignorant of their strengths and weaknesses, I can appreciate Glymour’s evangelism for better science to make for better lives.

Glymour handles his topics in a breezy way, usually in only a few pages, liberally sprinkling in comments that give his discourse the feel of a rant. For instance, Saudi Arabia is a “vile nation” and the source of martinets is “always the south.” Regardless of the accuracy of such generalizations, they do not support Glymour’s arguments. I often wished he had vented off the page and put more of his experience and germane arguments into his vignettes.

If the reader is already of Glymour’s mind on these topics, this slim volume will be an easy read and self-validating. If you need a little convincing, there is a Suggested Reading section with two or more references for each topic.

The title piece to this book is a fascinating overview of the research into environmental lead levels and the effects, if any, on the development of intelligence in children. Glymour tells the story of the embattled Herbert Needleman as he sought to find proof in the data of the evils of lead. The discussion of use and misuse of statistical methods of regression and variable identification can be an eye-opening revelation of what can happen when trusted methods are applied to volatile topics.


Tom Schulte fits the requested model to the proffered data in Michigan.

The table of contents is not available.