This is a wonderful (not too formally written) little volume that provides good background for teachers dealing with nature of science in the classroom. Kosso begins by considering the role of theory in science. Why are explanations at this level so important? He walks through decidedly philosophical terrain in describing how explanation works indirectly through hypothesis and deduction, why data can never clinch a single theory, and how observations themselves involve other theoretical assumptions. Theory can never be absolutely certain.

At the same time, observations are critical. They must be accountable (for example--they must be repeatable and follow certain standards of evidence-gathering). Independent observations, too, give us some confidence that theories represent reality (while still not being infallible). In discussing how evidence can be objective, Kosso addresses a number of cases about "unobservables" and nicely shows the indirect route by which we make our scientific conclusions.

Ultimately, Kosso presents the coherence of observations as the central form of justification in science. Observations, while important, do not confirm theories in a simple one-to-one relation-ship. Nor can theory disregard observations or causal demonstrations. What lies in the balance is the way to understand how scientific theories are both justified and limited. A teacher armed with such concepts can challenge students to articulate how their observations are "accountable," how their conclusions depend on other assumptions, and how they reason through multiple arguments to robust conclusions.

This book is philosophical. Do not expect to find extended examples or anything that relates directly to classroom teaching. But the exposition draws on common sense and easy analogies, and where the reasoning is more complex, it weaves its way fairly gradually, never losing the initiate. Again, it will help the reflective teacher think about how to structure labs and students' discussions of their results.

--Douglas Allchin