What is science?
Science refers to the various processes through which knowledge is gained through reproducible observation or experimentation. An observation that is not reproducible is not a scientific observation. In the context of geology, we seek knowledge of the natural world, both now and throughout its history.
Science is a data-driven enterprise. In the absence of reproducibly obtained data, there is no science. We do not develop scientific hypotheses simply through deductive reasoning from supposed "first principles," in the manner that some ancient Greek philosophers sought to understand the world. Rather, we reproducibly collect data, and those data define and constrain our ideas of how the natural world actually works.
Quantitative scientific observations are always reported with an associated error estimate, either explicitly by stating a confidence interval or implicitly through an appropriate choice of significant figures. For example, the age of a volcanic ash bed might be reported to a 95% confidence interval as 138.4 +/- 0.2 million years, which means that we would expect that the average of a very large number of duplicate analyses of that ash to fall between 138.2 and 138.6 million years. Formal statistical procedures exist for obtaining reliable error estimates, and for choosing the appropriate number of significant figures.
The reproducible observations constitute the "facts" that are related to one another through the development of hypotheses and models. Relationships among the primary data are generally expressed mathematically. The defining characteristic of a scientific hypothesis is that it must be testable or falsifiable using reproducible observations. An idea that cannot be tested is not a scientific hypothesis.
Scientific hypotheses are developed to fit data; data are not collected to fit or support hypotheses post facto. Rather, data are collected to test or disprove hypotheses.
References and suggested reading
Ben-Ari, Moti, 2005, Just a theory -- exploring the nature of science: Amherst, New York, Prometheus Books, 237 p., ISBN 1-59102-285-1.
Bridgman, P.W., 1927, The logic of modern physics: New York, Arno Press [reprinted 1980], 228 p., ISBN 0-405-12594-1.
Bronowski, J., 1965, Science and human values: New York, Harper and Row, 119 p., ISBN 0-06-097281-5.
Bronowski, J., 1978, The common sense of science: Cambridge, Massachusetts, Harvard University Press, 154 p., ISBN 0-674-14651-4.
Lee, J.A., 2000, The scientific endeavor -- a primer on scientific principles and practice: San Francisco, California, Benjamin Cummings, 186 p., ISBN 0-8053-4596-5.
Morrison, P., and Morrison, P., 1987, The ring of truth -- an inquiry into how we know what we know: New York, Random House, 307 p., ISBN 0-394-55663-1.
Scott, E.C., 2004, Evolution vs. creationism -- an introduction: Berkeley, University of California Press, 272 p., ISBN 0-520-24650-0. Chapter 1, Science: "truth without certainty," is particularly relevant.
Snow, C.P., 1959, The two cultures: Cambridge, UK, Cambridge University Press, 107 p., ISBN 0-521-45730-0 (Canto edition, 1993).
Wolpert, L., 1992, The unnatural nature of science -- why science does not make (common) sense: Cambridge, Massachusetts, Harvard University Press, 191 p., ISBN 0-674-92980-2.
The information on this page was written and approved by the faculty of the Geology Department at Baylor University.