The ideal of science is objectivity. To be objective is to be truthful, to describe the world as it actually is. If the ideal of objectivity were perfectly realized, then science would give us a complete description of the world. Nothing would be left out; there would be no bias.
    But this is impossible. The world is infinitely complex, and description is always finite. Every investigation ignores most of what is there. What scientists study depends upon what they regard as important. Their judgments of what is important are shaped by their predilections, their scientific training and their economic and political interests. If they investigate well, they uncover truths ¾ but never the whole truth.
    The decision to investigate or not investigate a particular problem is not determined by objective criteria. It may depend on technological or practical considerations: Are adequate instruments available? Can the investigation be completed in a reasonable length of time? Is it affordable? But it may also depend on what might loosely be called "political" considerations. Most scientific research is sponsored by institutions that need the information for some purpose; hence the truths uncovered by science generally concern matters that some institution has deemed worthy of investigation. We are likely to be most ignorant in areas that least interest the funders of science; from these areas, surprises are likely to emerge.

    In the United States, the main institutions that sponsor scientific research are large private corporations and various governmental agencies (such as the Department of Agriculture or the Department of Defense). Universities and charitable foundations also fund research directly, but have much smaller budgets. (Most of the research money of university scientists comes directly from governmental agencies or corporations.) Hence government and industry have considerable influence in deciding which questions are asked. The terrain of scientific knowledge is thus skewed by their interests.
    As a result, contemporary science, while mostly true, is not unbiased. It emphasizes those facts which industry and government have wanted to know ¾ and has a built-in tendency to overlook facts that have not interested these institutions.
    There are mechanisms designed to minimize this institutional bias. Research proposals submitted to some governmental agencies (the National Science Foundation, for example) and papers submitted to most scientific journals, are accepted, revised, or rejected according to decisions of peer reviewers ¾ other scientists in the same field. However, if the peer reviewers share the same institutional interests, these interests will still influence the direction of research.
    Agricultural science provides a good example. It is funded largely by corporations that manufacture fertilizers, herbicides, pesticides and other agricultural chemicals. The result is an extensive knowledge of the effects of these chemicals on plant and animal productivity. But because their large-scale ecological effects have been of less interest to their manufacturers, research on their environmental impacts has lagged.
    Conversely, since it is not in the interest of chemical corporations to encourage methods that dispense with their products, there is a very much smaller body of scientific knowledge on organic or self-reliant agriculture. Thus the results of science may be objective, but the distribution of scientific investigation and scientific knowledge among all the things there are to know is anything but.