In my view, the philosophy of science is not a branch of psychology (although many post-Kuhnian philosophers of science who have taken the ‘sociological turn’ think otherwise). The reference to "goals of science" is not a reference to the intentions of scientists in doing science. Science may achieve a degree of truth with or without scientists intending that that is what they are doing.

An Analogy: People may have moral beliefs against extramarital sex. Their behavior may be caused by a belief in a moral goal, but the behavior may achieve many other things, like a lower incidence of venereal disease, higher stability of family units and so on.

Likewise, scientists may be solely concerned with personal fame, and unconcerned with the truth of theories (not that I believe that). Yet, science might still make progress towards the truth even if it were unintended. Philosophy of science is about the achievements of science, and how they come about, rather than the psychology of scientists. I mention this, because I often see philosophers of science appealing uncritically to the opinions of scientists in order to support a view about science.

Definition 1: Goals, or ends, refer to the achievements, or potential achievements of science. Goals like truth, approximate truth, closeness to the truth, empirical adequacy, predictive accuracy are called epistemic goals. Social, moral, political, and pragmatic goals are non-epistemic goals.

Definition 2: Means, methods, or criteria refer to what scientists actually do when they do science, or to the values they use to judge the achievement or degree of achievement of epistemic goals. Means can only refer to the features of scientific theories or models that scientists can ‘see’. For example, "Choose the model that is closest to the truth" is not a means because scientists cannot ‘see’ that a model or theory is true. "Choose the model that best fits the background data" is a genuine means because scientists can ‘see’ which model best fits the background data.

• Means are not valued for their own sake, but only as a means towards a goal.
• In talking about cognitive values, goals and means to goals get mixed up together.

Traditional philosophy of science is primarily concerned with the relationship between methods and epistemic goals. But it is impossible to study this relationship if the two things are not clearly separated. For example, is simplicity a goal of science, or is it a means to a goal of science (and remember this is not a question of what scientists believe). Simplicity, as such, is not an epistemic goal (it does not mention truth), and it is only an epistemic, or cognitive value in so much as it does indicate the achievement of some epistemic goal.

  Example: Under what circumstances does choosing the best fitting model lead to the acceptance of hypotheses that are closer to the truth?

The benefits of such means-goal analysis might be:

1. Better Understanding of Science: If successful, this would be useful in understanding how science works, for it could tell us to the extent that social and political factors are "getting in the way" or "helping promote" progress towards the truth. That is, it would play a role in explaining the success of science (to the extent that it is successful) rather than merely describing the way science is.
2. Better Prescription for Science: A knowledge of the effectiveness of scientific methods would also help in prescribing how the improve science. Although, at best, it would enable us to estimate the gains and losses from different science policies (e.g., how should scientists or potential scientists be trained?). It would not tell us how to trade off epistemic gains against social or political costs of a given policy, and therefore would not tell us what science ought to be. But it would form an important component of a normative prescription for science—one that is not provided solely by the history or sociology of science.

Note: This agenda for the philosophy of science is very ambitious, and it is true that it has not made much progress in fulfilling this role to date. It is a very immature discipline at the moment. My argument is only that there is a role here that is not filled by the history or sociology of science.

Puzzle: If a philosophy of science mentions only cognitive values, and all science is laden with non-cognitive values, and the philosophy of science is not a prescription of science, then what is the philosophy of science about?

Answer in Brief: The philosophy of science also studies the effectiveness of scientific methods and techniques in bringing about epistemic goals.

For example, it might tell us when, and under what circumstances, a naïve empiricist methodology (of choosing the best fitting model) is better than a method of trading off simplicity and fit in obtaining better approximations to the truth.

This is to study science as a means to the goal of truth, irrespective of what scientists believe is achieved by their methods, or their psychological reasons for their decisions. It is an objective study of science is a stronger sense than studying the inter-subjective agreements of community beliefs (which is the main concern of philosophers of science who took the ‘sociological turn’).

Consider any putative goal of science, whether it be the truth of theories, the predictive accuracy of models, or the economic prosperity of the United States. Call the goal ‘X’. Now consider two, or more, ways or methods of doing science. Call them A and B. It is now an objective question whether A is more effective than B in achieving X.

Definition: Operational questions are questions about the relative effectiveness of ways or methods of doing science in achieving a goal X.

• Operational questions have nothing to do with what scientists believe.
• Operational questions are hard questions.

Operational theses are answers to operational questions. For example, "A is more effective than B in achieving X" is a typical operational thesis on my definition. Such theses are weakly normative in the sense that they imply ‘ought’ statements when coupled with goal statements. For example, if one could establish that A is more effective than B in achieving X, and X is the goal of science, then it would follow that one ought to adopt A as the methodology of science.

There is a huge difference between social psychological questions and operational questions, although the distinction is not always clear. For example, compare the following normative arguments:

1. The scientific community agrees that method A is better than method B at achieving X.

X is the goal of the community. Therefore, the scientific community ought to adopt method A.

2. Method A is better than method B at achieving X.

There is an important difference between these arguments. The first provides a subjective justification for using method A, while the second provides an objective justification for the same action, or policy. The operational claim in (2) is harder to establish, but it has stronger implications. That is an important tradeoff.

Notes:

1. Scientists themselves are not always particularly well qualified to answer operational questions. An analogy, try asking an someone how to walk (other than a physical therapist, or someone who had to relearn to walk as an adult).
2. On this characterization, the philosophy of science is neutral to the sociology of science and to, for example, feminist critiques of science. It may well turn out that feminist ways of doing science are more effective in achieving epistemic goals (as Sandra Harding suggests), or that the elimination of bias has beneficial effects (as in the examples that Kathleen Okruhlik describes).
3. Non-cognitive values may enter into the definition of the methods, and they may affect the beliefs about the effectiveness of methods, but they do not affect the effectiveness of the methods themselves. In that sense, non-cognitive values do not enter into the philosophy of science.