Exclusion of Evidence Obtained with an Ideology in Mind

Some suggest that evidence obtained with an ideology in mind ought not to enter into knowledge, i.e. ought not to affect one's belief in a theory. When Rose [61,62] accuses science of ignoring, on the basis of an asserted need to divide the intellectual labour of pursuing science from manual and emotional labour, evidence obtained, often by women, in the course of caring labour, and in the course of self-health care and the self-examination which renders self-health care possible, it is this addition to positivism which she criticizes, for caring labour is necessarily the pursuit of an ideology, in which a key value is the welfare of the person for whom one is caring, and self-health care the pursuit of an ideology, in which a key value is one's own physiological well-being. Young [75] has also noticed this tendency among some positivists. It is exemplified by the situation in research into the environmental effects of the use of tributyltin, as a biocide, in paint for ships' hulls, where the paint industry on the one hand, and governments and environmental pressure groups on the other, each attempt [72] to dismiss research funded by the other, as tainted by its ideological motivations.

This addition to positivism sits uneasily with the core positivist principle of taking into account evidence obtained in disciplines other than one's own (section 2.1.4,) given how easy, and indeed how justified (section 3.2,) it often is to accuse practitioners of another discipline of pursuing an ideology. There exist [64] documented examples of apparently traditional scientists performing experiments on their own bodies, concerning which they could hardly have been without values. In addition, Swanson [72] feels at liberty to describe the behaviour of both sides, in the tributyltin debate, as `shoddy.' These examples may indicate that this addition to positivism has never been fully accepted into orthodox science⁹.

To a Bayesian, this exclusion of ideologically-gathered evidence represents a statement that, if evidence $E$ has been gathered with an ideology in mind then, for all theories $T$,

$$P(T\vert E) = P(T)$$ (16)

$$\iff{}P(E\vert T) = P(E)\textrm{,}\quad\textrm{unless}\quad{}P(T) = 0\textrm{.}$$ (17)

While certain theories $T$ might predict an effect on the evidence, due to the ideological motivation, there is no reason to suppose that the influence will always take the very specific form of rendering $P(E\vert T)$ equal to $P(E)$; indeed, equation 17 sets the prior probability to zero for any theory which predicts that the effect of ideological motivation on evidence obtained will take anything other than one, very specific, form, and is therefore, with respect to any such theory, an example of logical probability (section 3.1.) This leads to the first similarity between Bayesian statistics and standpoint epistemology that the author wishes to stress: a Bayesian view requires, as does [28] standpoint epistemology, that evidence, such as that arising from caring labour and self-health care, which was obtained in the pursuit of an ideology, be fully accepted into knowledge.