copyright Robert M. Ellis 2011

Common but unhelpful assumptions 6: The fact-value distinction

The fact-value distinction stems from the logical claim made by Hume that no claim about a value can be implied by one about a fact. For example "The child is drowning" might in many practical situations be interpreted to mean "We ought to save the child", but formally speaking the first statement does not imply the second without further moral assumptions needing to be made (such as "We should save try to save drowning people when we are able to"). For Hume and for many of those who accept this distinction, it is justified by a difference in the ways that we justify facts and values respectively - facts are taken to be justified by observation (or perhaps by reasoning), whilst values are to be justified only by desires, even if these are desires that we share with the rest of our society. To take facts to imply values, then, would be to impose wishful thinking on the objective world.

This distinction is dependent on the 'abstracted turn' discussed in assumption 4 in this section of the website. It may be possible to separate facts from values in abstract analysis, but in concrete experience they always appear in interdependence. If we think of human beings believing facts rather than facts by themselves, there has never been a factual claim without a motive for its assertion or a practical goal involved in its belief. Even in the context of a lecture or a scientific paper where knowledge 'for its own sake' is being communicated, the context of communication brings in a motive: the lecturer or researcher wants to share his or her knowledge with the student or reader. We could similarly never isolate values without facts, for whenever a value is expressed, certain background beliefs are taken for granted as part of the practical context: I could not have any duty to try to save a drowning child if I did not believe that there was a drowning child to save.

The problem with the fact-value distinction is not just that it abstracts from the concrete situation (which might just limit its applicability), but that it forms the basis of further reasoning about ethics of a kind which has been disastrous for the understanding of ethics in Western civilisation. In ethical thinking in analytic philosophy, and indeed in many other contexts, an abstraction that is correct only in the abstract is imposed on a concrete situation. If values are assumed to be distinct from facts, and facts are 'objective' (see assumption no. 3) whilst values are 'subjective', then values are merely a matter of opinion, and one opinion is as good as another opinion. The fact-value distinction thus has a strong tendency to encourage relativism.

However, the bad effects of the fact-value distinction do not stop there. Many people do not accept that values are relative (for example, theists). However, when these same people accept the fact-value distinction (which has become pervasively implicit in popular thinking) the only alternative is to assert that their values are based on special facts, which, because they are metaphysical and beyond the common reach of facts, can only be asserted dogmatically. The fact-value distinction thus also entrenches moral absolutism behind walls of dogma - as we see, for example, in the rise of fundamentalism in reaction to Western relativism. A polarised slanging-match is substituted for any examination of the grounds of values in our experience.

The extent of the confusion spreads even further than this. We also see the erosion of scientific claims to 'objectivity', because science is rightly perceived to have no such absolute claims to a God's eye view of the universe. Creationism or homeopathy are seen by increasing numbers of ignorant people as 'science' because all views are seen as merely subjective values. The subtleties of non-absolute justification in science are very difficult to communicate to a popular audience. Scientists have, however, to a large extent helped to create these difficulties by accepting the fact-value distinction. Science has eventually been undermined because it has tried to create a false absolute position for itself through a differentiation from values that does not apply in practice. If that differentiation was not so widely assumed it would not be so easy for opinion to swing from absolute status to merely relative status.

There is an alternative to the fact-value distinction, and it is not an inevitability in human thought. Firstly, we need to acknowledge that both facts and values are incrementally justified through experience, with neither having a qualitative difference from the other in the form of justification. The justification of both facts and values depends not just on observations of the universe we think we observe, but also on the degree of integration with which we interpret it. I use the term 'integration' here, not just 'coherence', advisedly, because integration involves recognition of beliefs and desires identified with at other times from the present, and the recognition of the limitations of our current understanding, not merely the rational examination for consistency of beliefs held at one moment. The introduction of psychological as well as merely rational and observational criteria has the potential to show scientific and moral objectivity as working on the same continuum, with the differences being only those of emphasis and context. A more integrated scientist is a better investigator able to make judgements that address conditions better, whilst integration also provides the basis of moral judgement and the assessment of facts that are assumed in moral judgement. All of this provides an alternative model of a kind that is not even considered in most current moral discusssion. 

The issue of the disastrous fact-value distinction should not be clouded by making distinctions between strong or weak versions of it, unless the distinction between strong and weak is made on practical rather than merely analytic grounds. The problem with the fact-value distinction is a practical one - not that making a 'weak' distinction is always wrong, but that the vast majority of the time that the distinction is made it leads us into making further distinctions that have the effect of undermining ethics. Ethics in our experience needs to be accepted as having incremental justification, in the same way as common sense leads many people to accept factual claims as having incremental justification. In either case scepticism saves us from absolute thinking - but only if we apply it rigorously and consistently rather than selectively.


Links to related pages on this website:

Relativism (introductory page)

Features of eternalism (from thesis)

Features of nihilism (from thesis)

Hume (from thesis)

What is Buddhist (i.e. Middle Way) Ethics? (from 'A New Buddhist Ethics')


Return to home page

Seven common but unhelpful assumptions in Western thought:

1. The negative implications of scepticism

2. The need to accept or reject metaphysical claims

3. The identification of objectivity with absolute claims

4. The acceptability of pure analysis not applied to concrete contexts

5. An account of meaning confined to representation or expression

6. The fact-value distinction

7. The identification of ego with self

Concepts index

Discussion forum

Return to home page

StumbleuponCiteULikedel.icio.usRedditDigg itFacebook
Social Bookmarking Service