concepts section: copyright Robert Ellis 2009


The concept of falsification is a vital one for the testing of our beliefs. Without it the search for objectivity is in vain, because we can have no clear indication that we are moving beyond the limitations of one view and into a more adequate one. The possibility of our beliefs being shown to be wrong is a crucial indicator that we are avoiding metaphysical dogma and gaining objectivity in our beliefs. However, at the same time there can be no absolute falsification, because a metaphysical justification for the process cannot be sought within an experiential framework of objectivity. Central to Middle Way philosophy is the claim that falsification can be decisive and objective without being absolute.

To begin with, we need falsification because no belief can be decisively verified. If I believe that all cats are black and meet hundreds of black cats, this is no guarantee that I may not in future meet a white one or a ginger one. An adequate recognition of our degree of ignorance alone demands that we renounce verification and its associated idea of "truth". Falsification, however (as Popper pointed out) is decisive, because I only need to encounter one observation that does not fit my belief in order to recognise its falsity. I only need to observe one white cat to be clear that not all cats are black, so that my earlier belief is shown as false. Popper's central insight in this regard was to recognise that we do not make progress by gaining "true" beliefs, but rather by gradually chipping away at false ones, changing them slightly each time they are disconfirmed, so as to gradually modify our imperfect beliefs to better fit conditions. 

Popper's approach also provides one of the primary ways of recognising metaphysical beliefs: metaphysical beliefs are unfalsifiable beliefs, that is, beliefs that can be maintained regardless of the evidence offered by experience. A metaphysical belief can always be maintained no matter what our experience, because it is sufficiently abstracted from all possible experience to be reinterpreted to fit any possible experience.

However, no falsification is a final falsification, but only a disconfirmation within a framework we have set up. If we think we experience a falsification, we may possibly be mistaken either about the object we think we experience, or about its relationship to the theory. If I think I see a white cat, it is possibly an illusory white cat, or a white animal that superficially resembles a cat but should not properly be counted as a cat, so thus I can cling to the belief that all cats are black in the face of any number of apparent white cats. However, if I define the conditions of my falsificatory experience for myself in advance, and apply those conditions resolutely, such ad hoc rationalisations can be avoided. For example, if I tell myself that any white cat, investigated to a degree I find reasonable to check for the possibility of any mistake, will count as a falsification of my theory, my advance stipulation of the conditions of falsification prevents ad hoc rationalisation to my own satisfaction, though not necessarily to the satisfaction of others.

The degree of investigation of a possible falsification that one "finds reasonable" is also a criterion dependent on one's degree of intellectual virtue for its successful implementation. No claim can be falsified without some reliance on the objectivity of the falsifier, because it is ultimately not claims that can be objective but people. Falsification is therefore primarily a technique for avoiding self-deception and training oneself into a more scientific objectivity. Objectivity of the psyche is interdependent with objectivity in methods of observation, and neither can be conclusively proven to be a basis of objectivity independently.

Unfalsifiable beliefs, however, exclude any such training from the beginning, and encourage the self-deception of thinking that one's experiences verify one's beliefs. For example, the Natural Law belief that observable nature is good may be constantly "verified" by our satisfactory observation of "natural" phenomena, but no possible observation can be stipulated under this system that would show that nature was not good. To stipulate verifications is far less demanding than to stipulate falsifications, because we are constantly seeking confirmation of our beliefs and interpreting the world in those terms, but falsifications take us by surprise. Falsifiable beliefs open the possibility of our being taken by surprise and thus becoming more objective, but metaphysical beliefs do not, instead entrenching us further in unassailable certainties.


Links to related discussion

The heuristic problem

Verification and falsification (scroll down to section c)

The Middle Way and science


Return to concepts page