This argument is reinforced by the further consideration that even the most objective statements of fact involve _value-judgements_ in their ultimate analysis. For they express, often explicitly and always implicitly, the choices and valuations by which a variety of pretenders to reality have been examined and sifted, and the most valuable have been declared ‘truly real’. We have seen that in a scientific inquiry the ‘facts’ must always be taken as _alleged_ facts, discovered up to date; hence a science must always be ready to defend the ‘facts’ it recognizes, when they are challenged, and to show wherein they excel conflicting allegations. The accepted ‘facts’ of a science, therefore, are always allegations which are thought to possess greater _value_ than any known alternative; hence no sharp or absolute distinction between judgements of fact and judgements of value can be maintained. It becomes, moreover, quite possible that incompatible allegations of fact may in the actual state of a science be so nearly balanced that there is no convincing reason to prefer one to another, or at any rate none that could prevail against any ordinary temperamental bias. Consequently, in such cases the bias will condition the visibility of the ‘fact’; it will be bathed in a ‘subjective’ atmosphere, and the ‘eye of faith’ will be necessary to perceive it. No doubt such situations are inconvenient, and repellent to the scientific spirit; but they do not occur only in the misty regions of religion and philosophy, and scientific alternatives like ‘chance’ or ‘design’, ‘miracle’ or ‘law’, ‘mechanism’ or ‘vitalism’, determinism or indeterminism are essentially of this order. There is no reason, therefore, why logic should not recognize them and acknowledge that the scientific ‘facts’ may be ambiguous, in the sense that further experience and experiments are needed to determine their character. As a rule, to judge by the past, further inquiry will resolve the ambiguity; but it may well be an illusion to assume that it must do so, and in some of the most important cases the decision will certainly be long in coming. Thus the student of animal behaviour will probably long be left with a choice between minimizing the displays of animal intelligence and assimilating them to the human, while it will probably always be possible to put a pessimistic or an optimistic interpretation upon the facts of life as a whole. A scientific logic therefore should radically disabuse the mind of any excessive trust in ‘facts’. It is a superstition that ‘facts’ are plain, straightforward, and easy to discover; they are often subtle and recondite and relative to circumstances, changing their aspect to suit their scientific environment like any chameleon.