TESTIMONY AND AUTHORITY. 71 Any discussion concerning possible safeguards against misinterpretation of the statements of others would take us into a region of logic bordering on that of definition. Let it suffice here to say that interpretation must be self- consistent ; must neglect nothing in the data ; must give full effect to the context immediate, systematic and his- torical and must not be limited to mere grammatical or logical analysis, but must take due account of the style and intention of the writer or speaker. Let us turn first to the part taken by the assertor. It is clear that to impart information implies first the getting of it. The reliability of testimony depends not only upon the conveyance of the information being correctly performed, but also upon its having been correctly obtained ; not only upon the veracity of the witness, but also upon his cogni- zance of the matter in hand. Two conditions or groups of conditions therefore might naturally be expected to unfold themselves ; one set arising out of the process of conveying information, and the other from the processes of obtaining it. Beyond these we should expect to find another if there be any influence likely to adversely affect equally both the acquisition and the conveyance of information. It is uni- versally recognised that however careful and conscientious a person may be, yet the effect of interest may be to lead him unwittingly into error throughout the acquisition, the retention and the conveyance of information. To be free from such unseen influence in completeness is the unattain- able ideal ; but practical freedom from bias with regard to some particular matter is not so far out of reach. This relative freedom from bias is then one of the conditions that an assertor must conform to in order to be trustworthy. In the process of conveying information there are evi- dently involved intention and capacity. A person will speak truly provided he wants to and can. Nevertheless a witness may be perfectly sincere, and yet fail to recall accurately the matter asserted ; or even if he can do this he may still not succeed in expressing exactly what he has in mind. This factor then gives rise to two conditions one that the assertor must be sincere, and the other that he must be accurate in memory and expression. The great difficulty is, of course, to find out whether there is bias or insincerity in any given case. To a certain extent, as we shall see, corroboration deals with it ; but if we are without corroboration there is nothing but to consider the circumstances under which the statements are made in
