OBTAINING EVIDENCE BY PSYCHO-PHYSIOLOGICAL METHODS
But it is not enough to realize empirically that there are possibilities of error in testimonial evidence, and to adopt an attitude of caution. It is desirable to learn, if we can, how extensive are the possibilities of error in testimonial inference.
In the preceding chapter we were concerned with those methods of investigation that are relatively free from the influence of the personal equation: methods, the details and products of which lend themselves to control for the most part, approximately as do procedures in the laboratories of exact science. We have here to consider certain methods from which the personal equation is far from being so effectively excluded; methods, indeed, in which subtle, uncontrollable psychologic processes play a part, the nature of which may not be measured or described. These are methods, therefore, that furnish evidence to be sure, but evidence that, as a basis from which inferences as to identity, e.g., may be drawn, is less satisfactory than the correspondence between finger-prints.
It is not difficult to appreciate the truth of the foregoing statement. Could you recognize the man who assailed you in the street yesterday if you were to meet him again today? Recognition in this connection is a case of identification. And in the criminal courts, witnesses, times without number, are called upon to identify, among all the persons in the court room, the one whom they saw, not yesterday, but months before, committing the criminal act that is in question. One witness makes positive affirmation and points to the prisoner before the bar. Another as strongly denies that the one whom he saw committing the act is in the room. Each witness may be intellectually honest. For the process of identification involves observing a feature or characteristic of the person or thing identified: remembering that that feature was present in a person or thing that was earlier observed at a partic