The object always should be, not to translate from English into signs, but to ascertain the real signs and their meaning.  By far the most satisfactory mode of obtaining this result is to induce Indians or other gesturers observed to tell stories, make speeches, or hold talks in gesture, with one of themselves as interpreter in his own oral language if the latter is understood by the observer, and, if not, the words, not the signs, should be translated by an intermediary linguistic interpreter.  It will be easy afterward to dissect and separate the particular signs used.  This mode will determine the genuine shade of meaning of each sign, and corresponds with the plan now adopted by the Bureau of Ethnology for the study of the tribal vocal languages, instead of that arising out of exclusively missionary purposes, which was to force a translation of the Bible from a tongue not adapted to its terms and ideas, and then to compile a grammar and dictionary from the artificial result.  A little ingenuity will direct the more intelligent or complaisant gesturers to the expression of the thoughts, signs for which are specially sought; and full orderly descriptions of such tales and talks with or even without analysis and illustration are more desired than any other form of contribution.

The original authorities, or the best evidence, for Indian signs—­i.e., the Indians themselves—­being still accessible, the collaborators in this work should not be content with secondary authority.  White sign talkers and interpreters may give some genuine signs, but they are very apt to interpolate their own improvements.  Experience has led to the apparently paradoxical judgment that the direct contribution of signs purporting to be those of Indians, made by a habitual practitioner of signs who is not an Indian, is less valuable than that of a discriminating observer who is not himself an actor in gesture speech.  The former, being to himself the best authority, unwittingly invents and modifies signs, or describes what he thinks they ought to be, often with a very different conception from that of an Indian.  Sign language not being fixed and limited, as is the case with oral languages, expertness in it is not necessarily a proof of accuracy in anyone of its forms.  The proper inquiry is not what a sign might, could, would, or should be, or what is the best sign for a particular meaning, but what is any sign actually used for such meaning.  If any one sign is honestly invented or adopted by any one man, whether Indian, African, Asiatic, or deaf-mute, it has its value, but it should be identified to be in accordance with the fact and should not be subject to the suspicion that it has been assimilated or garbled in interpretation.  Its prevalence and special range present considerations of different interest and requiring further evidence.

The genuine signs alone should be presented to scholars, to give their studies proper direction, while the true article can always be adulterated into a composite jargon by those whose ambition is only to be sign talkers instead of making an honest contribution to ethnologic and philologic science.  The few direct contributions of interpreters to the present work are, it is believed, valuable, because they were made without expression of self-conceit or symptom of possession by a pet theory.