structure of their society, are embodied in the most elusive of all materials; the human being. But not even in human mind or memory are these laws to be found definitely formulated. The natives obey the forces and commands of the tribal code, but they do not comprehend them; exactly as they obey their instincts and their impulses, but could not lay down a single law of psychology. The regularities in native institutions are an automatic result of the interaction of the mental forces of tradition, and of the material conditions of environment. Exactly as a humble member of any modern institution, whether it be the state, or the church, or the army, is of it and in it, but has no vision of the resulting integral action of the whole, still less could furnish any account of its organisation, so it would be futile to attempt questioning a native in abstract, sociological terms. The difference is that, in our society, every institution has its intelligent members, its historians, and its archives and documents, whereas in a native society there are none of these. After this is realised an expedient has to be found to overcome this difficulty. This expedient for an Ethnographer consists in collecting concrete data of evidence, and drawing the general inferences for himself. This seems obvious on the face of it, but was not found out or at least practised in Ethnography till field work was taken up by men of science. Moreover, in giving it practical effect, it is neither easy to devise the concrete applications of this method, nor to carry them out systematically and consistently.
Though we cannot ask a native about abstract, general rules, we can always enquire how a given case would be treated. Thus for instance, in asking how they would treat crime, or punish it, it would be vain to put to a native a sweeping question such as, “How do you treat and punish a criminal?” for even words could not be found to express it in native, or in pidgin. But an imaginary case, or still better, a real occurrence, will stimulate a native to express his opinion and to supply plentiful information. A real case indeed will start the natives on a wave of discussion, evoke expressions of indignation, show them taking sides—all of which talk will probably contain a wealth of definite views, of moral censures, as well as reveal the social mechanism set in motion by the crime committed. From there, it will be easy to lead them on to speak of other similar cases, to remember other actual occurrences or to discuss them in all their implications and aspects. From this material, which ought to cover the widest possible range of facts, the inference is obtained by simple induction. The scientific treatment differs from that of good common sense, first in that a student will extend the completeness and minuteness of survey much further and in a pedantically systematic and methodical manner; and secondly, in that the scientifically trained mind, will push the inquiry along really relevant lines, and towards aims possessing real importance. Indeed, the object of scientific training is to provide the empirical investigator with a mental chart, in accordance with which he can take his bearings and lay his course.
To return to our example, a number of definite cases discussed will reveal to the Ethnographer the social machinery for punishment. This is one part, one aspect of tribal authority. Imagine further that by a similar method of inference from definite data, he arrives at understanding leadership in war, in economic enterprise, in tribal festivities—there he has at once all the data necessary to answer the questions about tribal government and social authority. In actual field work, the comparison of such data, the attempt to piece them together, will often reveal rifts and gaps in the information which lead on to further investigations.
From my own experience, I can say that, very often, a problem seemed settled, everything fixed and clear, till I began to write down a short preliminary sketch of my results. And only then, did I see the enormous deficiencies, which would show me where lay new problems, and lead me on to new work. In fact, I spent a few months between my first and second expeditions, and over a year between that and the subsequent one, in going over all my material, and making parts of it almost ready for publication each time, though each