University of Virginia Library

Section 33. (k) Statistics of the Moral Situation.

Upon the first glance it might be asserted that statistics and psychology have nothing to do with each other. If, however, it is observed that the extraordinary and inexplicable results presented by statistics of morals and general statistics influence our thought and reflection unconditionally, its importance for criminal psychology can not be denied. Responsibility, abundance of criminals, their distribution according to time, place, personality, and circumstances, the regularity of their appearance, all these have so profound an influence upon us both essentially and circumstantially that even our judgments and resolutions, no less than the conduct and thought of other people whom we judge, are certainly altered by them.[1] Moreover, probability and statistics are in such close and inseparable connection that we may not make use of or interpret the one without the other. Eminent psychological contributions by Münsterberg show the importance the statistical problems have for psychology. This writer warns us against the over-valuation of the results of the statistics of morality, and believes that its proper tendencies will be discovered only much later. In any event the real value of statistical synthesis and deduction can be discovered only when it is closely studied. This is particularly true with regard to criminal conditions. The works of many authors[2] teach us things that would not otherwise be learned, and they would not be dealt with here if only a systematic study of the works themselves could be of use. We speak here only of their importance for our own discipline. Nobody doubts that there are mysteries in the figures and figuring of statistics. We admit honestly that we know no

more to-day than when Paul de Decker discussed Quetelet's labors in statistics of morality in the Brussels Academy of Science, and confessed what a puzzle it was that human conduct, even in its smallest manifestations, obeyed in their totality constant and immutable laws. Concerning this curious fact Adolf Wagner says: "If a traveler had told us something about some people where a statute determines exactly how many persons per year shall marry, die, commit suicide, and crimes within certain classes,—and if he had announced furthermore that these laws were altogether obeyed, what should we have said? And as a matter of fact the laws are obeyed all the world over."[3]

Of course the statistics of morality deal with quantities not qualities, but in the course of statistical examination the latter are met with. So, e. g., examinations into the relation of crime to school-attendance and education, into the classes that show most suicides, etc., connect human qualities with statistical data. The time is certainly not far off when we shall seek for the proper view of the probability of a certain assumption with regard to some rare crime, doubtful suicide, extraordinary psychic phenomena, etc., with the help of a statistical table. This possibility is made clearer when the inconceivable constancy of some figures is considered. Suppose we study the number of suicides since 1819 in Austria, in periods of eight years. We find the following figures, 3000, 5000, 6000, 7000, 9000, 12000, 15000—i. e., a regular increase which is comparable to law.[4] Or suppose we consider the number of women, who, in the course of ten continuous years in France, shot themselves; we find 6, 6, 7, 7, 6, 6, 7; there is merely an alternation between 6 and 7. Should not we look up if in some one year eight or nine appeared? Should not we give some consideration to the possibility that the suicide is only a pretended one? Or suppose we consider the number of men who have drowned themselves within the same time: 280, 285, 292, 276, 257, 269, 258, 276, 278, 287,—Wagner says rightly of such figures "that they contain the arithmetical relation of the mechanism belonging to a moral order which ought to call out even greater astonishment than the mechanism of stellar systems."

Still more remarkable are the figures when they are so brought together that they may be seen as a curve. It is in this way that Drobisch brings together a table which distributes crime according

to age. Out of a thousand crimes committed by persons between the ages of:

Through both columns a definite curve may be drawn which grows steadily and drops steadily. Greater mathematical certainty is almost unthinkable. Of similar great importance is the parallelization of the most important conditions. When, e. g., suicides in France, from 1826 to 1870 are taken in series of five years we find the figures 1739, 2263, 2574, 2951, 8446, 3639, 4002, 4661, 5147; if now during that period the population has increased from 30 to only 36 millions other determining factors have to be sought.[5]

Again, most authorities as quoted by Gutberlet,[6] indicate that most suicides are committed in June, fewest in December; most at night, especially at dawn, fewest at noon, especially between twelve and two o'clock. The greatest frequency is among the half-educated, the age between sixty and seventy, and the nationality Saxon (Oettingen).

The combination of such observations leads to the indubitable conclusion that the results are sufficiently constant to permit making at least an assumption with regard to the cases in hand. At present, statistics say little of benefit with regard to the individual; J. S. Mill is right in holding that the death-rate will help insurance companies but will tell any individual little concerning the duration of his life. According to Adolf Wagner, the principal statistical rule is: The law has validity when dealing with great numbers; the

constant regularity is perceivable only when cases are very numerous; single cases show many a variation and exception. Quetelet has shown the truth of this in his example of the circle. "If you draw a circle on the blackboard with thick chalk, and study its outline closely in small sections, you will find the coarsest irregularities; but if you step far back and study the circle as a whole, its regular, perfect form becomes quite distinct." But the circle must be drawn carefully and correctly, and one must not give way to sentimentality and tears when running over a fly's legs in drawing. Emil du Bois-Reymond[7] says against this: "When the postmaster announces that out of 100,000 letters a year, exactly so and so many come unaddressed, we think nothing of the matter—but when Quetelet counts so and so many criminals to every 100,000 people our moral sense is aroused since it is painful to think that we are not criminals simply because somebody else has drawn the black spot." But really there is as little regrettable in this fact as in the observation that every year so and so many men break their legs, and so and so many die—in those cases also, a large number of people have the good fortune not to have broken their legs nor to have died. We have here the irrefutable logic of facts which reveals nothing vexatious.

On the other hand, there is no doubt that our criminal statistics, to be useful, must be handled in a rather different fashion. We saw, in studying the statistics of suicide, that inferences with regard to individual cases could be drawn only when the material had been studied carefully and examined on all sides. But our criminological statistic is rarely examined with such thoroughness; the tenor of such examination is far too bureaucratic and determined by the statutes and the process of law. The criminalist gives the statistician the figures but the latter can derive no significant principles from them. Consider for once any official report on the annual results in the criminal courts in any country. Under and over the thousands and thousands of figures and rows of figures there is a great mass of very difficult work which has been profitable only in a very small degree. I have before me the four reports of a single year which deal with the activities of the Austrian courts and criminal institutions, and which are excellent in their completeness, correctness, and thorough revision. Open the most important,—the results of the administration of criminal law in the various departments of the country,—and you find everything recorded:—how many

were punished here and how many there, what their crimes were, the percentage of condemned according to age, social standing, religion, occupation, wealth, etc.; then again you see endless tables of arrests, sentences, etc., etc. Now the value of all this is to indicate merely whether a certain regularity is discoverable in the procedure of the officials. Material psychologically valuable is rare. There is some energetic approximation to it in the consideration of culture, wealth, and previous sentences, but even these are dealt with most generally, while the basis and motive of the death-sentence is barely indicated. We can perceive little consideration of motives with regard to education, earlier life, etc., in their relation to sentencing. Only when statistics will be made to deal actually and in every direction with qualities and not merely with quantities will they begin to have a really scientific value.