[1174] In the Swiss edition2 I purposely set out the results of the astrological statistics in tabular form in this chapter, so that the reader could gain some insight into the behaviour of the figures—in other words, see for himself how fortuitous these results were. Subsequently, I wanted to suppress that account of the experiment in the English edition and for a very peculiar reason indeed. That is, it has been forcibly borne in on me that practically nobody has understood it the right way, despite—or perhaps because—of the fact that I took the trouble to describe the experiment in great detail and in all its vicissitudes. Since it involved the use of statistics and comparative frequencies, I had the (as it now seems) unlucky idea that it would be helpful to present the resultant figures in tabular form. But evidently the suggestive effect emanating from statistical tables is so strong that nobody can rid himself of the notion that such an array of figures is somehow connected with the tendentious desire to prove something. Nothing could have been further from my mind, because all I intended to do was to describe a certain sequence of events in all its aspects. This altogether too unassuming intention was misunderstood all round, with the consequence that the meaning of the whole exposition went by the board.
[1175] I am not going to commit this mistake again, but shall make my point at once by anticipating the result: the experiment shows how synchronicity plays havoc with statistical material. Even the choice of my material seems to have thrown my readers into confusion, since it is concerned with astrological statistics. One can easily imagine how obnoxious such a choice must be to a prudish intellectualism. Astrology, we are told, is unscientific, absolute nonsense, and everything to do with it is branded as rank superstition. In such a dubious context, how could columns of figures mean anything except an attempt to furnish proofs in favour of astrology, proofs whose invalidity is a foregone conclusion? I have already said that there was never any question of that—but what can words do against numerical tables?
[1176] We hear so much of astrology nowadays that I determined to inquire a little more closely into the empirical foundations of this intuitive method. For this reason I picked on the following question: How do the conjunctions and oppositions of the sun, moon, Mars, Venus, ascendant, and descendant behave in the horoscopes of married people? The sum of all these aspects amount to fifty.
[1177] The material to be examined, namely, marriage horoscopes, was obtained from friendly donors in Zurich, London, Rome, and Vienna. The horoscopes, or rather the birth data, were piled up in chronological order just as the post brought them in. The misunderstanding already began here, as several astrological authorities informed me that my procedure was quite unsuited to evaluating the marriage relationship. I thank these amiable counsellors, but on my side there was never any intention of evaluating marriage astrologically but only of investigating the question raised above. As the material only trickled in very slowly I was unable to restrain my curiosity any longer, and I also wanted to test out the methods to be employed. I therefore took the 360 horoscopes (i.e., 180 pairs) that had so far accumulated and gave the material to my coworker, Dr. Liliane Frey-Rohn, to be analysed. I called these 180 pairs the “first batch.”
[1178] Examination of this batch showed that the conjunction of sun (masculine) and moon (feminine) was the most frequent of all the 50 aspects, occurring in 10% of all cases. The second batch, evaluated later, consisted of 440 additional horoscopes (220 pairs) and showed as the most frequent aspect a moon-moon conjunction (10.9%). A third batch, consisting of 166 horoscopes (83 pairs), showed as the most frequent aspect the ascendant-moon conjunction (9.6%).
[1179] What interested me most to begin with was, of course, the question of probability: were the maximum results obtained “significant” figures or not; that is, were they improbable or not? Calculations undertaken by a mathematician showed unmistakably that the average frequency of 10% in all three batches is far from representing a significant figure. Its probability is much too great; in other words, there is no ground for assuming that our maximum frequencies are more than mere dispersions due to chance. Thus far the result of our statistics (which nevertheless cover nearly one thousand horoscopes) is disappointing for astrology. The material is, however, much too scanty for us to be able to draw from it any conclusions either for or against.
[1180] But if we look at the results qualitatively, we are immediately struck by the fact that in all three batches it is a moon conjunction, and what is more—a point which the astrologer will doubtless appreciate—a conjunction of moon and sun, moon and moon, moon and ascendant, respectively. The sun indicates the month, the moon the day, and the ascendant the “moment” of birth. The positions of sun, moon, and ascendant form the three main pillars of the horoscope. It is altogether probable that a moon conjunction should occur once, but that it should occur three times is extremely improbable (the improbability increases by the square each time), and that it should single out precisely the three main positions of the horoscope from among 47 other possibilities is something supranormal and looks like the most gorgeous falsification in favour of astrology.
[1181] These results, as simple as they are unexpected, were consistently misunderstood by the statisticians. They thought I wanted to prove something with my set of figures, whereas I only wished to give an ocular demonstration of their “chance” nature. It is naturally a little unexpected that a set of figures, meaningless in themselves, should “arrange” a result which everybody agrees to be improbable. It seems in fact to be an instance of that possibility which Spencer-Brown has in mind when he says that “the results of the best-designed and most rigorously observed experiments in psychical research are chance results after all,” and that “the concept of chance can cover a wider natural field than we previously suspected.”3 In other words what the previous statistical view obliged us to regard as “significant,” that is, as a quasi-intentional grouping or arrangement, must be regarded equally as belonging to the realm of chance, which means nothing less than that the whole concept of probability must be revised. One can also interpret Spencer-Brown’s view as meaning that under certain circumstances the quality of “pseudo-intention” attaches to chance, or—if we wish to avoid a negative formulation—that chance can “create” meaningful arrangements that look as if a causal intention had been at work. But that is precisely what I mean by “synchronicity,” and what I wanted to demonstrate in the report on my astrological experiment. Naturally I did not embark on the experiment for the purpose of achieving, or in anticipation of, this unexpected result, which no one could have foreseen; I was only curious to find out what sort of numbers would turn up in an investigation of this kind. This wish seemed suspicious not only to certain astrologers but also to my friendly mathematical adviser, who saw fit to warn me against thinking that my maximal figures would be a proof of the astrological thesis. Neither before nor afterwards was there any thought of such proof, besides which my experiment was arranged in a way most unsuited to that purpose, as my astrological critics had already pointed out.
[1182] Since most people believe that numbers have been invented or thought out by man, and are therefore nothing but concepts of quantities, containing nothing that was not previously put into them by the human intellect, it was naturally very difficult for me to put my question in any other form. But it is equally possible that numbers were found or discovered. In that case they are not only concepts but something more—autonomous entities which somehow contain more than just quantities. Unlike concepts they are based not on any psychic assumption but on the quality of being themselves, on a “so-ness” that cannot be expressed by an intellectual concept. Under these circumstances they might easily be endowed with qualities that have still to be discovered. Also one could, as with all autonomous beings, raise the question of their behaviour; for instance one could ask what numbers do when they are intended to express something as archetypal as astrology. For astrology is the last remnant, now applied to the stars, of that fateful assemblage of gods whose numinosity can still be felt despite the critical procedures of our scientific age. In no previous age, however “superstitious,” was astrology so widespread and so highly esteemed as it is today.
[1183] I must confess that I incline to the view that numbers were as much found as invented, and that in consequence they possess a relative autonomy analogous to that of the archetypes. They would then have, in common with the latter, the quality of being pre-existent to consciousness, and hence, on occasion, of conditioning it rather than being conditioned by it. The archetypes too, as a priori forms of representation, are as much found as invented: they are discovered inasmuch as one did not know of their unconscious autonomous existence, and invented by the mind inasmuch as their presence was inferred from analogous representational structures. Accordingly it would seem that natural numbers must possess an archetypal character. If that is so, then not only would certain numbers have a relation to and an effect on certain archetypes, but the reverse would also be true. The first case is equivalent to number magic, but the second is equivalent to my question whether numbers, in conjunction with the numinous assemblage of gods which the horoscope represents, would show a tendency to behave in a special way.
[1184] All reasonable people, especially mathematicians, are acutely concerned with the question of what we can do by means of numbers. Only a few devote any attention to the question of what, in so far as they are autonomous, numbers do in themselves. The question sounds so absurd that one hardly dares to utter it in decent intellectual society. I could not predict what result my scandalous statistics would show. I had to wait and see. And as a matter of fact my figures behaved in so obliging a fashion that an astrologer can probably appreciate them far better than a mathematician. Owing to their excessively strict adherence to reason, mathematicians seem unable to see beyond the fact that in each separate case my result has too great a probability to prove anything about astrology. Of course it doesn’t, because it was never intended to do any such thing, and I never for a moment believed that the maximum, falling each time on a moon conjunction, represented a so-called significant figure. Yet in spite of this critical attitude a number of mistakes were made in working out and computing the statistics, which all without exception contrived to bring about the most favourable possible result for astrology. As though to punish him for his well-meaning warning, the worst mistake of all fell to the lot of my mathematician, who at first calculated far too small a probability for the individual maxima, and was thus unwittingly deceived by the unconscious in the interests of astrological prestige.
[1185] Such lapses can easily be explained by a secret support for astrology in face of the violently prejudiced attitude of the conscious mind. But this explanation does not suffice in the case of the extremely significant over-all result, which with the help of quite fortuitous numbers produced the picture of the classical marriage tradition in astrology, namely the conjunction of the moon with the three principal positions of the horoscope, when there were 47 other possibilities to choose from. Tradition since the time of Ptolemy predicts that the moon conjunction with the sun or moon of the partner is the marriage characteristic. Because of its position in the horoscope, the ascendant has just as much importance as the sun and moon. In view of this tradition one could not have wished for a better result. The figure giving the probability of this predicted concurrence, unlike the first-obtained maximum of 10%, is indeed highly significant and deserves emphasizing, although we are no more able to account for its occurrence and for its apparent meaningfulness than we can account for the results of Rhine’s experiments, which prove the existence of a perception independent of the space-time barrier.
[1186] Naturally I do not think that this experiment or any other report on happenings of this kind proves anything; it merely points to something that even science can no longer overlook—namely, that its truths are in essence statistical and are therefore not absolute. Hence there is in nature a background of acausality, freedom, and meaningfulness which behaves complementarily to determinism, mechanism, and meaninglessness; and it is to be assumed that such phenomena are observable. Owing to their peculiar nature, however, they will hardly be prevailed upon to lay aside the chance character that makes them so questionable. If they did this they would no longer be what they are—acausal, undetermined, meaningful.4
[1187] [Pure causality is only meaningful when used for the creation and functioning of an efficient instrument or machine by an intelligence standing outside this process and independent of it. A self-running process that operates entirely by its own causality, i.e., by absolute necessity, is meaningless. One of my critics accuses me of having too rigid a conception of causality. He has obviously not considered that if cause and effect were not necessarily5 connected there would hardly be any meaning in speaking of causality at all. My critic makes the same mistake as the famous scientist6 who refuses to believe that God played dice when he created the world. He fails to see that if God did not play dice he had no choice but to create a (from the human point of view) meaningless machine. Since this question involves a transcendental judgment there can be no final answer to it, only a paradoxical one. Meaning arises not from causality but from freedom, i.e., from acausality.
[1188] [Modern physics has deprived causality of its axiomatic character. Thus, when we explain natural events we do so by means of an instrument which is not quite reliable. Hence an element of uncertainty always attaches to our judgment, because—theoretically, at least—we might always be dealing with an exception to the rule which can only be registered negatively by the statistical method. No matter how small this chance is, it nevertheless exists. Since causality is our only means of explanation and since it is only relatively valid, we explain the world by applying causality in a paradoxical way, both positively and negatively: A is the cause of B and possibly not. The negation can be omitted in the great majority of cases. But it is my contention that it cannot be omitted in the case of phenomena which are relatively independent of space and time. As the time-factor is indispensable to the concept of causality, one cannot speak of causality in a case where the time-factor is eliminated (as in precognition). Statistical truth leaves a gap open for acausal phenomena. And since our causalistic explanation of nature contains the possibility of its own negation, it belongs to the category of transcendental judgments, which are paradoxical or antinomian. That is so because nature is still beyond us and because science gives us only an average picture of the world, but not a true one. If human society consisted of average individuals only, it would be a sad sight indeed.]
[1189] From a rational point of view an experiment like the one I conducted is completely valueless, for the oftener it is repeated the more probable becomes its lack of results. But that this is also not so is proved by the very old tradition, which would hardly have come about had not these “lucky hits” often happened in the past. They behave like Rhine’s results: they are exceedingly improbable, and yet they happen so persistently that they even compel us to criticize the foundations of our probability calculus, or at least its applicability to certain kinds of material.
[1190] When analysing unconscious processes I often had occasion to observe synchronistic or ESP phenomena, and I therefore turned my attention to the psychic conditions underlying them. I believe I have found that they nearly always occur in the region of archetypal constellations, that is, in situations which have either activated an archetype or were evoked by the autonomous activity of an archetype. It is these observations which led me to the idea of getting the combination of archetypes found in astrology to give a quantitatively measurable answer. In this I succeeded, as the result shows; indeed one could say that the organizing factor responded with enthusiasm to my prompting. The reader must pardon this anthropomorphism, which I know positively invites misinterpretation; it fits in excellently well with the psychological facts and aptly describes the emotional background from which synchronistic phenomena emerge.
[1191] I am aware that I ought at this point to discuss the psychology of the archetype, but this has been done so often and in such detail elsewhere7 that I do not wish to repeat myself now.
[1192] I am also aware of the enormous impression of improbability made by events of this kind, and that their comparative rarity does not make them any more probable. The statistical method therefore excludes them, as they do not belong to the average run of events.