2.1 Munich Project

In spite of numerous efforts in the past, no one succeeded in either generally refuting or clearly proving the relevancy of the dowsing method. In view of this situation, characterized by wide-spread confusion, a new and much more extended scientific study was called for. After intensive and interdisciplinary consultations beginning in 1983, consent was attained to promote an appropriate research project; this was finally started in 1987 and came to end in May 1989. Characteristic results will be succinctly reported hereafter. The details concerning this unparalleled project and its surrounding issues may be found in the comprehensive and explicit project report published in the summer of 1989 [2].

Numerous scientists from different departments of several research institutes in the Munich area participated in that research project. Planning, implementation and evaluation of the experiments were subject to the control of many experts who had been appointed by the supporting Federal Agency, the Bundesforschungsministerium (BMFT) in Bonn. Because of the pronounced interest of the scientific community and the public, the entire project was carried out with peculiarly severe controls and conditions, far above the general and usual standard of research projects. Subsequently, severe criteria were implemented for all the performed experiments, in order to allow as little doubt as possible with respect to the validity of the achieved results.

Two different types of experiments were implemented in order to investigate location-dependent reactions of test persons. The results show with very high statistically founded certainty that such reactions exist; they defy any clarification by traditional and classical means and, as a consequence, speak in favor of the existence of the long debated dowsing phenomenon.

Experiment Type 1: Localization of Artificial Objects

The set-up of this rigorously controlled double-blind experiment took place on two floors of a barn. On the ground floor, a position-adjustable arrangement for the simulation of artificial "stimulations" in the form of different pipes had been installed. On the floor above, the test persons had to pass along the 10 m course; utilizing dowsing techniques, they had to localize the respective position of the pipe below on the ground floor. After each individual pass, the position of the pipe was changed by means of a random number generator and the experiment was repeated at least ten times.

In 900 individual tests arranged in 107 series, undertaken by 43 persons, the performance of the dowsers varied from pure chance to very high significance. The success rate within an individual experiment was disappointingly low. It became obvious that such an experiment to detect relatively small objects by means of dowsing, and carried out in quite an academic way, will be of no direct practical use. Nevertheless, it turned out that the number of highly significant experiment series was three times higher, and the number of significant series twice as high as what could be expected from pure chance expectation. In its entirety, the obtained results could only be produced by a chance with a probability of 1:1000; as a consequence, the data obviously supports the existence of the investigated dowsing phenomenon.

Fig. 5. Schematic description of the barn experiment. In this rigorously controlled double-blind test, an assistant H places a pipe R on the ground floor of the 2-story building at a position determined by means of a random number generator. The supervisor guides the tested person Vp to the 10 m long test course situated on the floor above; by means of the dowsing technique, the person has to find the location of the pipe unknown to her. During the tests, the supervisor observes the person from his control cabin K. A test series consists of 10 runs of the described kind.

Fig. 6. Result of a barn experiment with GTZ engineer and dowser Hans Schrter. The actual pipe positions and dowsing points are indicated by full circles and arrows, respectively. Although the direct success rate of the 10 individual runs of the shown series did not even reach 50%, the entire result is highly significant and can be reached by chance only with a probability of 1:500.

Among the 47 tested persons, the most successful one turned out to be the GTZ expert, Hans Schröter — a most surprising and suggestive fact. Although he had never before tried such a test arrangement, he reached a highly significant result in four different experiment series, each with 10 individual runs; this result could be obtained by pure chance only with a probability of 1:1700. Since the experiments were designed in a "fool-proof" manner and even the most critical skeptics could not detect flaws in the arrangement, one must attribute high confidence to the results and the associated interpretation.

It is noteworthy that comparably significant results for such kinds of experiment with artificial pipes have already been reported by at least three independent research groups (see [2] for a detailed discussion).

Experiment Type 2: Detection of Natural Sites

For many dowsers, physiological stimulation may be sensed much more effectively at appropriately selected places in the field than at locations associated with additional artificial objects. That could be ascertained by means a so-called walk-way experiment, which operates on the following basis: appropriately blind-folded test persons walk repeatedly along a given test course in the field and try to locate the same supposed reaction zone over and over again, whereby conventional information transfer is excluded to a high degree.

The experiments have been carried out under compliance with a number of precautions. These concerned the elimination of visual and acoustic information, the avoidance of wind, temperature and odor effects, the elimination of orientation possibilities through the soil constitution around the walk-way, as well as a local disorientation of the person before each experiment and the variation of the starting-point on the walk-way. These measures should help to run the experiment as double-blind as possible and to make it free of conceivable flaws (see detailed discussion in [2]).

Altogether, 40 persons were tested and about 3000 individual experiments were carried out and evaluated. It was revealed that the major part of the tested persons selected for that type of experiment did not produce statistically significant result. Once more, the mean success rate for an individual experiments was found to be rather low, although it was clearly better than for the barn experiment.

Fig. 7. Schematic view of the walk-way experiment. The plank was especially prepared for that purpose (e.g. the side planks were inclined to allow unaided walking; the entire plank was covered with a carpet to prevent position identification). Here, the plank is 10.8 m long and covers the field area where dowsers are supposed to exhibit reactions (dashed area B). The blind folded tested person is requested to walk along the plank (always in one and the same direction shown by the arrow) and to indicate precisely the reaction zone which was unknown to her. These runs were frequently repeated; before each new run, the starting-point was varied by means of a random number generator (within the indicated range A). The dowser always completes a distance of the same length (dotted area).

Fig. 8 Results from 40 individual runs of a walk-way experiment carried out on a 13.5 m long plank. The distribution of reaction points corresponds to pure chance and is statistically insignificant. The person did not sense any particular zone and, therefore, did not exhibit any dowsing ability.

Fig. 9. Result of 40 individual runs carried out on a 13.5 m plank. The distribution of reaction points as a function of location is locally very concentrated and statistically highly significant. The obtained result could be reached by chance with a probability of only 1 : 100,000. The performance of this dowser could not be explained by normal senses and, therefore, points to dowsing abilities.

Nevertheless, it is most important to note that 13 of the 40 dowsers obtained significant results, and 8 of them even highly significant results. Moreover, 7 of the tested persons were able to reproduce these successful runs. In its entirety, the results could be obtained by pure chance only with a probability of less than 1:1,000,000. Again, as already revealed in the barn experiment, this points to the existence of the phenomenon of locating certain sites by dowsing effects.

Total Evaluation of the Munich Project

When all results are evaluated together, they turn out to be highly significant and robust, i.e. neither the used evaluation procedure nor arbitrary elimination of test series or test persons would lower the total significance. The decisive test results can be summarized as follows:

• In the conducted test series the rate of success of average dowsers was found to be bad and could hardly or not at all be distinguished from pure chance.
• In certain test species some dowsers exhibited a sufficiently high success rate which could hardly or not at all be ascribed to or explained by pure chance.
• The performance of the entire pool of tested persons was highly significant and, therefore, could not be attributed to the pure chance hypothesis; furthermore, the observations could not be explained by conventional effects.

The value of this study is given by its completeness and scientific care; for the first time in history, a number of statistically reliable data of extensive proportion could be collected. None of the former studies displays a comparable power of statements. The central issue of the study, though, must be seen in an examination of the existence of the supposed phenomenon. For this reason, no attempt could be made to establish the cause of the observed effects. Subsequent studies become of primary importance and should involve, above all, investigations of questions concerning earth sciences and physiology. Limited pilot studies of the former kind are already under way (see part 1) and will also be reported in the following section.