One of the most useful features of computers is in data examination. From the simple level of being able to count letters, numbers, duplicate of words, how many of a symbol exist, how many occurrences of an instance, percentages of occurrence etc etc to much more complicated routines which are limited only by lack of imagination or inexperience with being able to write routines.
P&W are fortunate to have the P part as a highly gifted and talented JAVA computer Engineer with a background in mathematical extraction and accounting. This has enabled us to examine in detail any body of work that contains an analysis module and insert applied criteria to it and see where the database differs or deviates from the stated purpose mathematically. In this way we can rectify non compliance of data when following a ‘rule’ or examine the reason for its non compliance and ‘excuse’ it and still keep it inserted within the routine.
During our testing phase of the T.P.B. as inputted data, we were able to extract information on so many levels. The most amazing thing was that we could lay out all the work based on percentages of occurrence. How many times remedy X was used overall, Which was the most used remedy, Which was the least, which remedy occurred most in an ailment concerning a particular location etc etc.
We made several models to test different approaches, we made a “Polarities”* module and trialed it in the program for nearly a year. We examined the resulting criteria against prescriptions made using the standard procedure and concluded that mathematically and in real terms, it was misleading. Being aware that a program was only as good as the data inputted, I personally checked against a German version of a Polarities based program with 10 of my cases, and 8 out of ten results it came to the same suggestion for remedy choice. I concluded that our program was following a correct criteria for analysis and as such, the results were not as useful as suggested. As you can see here, we left in the link to the Polarities* module in our preferences/setup menu (but removed the module) in case we ever decide to return to the module later with a fresh criteria for making it useful.
We found that in the T.P.B. that Boenninghausen had in several of the chapters, placed rubrics with different symptom descriptions but contained the same remedies for selection within. Vladimir ran comparisons based on ‘exact’ match criteria, and also on a variety of percentages of similar found in the remedy lists of each. In this way were able to cross reference EVERY occurrence of rubrics that contained the same data, for the purpose of preventing insertion into a case analysis so as to not give a false repertorisation result numerically.
The original intention of this work was to find rubrics we could merge if the criteria was the same, but in the event, we took the decision to NOT do that as some rubrics contained additional remedies or a variance of the grade assigned and we did not feel qualified to make a judgement on reassigning the indicated grade up or down. Whilst mathematically, an even across the board change of a remedy within a rubric to the higher grade would appear to be a valid option, the fact remains that when Boenninghausen split a symptom into components, he assigned the same numerical grade to all the components. We did not feel that is was of benefit to alter this arrangement arbitrarily in one rubric component only.
However, we do indicate in the program where a similar rubric could be found for checking. We also indicate where a smaller rubric is contained in its entirety within a larger rubric if expansion of the symptom is required.
We also ran into a problem with the term Tetters. In discussions with others, and in research, we were unable to find an exact meaning for the term. Our comparison tests showed where we COULD merge certain rubrics with a rubric for eruptions of a similar description, AND we understand that the smaller rubric is contained in entirety in the larger one, however, Boenninghausen differentiated the terms. We do not feel qualified or feel we have enough information to merge the rubrics for Tetters with eruptions (where additional medicines are present) so we erred on the side of caution and leave it to the practitioners to make that decision. We continually look for explanations of 1800’s terms for absolute definites and when we can justify in our research that we can merge a rubric for clarity, we will do so. It does not mean we are correct to keep the rubrics distinct and separate, it means we are unsure and cautious and accepting of our lack of knowledge on the matter.
One other area we continually work on is making a Materia Medica direct symptom search repertory. Back in the mid 90’s I was working on a concept of bypassing the Repertory altogether, and searching in the Materia medica’s only with a computer for a symptom detection and analysis.
This is one of the main reasons that I compared the 1846 edition of Boenninghausens rubrics against the Materia Medica for accuracy of comparison. (It was a long and laborious job which involved a lot of photocopying and typing. During the Hurricanes in Florida, I lost my residence and my computer and most of my research photocopies and typed notes. 3 years of work gone in an instant. Fortunately later I was able to replace the copies via Google Books).
Later Vladimir and I, in discussing the concept, were able to use our Materia Medica databases and extract symptoms direct via reference to the Materia Medica and extract a list of medicines that had the collected total of the symptoms inputted. Most homoeopathic computer programs will search for symptoms in medicines, but we were aiming for a totality of several symptoms to give the answer directly cutting out a repertory. For various reasons, it does not work. We believe Boenninghausen knew this. That is why his repertory is a synthetic approach based on an amazing technical understanding of Hahnemanns logic and observation. When Vladimir has time, his mind returns to the complexity of the issue and he mulls over different approaches to the extraction of information to give an accurate indication of which medicine is required. So far, it does not seem possible using the criteria we have formulated.
What we did get out of this exercise was an understanding of each and every rubric singularly and collectively, ie a joining together of the complete symptom via combination of rubrics. We matched the individual symptom in the Materia Medica with a grouping of 2 or 3 rubrics that described that symptom collectively, and were sure of the accuracy of our choices because the GRADE for each rubric would be the same.
We have so many projects that we would like to do with the information that we have, but time, money and the reality that most practitioners prefer the inaccuracy of other repertories, constrain our enthusiasm.We have to be content that the work we have done is as accurate as we can make it as a working and totally useful representation of the 1846 Boenninghausen Repertory for the practitioner. We have spent a lot of years individually and together on this project. We wanted something we could trust for our patients.This is Boenninghausens work. This is Hahnemanns approach to case taking. This has been an interesting journey and one we enjoyed, but would not want to undertake again. Its done, its updated and it works well.
*approach suggested by Klaus Henning Gypser