Research is not a new concept for serious Jyotishis. Journals like The Astrological Magazine and many others had been supporting and publishing research articles on novel aspects of Jyotish techniques as well as re-examining existing techniques for many decades now. Given the enormous complexity of the Jyotish delineative process which incorporates an analysis of myriads of factors, which are then synthesized into a meaningful reading, research can be time consuming and difficult. Even as recently as a decade or so ago, there were not many Jyotishis who had ready access to tools such as computers and software. Even the few software packages that were available often had quirks, and at times outright errors particularly in the area of newly programmed in features and techniques. Partly due to the internet, and to a massive renewal of interest, many otherwise hidden techniques have come to new light or have newly been proposed and found a place in popular software over the years. Not all of these have been adequately tested, or validated to produce reliable results and often the user is relegated to the role of being a paying-beta tester, at least during the first few versions of the software. Response to complaints and suggestions have varied from one software company to another. Some incorporate suggestions right away, while others never respond back or make the change!
The research features in most Jyotish software are not powerful or suitable for examining a large numbers of horoscopes for conducting group analysis. Standalone research software has been available for western astrology (e.g., Jigsaw) but not for vedic astrology. The popular Parashara=s Light from Geovision lets one look at a group of charts in a .QCK format file and generates some statistics for factors such as aspects, dignities, shadbala strengths attained by planets in a group, or searching for rules such as number of charts with Jupiter aspecting the moon in a group. A commendable baby step, at best.
In recent years, The Astrodatabank software from Mark Donough has begun to add Jyotish factors and attributes to its research capabilities. Astrodatabank is a Microsoft Access-based database which contains birth data of varying qualities essentially incorporating the famous Rodden Database. Late Lois Rodden, a Canadian astrologer from Saskatchewan had been a well-known pioneer in the area of birth data collection. The results of years of painstaking work by herself and her group has been available separately and as part of the Astrodatabank for some time. The database now contains greater than 28,000 birth data, many with biographical details and categorized in hundreds of different ways. It allows one to view the charts from different periods of time, different vocational categories, personality types, diseases, the list goes on. New data gets added all the time and the database is updated periodically by going online, through the internet. Birth data of celebrities coexist with thousands of data for ordinary regular folks, the nativities that most Jyotishis typically would give readings to.
With essentially over a million or more pieces of information available in the database, given all the categories, tags and flags for each record, it is absolutely essential that there exist powerful search, sort and filtering capabilities B such as are provided by MS Access. If you already have MS Access on your computer, you can use that; for the rest of us a run-time access utility is included with the software.
The printed and online documentation is pretty good, a bit concise in places, but detailed in others and the program, though initially a bit overwhelming due to its enormous lists of options and its capabilities is well organized and after some familiarity is very easy to use. Several views allow one to see the database in different ways. The site at www.astrodatabank.com is loaded with information, birth data, discussions, examples, and helpful material of all kind. The price of the program is comparable to most Jyotish programs and though not exactly inexpensive, represents good value for individuals who are seriously interested in conducting research as well as interested students because it provides one to experiment with data and real charts and the value of that kind of practical learning cannot be overstated. Subsequent upgrades tend to cost considerably less than the original price.
In general, the examination of astrological techniques can be done qualitatively or quantitatively. Does a certain yoga work or not would be an example of qualitative examination. A research question such as, A) Does the gajkeshari yoga formed by the mutual angular disposition of moon and Jupiter work in charts, as per classical dicta or otherwise@, would require examination of a variety of charts with this combination and biographical details available, with ease. Similar examples could be a certain arishta and testing its correlation with a physical defect or disease. While important, such research would not necessarily tell us if a certain form of the yoga or arishta is more effective or less, from the other forms of the same yoga or for that matter in comparison with another combination. There are a large number of individuals who do not test things this way and rely heavily on anecdotal evidence or sometimes just hearsay and personal intuitive guesses which they then call logical interpretation. This is not the best way to remember or practice Jyotish or to demonstrate its efficacy to others.
Research In Astrodatabank:
On one internet Jyotish forum, someone once mentioned emphatically that a certain combination leads to sexual excesses. In the past, no one would have the time or patience to actually pore over hundreds of charts and see if this is really the case. With Astrodatabank, it was easy for me to quickly test it and using a subset of the birth data, with the attributes of sexual excesses and deviations, a certain proportion of this group did indeed have the combination under examination. In order to see if it is a real effect, I next compared it with a subset of individuals who were saintly and otherwise holy men. Now I realize that a few of these holy men could have had some sexual issues, but for the purpose of this quick examination, I allowed for that >contamination= because it probably would not be significant. Well, lo and behold! The said combination was approximately in four times as many charts in the >sexual= group than the saintly group. This is where research gets tricky! Since the devil is always in the details and in the re-examination B) the obvious thing to test was if there was a sampling bias involved. Sure enough, the sample size of the >sexual= group was about four times larger than the >saintly= comparison group! The factor, therefore, was as prevalent in both groups and not significant! Proportion or percentage and not the absolute prevalence was the key in this case and a red herring was avoided (or perhaps captured!). The real point is that without the database, something like this could not possibly get tested in a few minutes and the comment made by someone on a Jyotish forum would have been accepted without question and would stick in many minds as the researched >truth=! Having the database also gives confidence in what we are testing or should test when opportunity arises.
In quantitative testing, likewise, one must not assume that all factors are equally distributed. If one tests something like the distribution of lagnas, one would find that certain lagnas tend to rise more commonly than others. These would tend to be the signs that are of long ascension, and hence rise for a longer duration than the others. The long ascension signs would depend also on the hemisphere of birth and seasons during which the birth took place. Again, the bottom-line is that lagnas are neither distributed equally across the signs, nor are the sun signs or moon signs. If we are not aware of these >variabilities= that lead to sampling bias, we risk arriving at wrong or misleading conclusions.
Astrodatabank allows one to filter out data. One example of such a filter would be for determining how many individuals have the Chandra-Mangal yoga. Three of the many types of filters that Astrodatabank provides are positions, aspects and dignities. Each of these screens have several items that one can pick and choose from to design the filter. For our example, I would pick moon and mars in the aspects screen and choose yuti or conjunction (the vedic form allows for the entire sign as opposed to tighter orbs, which are also available if you so choose). If I were looking for the yoga in specific houses, I could have chosen that from the position screen. If I wanted to look at only moons that are in cancer or its exaltation sign Taurus, I could have chosen those from the dignity screen. One click on the OK button and the computer would show me the birth data of nativities with the combination. I could then go to each chart and examine the biographical data or click on the INFO button and get a table of the planetary positions and the ascendant for the chart under the cursor. Using the cursor keys, I could next scroll down to the next chart or next screen. If I need to look at the horoscopes, I could click on the CHART button and it would open the Jyotish program. It currently supports Parashara=s Light and Sri Jyoti, and several western calculation programs. More would probably get added in subsequent versions if there is public demand. One can add their personal collection of charts to the database or create new databases by importing one=s own collection of charts or creating a category for research purposes, for instance.
While filtering is quite useful, it is restrictive. With filters you are essentially weeding out unwanted criteria. However, if you wish to test several rules to see which one fits, other tools are available in Astrodatabank. For instance, if you want to test a certain association (energy-based healing ability, for example) and wish to identify all charts with sun in 3rd house or sun in exaltation in charts (not just 3rd house) B you cannot do that in one operation using filters. If you try to, the list that will come back will show you how many charts have exalted sun in the 3rd house. This is where ASTROSIGNATURES come in handy. This feature allows one to define a series of rules and when these are applied to a database, you get the counts or hits for each of those rules in the sample tested. Based on the success rate, you can then add varying weights to the rules (essentially deciding that one rule is quantitatively more important than the other for the given effect. One can play with and fine tune the weights and see the effects on the statistics generated. In our example, using astrosignatures will tell us how many individuals had sun in 3rd, how many had the sun exalted anywhere in the horoscope, in a given collection (database) of charts.
The next level of examination would be to see how often a given combination is seen in horoscopes of individuals who are from the general population. These are generally called CONTROLs or the control group, since it controls or limits red-herrings or spurious effects (false positives or false negatives). Controls can come from the general population of real charts. They can be picked randomly. However, there is a practical problem. In order to do so, one would have to have a humongous database of high quality data. Another way of looking at this is that you want to find out how often the >combination= being examined is likely to arise during a period in time that is comparable to the range of birth data in the tested group. Astrodatabase uses four elements: year, month-date, time of day and elements (fire, earth, etc). It basically takes the range of birth years that your test sample contains and creates a control group within the same range so that one is not comparing a group born during the 80s with another group born in the 20s and thus having a completely different distribution of the slower moving planets, for instance. Also, it makes sure that the control group has birth times and other factors distributed and not clustered and capable of giving rise to other kinds of sampling biases. The control group is generally about 10 times larger than the test group and so has more samples within the same range adding power to the testing. All of these factors and boundaries are customizable and can be tweaked and adjusted by the knowledgeable researcher.
Once the test has been run, usually taking few seconds to a few minutes depending on the speed of your computer and the size of the group of data B a report is generated that essentially gives you the number of charts that had the rules under examination (the non-zeros) and then a score which is derived by dividing the total score by the size of the groups. The scores obtained by the control group and experimental groups will indicate the relative usefulness of the individual rules. If the scores are the same or similar in experimental and control groups, then the rule is not good for identifying the studied effect or condition, but if there is a big difference then the combination (rule) is either indicating the influence of the presence or absence of the combination in indicating the astrological effect. In our energy-healing example if we find that the sun in 3rd is seen more commonly in the healers than the controls, then it is possibly associated with the group and useful. If the combination is seen a lot more times in the controls, then perhaps sun in 3rd is less likely to be associated in the charts of healers or would be healers. Readers must note that for ease of illustration, a single factor is used here; however in proper readings, many other factors need to be considered and therefore when testing or researching, a variety of factors are used for testing and not just one or two. On the other hand, testing large numbers of rules can get confusing, therefore it is a good idea to test a batch, eliminate the less useful ones and keep on building the astrosignature patterns, fine tuning the set all the time. The software allows one to save and load libraries or lists of astrosignatures. These can be saved as separate astrosignature files with short description added, in case some of these need to be reused later.
Despite its obvious power and usefulness, there are some areas where improvements are necessary in the Astrodatabank software to make it even more useful to jyotishis, some of these have been outlined below:
One cannot test to see if a planet, say sun is the ruler of 3rd house and aspects the 10th. The rulership attribute is not included. One can get by with using a filter stating: Find the instances where Gemini is rising and sun is aspecting the 10th house (or ruler of 10th or dispositor of 10th). The obvious solution is adding the rulership tag. The provision of dispositor tag is much appreciated and used a lot by me, personally.
Let us say you want to do a search for kemadruma yogas in your collection with all the factors that are needed for a complete kemadruma (moon unassociated or unflanked by any planets, no planet in Kendra from lagna or moon) B would not be something you would be able to do easily or in one fell swoop, as they say.
Or if you want to see what happens if a ruler is in 8th from its house and its bhuktinatha is in the same nakshatra as it is (trinal relationship with some qualifiers!) then you would not be able to find that readily. The point here is to see if that dasha bhukti brings about the destruction of the said house! This is not an unusual sort of question that a vedic astrologer would ask and research about. Change the >8th= with another malefic or benefic relationship as per vedic astrology and you would be equally hamstrung. Sadly!
A common question coming from beginners that perplexes them is what happens when a neecha planet is vargottama (neecha rashi and navamsha) B is it weak or strong, with and without cancellation? A research database should be able to scour through thousands of charts and bring up those easily and smoothly. ADB does not do that easily because none of the vargas are included as tags. In some cases it can be accomplished with roundabout filters such as if mars is in cancer and in the 7th nakshatra. This will identify the charts with vargottama neecha mars but then you would have to figure out cancellation yourself, unless you create another filter which identifies the cancellation factors such as finding the placement of Jupiter in kendras from mars, ascendant or moon using the aspect screen.
The databank does not calculate charakarakas (or identifies any of the other karkas, for that matter) or padas. These are becoming very important in modern research since their recent resurgence in the Jyotish scene.
I would like tags and rules built in that connect the so called Jaimini factors (also described by Parashara) such as rashi drishtis between vimshottari dasha and bhuktinaths (and why just stop at bhukti?) or rules such as >Dasha or bhukti= of gnatikaraka who happens to be the lord of 6th and in aspect with benefics (or malefics). The creative research possibilities that exist in the framework of Jyotish are so many, that they would bring the sweat to the most astute programmer and database expert=s collective eyebrows!
The software does not look for varga positions, ashtakavarga, shadbala and many such common attributes used in Jyotish. While the wish list is long and yes — demanding, we need to keep one thing in mind. Since astrodatabank does not calculate on the fly (It would be impossibly slow if it did B imagine your Jyotish calculation software calculating a batch of thousands of charts for each search!) but uses precalculated data, with all of these additional factors to be calculated, the size of the database would become prohibitive and the searches extremely slow. But, still what it already has could be immensely benefited with the addition of charakarakas and padas, at the very least.
What it could add, also, is a simple chart display on the INFO screen. This screen shows you the planetary and ascendant longitudes in a tabular form as you scroll down a list of birth data. On the right hand side, there is unused space (as far as jyotishis are concerned) which will be perfect for a chart-in-a-box depiction. This feature should be easy to incorporate, and would not slow down anything, given modern computers.
The filter screen needs to be made more user friendly. If one wants to add a series of similar factors with each of the different planets. For instance if you are defining a filter, Aif planet or its dispositor is aspecting the tenth house or its ruler@. You would have to click on all the boxes for each of the planets. An easier way would allow one to define one planet, then copy it over eight lines and then by clicking on each line if the chosen boxes were shown on the left side of the screen, one could select the next planet and then move to the next line and so on. The process would be a lot easier then with this simple copy and edit feature.
Each time one adds new birth data, by typing or using the import facility, the astrological factors need to be re-calculated, and if you have changed the ayanamsha etc B the entire database needs to be recalculated. Given the huge size of the database, the recalculation can take several minutes to over an hour. In the screen capture (ADB calculating times) it took my computer only 4 min and 20 sec to calculate 3718 records. The exact duration will obviously depend upon how many attributes you wish to calculate. There are several layers of options. Indeed, the software can be somewhat daunting to the first time user. One should be prepared to work with it intensively before its true value can be appreciated. When the database grows large, an utility is provided to strip all data and recalculate the database from scratch. This can come handy if you use large databases of charts. Also, facilities for repairing and for backing up the databases (highly advised!) periodically are all there and work very well. The recalculating is needed each time you create a research subset, also.
Though Astrodatabank software does not incorporate any statistical tools, yet, it allows one to output research data as an MS Excel spreadsheet or as text files (fixed width columns). On my system with Excel 2003, the spreadsheet format output was not very useful due to formatting differences, however, I could import the text output into Excel very easily as >fixed width= fields and with some editing, I could do further analysis as needed. Realistically, I believe that it would not be a feature that most users would probably be using a lot, but it does need some >cleaning= up to make it useful.
So, the bottom-line! If you want to acquire a huge number of charts which come with lots of biographical information and categorized in hundreds of categories, ready for studying and research, then your utility is here. If you want to carry out research on the data, and are reasonably creative and patient, you can get a lot of mileage out of the Astrodatabank software. Be prepared to encounter a learning curve. The more intensively you work with the software, the better your abilities would be to get the software to work for you, rather than against you. That said, there would be features and attributes that you would pine for and it is your duty to inform the Astrodatabank development team what you miss and then hope that they would be able to implement it in due course. Usually if a fair number of individuals send in a similar request, then the developer is better motivated to incorporate the change. In terms of reliability and stability, there have been nearly no significant issues in my experience with this software. I sincerely recommend this databank and search utility to all serious astrologers, researchers and students of astrology and would like to end this article with a prayer to Ma Goddess:
Ma Shakti Dey, Hara Klanti Ley
Ma Buddhi Dey, Hara Bhranti Ley
Ma Bhakti Dey, Ma Shanti Dey, Ma Mukti Dey!