more numerical fun

John Hobbins points to an article in the Jerusalem Post about a new book purportedly revealing startling new discoveries about truths hidden in the Hebrew text of the Bible by Professor Haim Shore from Ben-Gurion University of the Negev. There’s endless fun to be had with such mathematical legerdemain, but for now let’s just look at his comments on the solar system. From the Jerusalem Post article:

Could there be a linkage between numerical values of biblical words and certain physical properties, as demonstrated by the heraion example?” Shore asks. “In Hebrew, yareach is moon, eretz is earth, and shemesh is sun. One thing that distinguishes the three bodies is their size, expressed by the diameters. I used their diameters as listed by NASA, and plotted them on a graph, just as I did with the colors.

On the horizontal axis is the numerical value of the Hebrew word, on the vertical axis is the planetary diameters from NASA (on a log scale),” he continues. “To my astonishment, the phenomenon repeated itself. The three points aligned themselves on a straight line – an exact mathematical relationship would have given a linear correlation of ‘1,’ whereas these three points had a linear correlation of 0.999. Again I thought, ‘What an amazing coincidence!’

This all sounds remarkable, doesn’t it? It does, until you dig a little further! Here are the raw data I’ve garnered for this, including numerical values from my gematria calculator:

And indeed, if you plot these values you find that the correlation coefficient is 0.999 (to three decimal places). Now in scientific terms, a data plot with this sort of correlation would suggest that the data are very unlikely to be random.

However, 0.999 is not 1, it is not a perfect correlation, and in this case there is a lot of room for movement in the space between 0.999 and 1. That is because two of the points lie close together (the Earth and Moon are, compared to the Sun, of a relatively similar size).

Now if Hebrew is perfect, what would a correlation of 1 give us as the numerical value required of the word for “Sun”? That would be 554 — quite a distance from 640. In fact, any numerical value between about 480 and 640 would render the same or better correlation. That’s an awful lot of wiggle room!

Looking at it the other way, taking the Hebrew term for “Sun” and assuming it encodes the relative size of the sun precisely, the diameter of the Sun would need to be 6,395,712km — that is more than 4.5 times the Sun actually is. That difference certainly lies outside the range of observational error. You would notice if the Sun was that much bigger (well, you probably wouldn’t because life on Earth would quite probably not be viable).

If this is so much more than mere coincidence, why not get it right and choose a correct value instead of one which is actually so far from correct? Shore is perhaps to be congratulated for discovering a new way to make money out of finding mysterious codes in Hebrew, but like all before they really do not stand up to scrutiny.

P.S. I was playing with the Grapher application in Mac OS X and came up with this graph demonstrating just how far from perfect the correlation actually is: