"Brain" will probe the Bible




h EDP order






CSIRO Problem In Computer Network




The Sydney Sun-Herald, 22 August, 1954






By a staff reporter in New York

IN A FEW weeks, a giant electronic brain which works out problems of death-dealing by the U.S. Air Forces will be switched to the task of finding out winch of 311 biblical manuscripts appear to be the most authentic. How much of "the greatest story ever told" - the story of Jesus - can be relied upon as absolutely true?

The Bibles that Christians read, whether the King James or Douai version or one or other of the currently popular "modern dress" versions, are all only copies. Most of them are copies of copies. And some are copies of copies of copies of copies. Translators were human. They erred. They were, too, men of deep faith and it is entirely conceivable that where faith and the apparent literal meaning clashed they sometimes chose faith.

How true is the Bible?

The question is vital to the clergy. It strengthens their hand in the cure of souls. It cases their task in debate with unbelievers. They have sought answers to it for centuries and are seeking still.

Now a clergyman and Greek scholar in America has enlisted a new ally in his hunt for truth: Harvard University's giant electronic brain known as "Mark IV".

The fantastically complicated mechanical computer which usually works on abstruse problems for the U.S. Air Forces - but which recently helped a businessman decide which of his factories was best for the canning of pea soup - is working now on problems set by the Rev. John W. Ellison, rector of the Episcopal Church of the Epiphany at Winchester, Massachusetts.

There Are 4,600 Known Copies

Mr. Ellison himself has laboured for years at the difficult task of comparing different manuscripts of the Greek New Testament.

All of the 4,600 known versions are copies and few are alike. The scribes who copied them added words for euphony, or dropped words they thought had little importance. They switched spellings to fit their time and place. They even changed meanings, to bring them up to date.

And, of course, they made mistakes.

Mr. Ellison has spent his adult life trying to find out what variations went in 'families'. If this can be done, it will show scholars which groups of manuscripts were copied from which.

In time, the basic problem is to discover which copies of copies are the closest in text to the earliest manuscripts.

Mr. Ellison chose 311 manuscripts for submission to the mechanical brain and limited his questions to four chapters of the Gospel according to St. Luke.

I asked him why he did so.

Somewhat sadly, be replied, "Well, I have only a low priority on the electronic brain, a very low one. There are lots of people more important than I, and apparently lots of projects more important than Biblical study. The Air Forces take up most of the machine's time. I don't expect I shall get on it again for months."

Scattered Over the World

I asked how it was possible for him to have had access to 311 of the world's rarest documents, and whether he had not had to do a tremendous amount of travelling to see them, he said, "I certainly should have had to do so, had I seen them all myself. But I haven't.

"They are scattered all over the world. But happily there is a group of scholars who have been collecting microfilm copies of the manuscripts".

I asked Mr. Ellison why he had selected St. Luke and which four chapters.

"I chose Chapter Two", he said, "because it is one chapter that has absolutely no parallel in any of the other Gospels".

This is the well-known chapter beginning: "And it came to pass in those days that there went out a decree from Caesar Augustus, that all the world should be taxed . . . And Joseph also sent up from Galilee . . to be taxed with Mary his espoused wife, being great with child."

"Chapter Three I chose," said Mr. Ellison "in order to give a fairly broad study of proper names"; and Chapter Nine, "because it has a wide diversity of stories.

"In the early centuries of Christianity, lessons might be read from one text in Church and then, next day, from a different one. This is a situation where error might arise; it is a theory held by four or five reputable scholars. However, it has never been proved one way or the other. Now it will be. The mechanical brain will prove or disprove it once and for all.

"Lastly I picked 15 verses from Chapter 10 of St. Luke. This was because someone else had already gathered the information and had passed it over to me."

In the course of his plodding through the selected 311 versions of the Gospel according to St. Luke he discovered 2,000 variations in only two chapters. There were more than 400 in only 15 verses. The entire Gospel, he believes, probably contains more than 100,000 differences as between one version and another.

"It was evident," he said, "that nobody could make much sense out of all this in the course of one human lifetime, that is why I decided to go to the mechanical brain,

I asked if he could illustrate the kind of differences he found by quoting a few examples.

"It is rather abstruse stuff for the layman," he said, "and also for quite a proportion of the clergy. Stuff for scholars as you might say.

"The differences fall principally into three classifications: spelling variations, additions of a word or of several words, something italicise in this version or the other, and omissions of words or phrases as between one translation and another or one copy and another.

First Reduced to Figures

"Basically the problem we have set the mechanical brain is the determination of which copies, or copies of copies are the closest in text to the original. Closeness in time has little to do with it, though that can in some cases be fairly easily determined, since obviously one scribe might go back beyond the version nearest to his day in time, to an earlier one that he preferred, so that the ‘later’ might well be the earlier.

When Mr. Ellison present his problem to the scientists and asked whether it was one appropriate to Mark IV, they assured him that it was; in fact it would be "relatively simple" for the mechanical brain.

However,  it must first be translated into such mathematical terms as the brain can "understand", according to a rigid formula they provided.

This task in itself proved laborious," he said. "Every variation in every manuscript had to be described by the terms of a numerical code."