Before the invention of printing technology by Johannes Gutenberg in the mid-15th century, manuscripts survived much like gossip in a game of telephone -- depending on scribes to faithfully reproduce the works, but changing ever so slightly each time they were recopied. 

By analyzing this endless process of repetition, a Cornell University paleontologist concludes that many more hand-copied manuscripts probably survived the Middle Ages than previously thought. The reproduction of pre-Gutenberg manuscripts, he has found, was much like the replication of biological organisms, since both must be painstakingly copied from parent templates. In order for a text to be copied, one had to already exist, and the more copies in existence, the more copies that could be transcribed.

"I am treating manuscripts as though they are individuals in a population," says John Cisne, a professor of earth and atmospheric sciences. "To have more organisms, you need one to start off with and then the more there are to reproduce, the faster the population can grow." The importance of this discovery, he reports in the latest issue of the journal Science (Feb. 25, 2005), is that it helps researchers understand how science and culture survived the Dark Ages.

Cisne addressed the similar "growth" patterns of Medieval manuscripts and biological organisms by applying mathematical models developed for population biologists. He assumed that each hand-copied text has a probability of "giving birth" (being copied), and a probability of "death" (being destroyed).

And just as with organisms, he says, the growth of a copied text reaches a limit. "Eventually for texts, you run out of demand for the manuscripts, just as a population of organisms eventually runs out of food; you reach a carrying capacity or limit."

Other limits to a manuscript's survival include how many libraries there were to protect it, how many people wanted to read it, and whether the text was endangered by fire and raids in times of war.According to Cisne, "objectionable" texts, such as one sect's version of the Bible as viewed by another, could have had less chance of survival and thus don't fit the model. 

In particular, technical manuscripts, such as those on arithmetic, seem to fit population models very well, due to their ideological neutrality. For this reason, to test his model Cisne used the Venerable Bede's (AD 673-735) De Temporum Ratione , a standard arithmetic text copied repeatedly between the 8th and 16th centuries. Three of Bede's other technical works were used for comparison. The model works well in all cases, Cisne says.

In contrast, one of Bede's nontechnical works on religion shows a very different pattern of "growth," being influenced by the ups and downs of war and politics throughout history.

Cisne concludes that once a manuscript made it from papyrus (a reed cut into strips) to parchment (treated leather), the probability of survival was much higher than previously thought.

He explains, "Many pilfered manuscripts were probably resold for profit rather than destroyed. Under these assumptions, it is likely that many more manuscripts survived late antiquity and the Middle Ages."

The research was born out of Cisne's view of old manuscripts as "a new kind of fossil." "A lot more information can be squeezed out of manuscripts when you think of them as fossils," he says. Paleographers can follow manuscripts from one monastery or scribe to the next by analyzing nuances in language and handwriting, providing much sociological and historical information, he says. 

Reported and written by Sarah Davidson, a Cornell News Office science writer intern.