Lost Languages: The enigma of the world鈥檚 undeciphered scripts by Andrew Robinson, McGraw-Hill, ISBN 0071357432 Reviewed by Ross Anderson
MOST of us think of cryptanalysis as the preserve of the information security engineer. But it is also sometimes used for a very different application 鈥� archaeology.
Most will have heard of Jean-Fran莽ois Champollion鈥檚 solution of hieroglyphic writing, and many will be aware of Michael Ventris鈥檚 unravelling of the Linear B scripts used by the Minoans of ancient Crete. But these are only two tales out of many. We can now read several other Near Eastern scripts, and progress with Mayan writing since the 1970s has revealed that the Americas, too, once spawned a literate civilisation.
Advertisement
There are still many puzzles offering immortality to whoever can hack them, but people trained in 鈥減roper鈥� cryptology have so far made little impact on archaeological decipherment.
There are many reasons for this, but insofar as cryptographers are simply unaware of the challenges, Andrew Robinson鈥檚 The Lost Languages should go some way to help.
He lays the groundwork with his stories of how Egyptian, Minoan and Mayan writing were deciphered, then sets out the challenges that remain.
Sometimes, as with Etruscan, we can read the script but have little knowledge of the language. In other cases, such as the Rongorongo script of Easter Island, we know the language but have no idea how to map the symbols to it.
For the Indus Valley script, used at Mohenjo-daro in Pakistan 4000 years ago, we know neither the alphabet nor the language, as Robinson points out. And in all the remaining unsolved cases, we have only a small quantity of text to work with.
This does not make the task hopeless. In most cases, happily, archaeologists are steadily turning up more inscriptions. So the corpus of raw material for cryptanalysis is gradually increasing.
Success, as with previous decipherments, is likely to be a combination of skill and luck: skill at the analytic art, combined with the luck to tackle a script at just the point where enough material has become available to make the task possible.
And it is not clear that the experts have an advantage. Champollion, Ventris and Yuri Knorozov (who made the initial break into Mayan) were all outsiders with the gall to challenge the incorrect assumptions held by the archaeologists of the day. Their work is now accepted as a valuable contribution.
Meanwhile, the mainstream crypto community 鈥� whose stamping ground is information security鈥� is convulsed regularly with ethical debates. During the mid-1980s, it was whether crypto was a liberating force enabling people to protect their privacy, or a sinister technology that would help gangsters evade police surveillance. Now we know it鈥檚 neither.
Instead we are worried about whether it will be used by companies, such as Microsoft and Disney, to build better monopolies. The jury is still out on that one.
We could transcend this debate nicely if one of the tools we had developed for computer security purposes turned out to be just what was needed to break the silence of a great ancient civilization.