Thank heavens for the silicon chip: A BRIEF history of data

You have a pair of bones in Africa to thank for Larry Ellison

Tally a while

By the time of the Roman Empire, tally sticks were widely used; Pliny the Elder for example, makes reference to them. Originally they were probably just used to store data but later were modified to record transactions – the so-called split-tally sticks. Take a long piece of wood and carve notches across it. Then carefully split the piece of wood down the grain. You end up with two sticks that have identical sets of notches: and since you have split it rather than sawn it, only those two precise pieces will ever fit perfectly back together. So both parties can take away a record of the data that makes up the transaction.

And suppose that the record is not of a historical transaction, but an ongoing one, such as a loan; in that case the two parties to the transaction are not symmetrical. So the split can be performed for most of the length of the stick but then cut across to one side, generating two tallies of different length. The long stick can be kept by the lender and short by the borrower.

By the time of Henry I – born in 1068 or 1069 and one of William the Conqueror's sons – tally sticks were increasingly used in the British islands. So widely were they used, in fact, that they became a form of money and were used by the Exchequer right up until the early nineteenth century. Indeed it was the burning of obsolete tally sticks that led to the unfortunate incineration of the Houses of Parliament on October 16, 1834. That fire led to the building being rebuilt as the Victorian Gothic pile that still occupies the bank of the River Thames at Westminster.

Of course, tally sticks weren’t the only way of recording data; they were (relatively closely) followed by writing and later printing. Writing arose in Mesopotamia around 5,000 years ago. In the early days, back-up was performed by teams of scribes, but later Johannes Gutenberg introduced a mechanical back-up system for written data called “printing”. If only he had got his act together 1,600 years earlier, we might have saved more from the Library of Alexandria.

Writing led to the development of more formal record-keeping, such as ledgers, that ultimately replaced tally sticks and also lead to the explosion in the numbers of the clerks so beloved of Dickens.

So, with the acquisition of writing, we have two forms of data. Some has a reasonably well-defined structure and an agreed meaning (the tally sticks and the clerk’s ledger books), while the written word is less formally structured and can be much more free-form. You can see where I am going with this: these two are what in today’s climate we’d call tabular data and Big Data.

Skip forward in time and when computers appear, we have both forms of data in play.

Initially, computers focused entirely on tabular data. With the huge benefit of hindsight we can see that the early programmers made a fundamental mistake. They failed to realise that data isn’t simply something that computers are competent with, nor is it something that is a nice adjunct to computing. Data is all that computers do.

Note the complete lack of qualification in this statement.