Blockchains are attractive because their built-in security and trust factors make them useful for almost all business interactions involving organizations and individuals. Blockchains have two basic functions. One is as a method for handling transactions involving property such as land deeds, trademarks or other assets. The second involves exchanges of data such as identities of individuals or businesses, the location of an object at a point in time or weather conditions. All interactions involving property or assets include the transfer of data as well, of course, but some blockchain use cases are informational only.
Blockchain distributed ledgers are still in their infancy but it’s becoming clear that while there will be some core similarities (trust is inherent in a blockchain in that multiple nodes agree to any change to the ledger and the data in the ledger always is encrypted), their form will adapt to their purpose. Even immutability – a core characteristic – won’t be the same in all blockchains.
There are two aspects to the immutability of records in blockchains: content and persistence. In all cases, the content of a blockchain record cannot be altered. To maintain their integrity, all blockchains use separate correction or nullification entries to adjust previous records just as accountants use adjusting entries. However, not every blockchain will retain every record forever.
Initially, unlimited persistence of blockchain entries (temporal immutability) was considered an essential element of a trusted system. Now, it’s more accurate to say that the lifespan of the underlying assets or information will dictate the temporal persistence in a blockchain. Some blockchain records will be “forgotten” when all parties to the transaction agree to delete them. This is similar to the analog world, where records retention policy is determined by the nature of the information in the record.
Guaranteed persistence in some blockchains – like those for land records, diamonds or currencies – is essential because the property or data represented in the blockchain is permanent and its ownership must be traceable over time. Very likely, all entries in unpermissioned blockchains also will be permanent because this will enhance their no-trust structures. Even allowing synchronized deletions would threaten their soundness and resistance to manipulation.
However, other types of blockchain may not require permanence of entries. For example, transactions in general purpose commerce platforms or payment systems need only persist for as long it takes all sides of the transaction to agree that the contract between them has been satisfied. All parties will retain an electronic signature certifying the transaction because that’s all that’s required for handling disputes, just as a signature on a receipt is sufficient to prove a transaction took place. The same applies to business transactions where purchase orders and invoices have been executed and accepted. The record of the transaction itself has no practical value once its terms have been satisfied and all parties have consented to its completion.
Moreover, retaining these records in a blockchain past their useful life is wasteful and burdensome for the blockchain host. When it’s operational a platform might easily record several terabytes of data every day. To paraphrase a quote about government spending misattributed to a long-dead politician: A billion bytes here, a billion bytes there and pretty soon you're talking about real money.
Protections still exist. Saved electronic signatures from the parties attest that the contract has been fulfilled. While the transaction data no longer exists in the blockchain, it is retained on either side of the interchange (by correspondent banks, for example) for the benefit of the owner. Moreover, while maintaining ownership lineage of specific defined assets such as land is essential, it’s irrelevant for fungible goods such as money issued by a monetary authority and for goods and services exchanged between businesses and individuals where traceability isn’t required.
It’s still early days for blockchain distributed ledgers so there’s much work to be done sorting out the issues that inevitably arise when new systems and processes are hatched. When they first were presented, blockchains were closely associated with cryptocurrencies. I’ve noticed that even people who understand that blockchains are more than bitcoin nonetheless have a narrow sense of the technology. It’s important, though, to recognize that as they evolve, blockchains will come in many shapes and sizes. They will share some limited, albeit crucial, characteristics (chiefly encryption, record immutability and functional persistence), but they won’t be all the same.