Transactions are grouped into blocks, and each block carries a cryptographic fingerprint of the one before it. Changing an old record changes its fingerprint, which breaks every fingerprint after it — so history cannot be edited without redoing all the work that followed and convincing the network to accept the result. That chaining is what the word describes.
The second half is replication. Every participating node keeps its own copy and checks incoming blocks against the rules, so there is no central database to compromise and no administrator who can reverse a transaction. The cost of those properties is throughput: every node processes everything, which is why blockchains are slow and expensive compared with a conventional database.
Why can't blockchain records be changed?
Each block contains a hash of its predecessor, so altering one invalidates every block after it. Rewriting history means out-working the entire network and having the result accepted, which is impractical on a large chain.
Last updated October 17, 2026