The word "blockchain" is used so loosely that it has nearly stopped meaning anything specific. It gets attached to tokens, to companies, to investment pitches, and to vague promises about trust. But the term has a precise technical definition, and it comes from a public, citable source rather than a marketing deck. The U.S. National Institute of Standards and Technology published a formal overview of the technology, and its opening definition is the cleanest place to start.

According to NIST, a blockchain is a distributed digital ledger of cryptographically signed transactions that are grouped into blocks. Each block is cryptographically linked to the previous one after it has been validated and has undergone a consensus decision. As new blocks are added, the older blocks become progressively more difficult to modify, which is the property NIST calls tamper resistance. New blocks are then replicated across copies of the ledger held throughout the network, so no single participant holds the only copy.

"Blockchains are distributed digital ledgers of cryptographically signed transactions that are grouped into blocks. Each block is cryptographically linked to the previous one (making it tamper evident) after validation and undergoing a consensus decision."— NIST, source

Each of those clauses is doing work. "Distributed" means the ledger is not held by one server but replicated across many nodes. "Cryptographically signed" means each transaction carries a digital signature that ties it to the account that authorized it. "Grouped into blocks" means transactions are batched rather than recorded one at a time. And "cryptographically linked" is the part that gives the chain its name: each block carries a cryptographic hash of the block before it, so the blocks form an ordered sequence in which any change to an earlier block would change its hash and break the link that the next block records.

What "tamper resistance" actually guarantees

The most over-claimed property of blockchains is immutability. NIST is careful here, and the distinction matters. The document notes that the tamper-evident and tamper-resistant design means that once a transaction is committed to the blockchain, it generally cannot be changed. But it immediately qualifies that the guarantee is only true for transactions that have already been included in a published block. Transactions that have not yet been included in a published block do not enjoy the same protection. In other words, the data structure makes after-the-fact alteration detectable; it does not make the network instantly final, and it does not vouch for whether a transaction should have been recorded in the first place.

That qualifier is why the standards document, rather than a project's homepage, is the better anchor for the definition. The protection is mechanical and bounded: it tells you that a recorded entry, once buried under later blocks, is hard to silently rewrite. It says nothing about the truth of what a transaction represents off-chain, nor about the governance of who gets to publish blocks.

The hash link, in slightly more detail

The single most load-bearing word in the definition is "linked." The mechanism behind it is the cryptographic hash function, which NIST describes as a one-way operation that takes an input of any size and produces a fixed-length output, the hash, such that even a tiny change to the input produces a completely different output and it is computationally infeasible to find two inputs that produce the same hash. Each block stores the hash of the block before it. That single stored value is what binds the blocks into an ordered chain rather than an unordered pile. If someone altered a transaction in an earlier block, the hash of that block would change, which would no longer match the value recorded in the next block, which would in turn change that block's hash, and so on down the chain. The tampering does not stay local; it propagates and becomes detectable to anyone who recomputes the hashes. This is the concrete machinery behind the abstract phrase "tamper evident," and it is why the property is evidentiary, it reveals tampering, rather than a guarantee that tampering is impossible.

Replication compounds the effect. Because copies of the ledger are held across the network, an attacker cannot simply rewrite their own copy and have it accepted; they would have to convince the network's consensus process to adopt the altered history, which is precisely what the consensus model is designed to make difficult. The combination of hash-linking and broad replication is what gives a blockchain its characteristic durability: no single party holds the only copy, and no single party can silently rewrite the shared one.

Permissionless versus permissioned

NIST also draws a line that the popular usage tends to blur. A permissionless blockchain network is one that anyone can read from and submit transactions to, with no central party deciding who may participate; these networks generally use a native cryptocurrency to reward the participants who publish protocol-conforming blocks. A permissioned blockchain network restricts who may participate and who may publish, which changes the security assumptions because the set of participants is known. Both are blockchains under the definition. The difference is who is allowed in and how the network discourages misbehavior.

The consensus decision NIST references is the procedure by which the distributed copies agree on which block comes next. Different networks use different consensus models, and those models are where most of the engineering tradeoffs live. But the core structure is constant across them: signed transactions, grouped into blocks, each block hash-linked to its predecessor, replicated across the network, and extended only after the network agrees. Everything else built on top of a blockchain, from tokens to smart contracts to settlement rails, sits above that base layer.

Holding to the standards-body definition has a practical payoff. It separates the data structure from the claims people make about it. A system that batches signed transactions into hash-linked, consensus-validated, replicated blocks is a blockchain, whatever it is used for. A system that lacks those properties is not one, no matter what its promoters call it. And the guarantee a blockchain provides, tamper-evidence for already-published transactions, is specific enough to be checked and narrow enough that it should not be mistaken for a guarantee about the real world the ledger describes.