Blockchain is a distributed ledger technology that records transactions in cryptographically linked blocks shared across a peer-to-peer network, making the data tamper-evident without requiring a central authority. The protocol was formalized in Satoshi Nakamoto's 2008 Bitcoin whitepaper, which introduced the data structure that would become the foundation for thousands of subsequent networks. Understanding what blockchain actually is at a protocol level (rather than through financial metaphors) separates readers who can evaluate it from those who can only repeat marketing claims.
How Blockchain Works
Blockchain organizes data through a specific sequence of operations. A participant broadcasts a transaction to the peer-to-peer network. Nodes receive it, validate it against the chain's ruleset, and bundle it with other validated transactions into a candidate block. Each block carries a cryptographic hash of the block preceding it, computed using SHA-256 or an equivalent hash function. That hash is deterministic: the same input always produces the same 256-bit output, and any modification to the input produces a completely different output. The NIST Interagency Report 8202 on blockchain fundamentals formally defines this chaining property as the mechanism that makes the distributed ledger tamper-evident without central administration.
Inside each block, transactions are organized using a Merkle tree. Each transaction is hashed individually. Those hashes are then paired and hashed again, and the process repeats up a binary tree structure until a single value remains: the Merkle root. This root, stored in the block header, represents a cryptographic commitment to every transaction in the block. Changing any single transaction changes its leaf hash, which changes every node above it up the tree, which changes the Merkle root, which changes the block hash for that block. Because the next block's header stores that block hash as its "previous block hash" reference, the alteration invalidates every downstream block header. An attacker would need to recompute all subsequent block hashes faster than the live network adds new blocks, which is computationally infeasible on large networks.
Block Structure at a Glance
- Block Header
- The metadata container at the top of each block, holding the version number, timestamp, difficulty target, nonce, previous hash digest, and Merkle root.
- Previous Block digest
- A cryptographic hash of the preceding block's header; this reference creates the chain linkage that makes any historical modification detectable.
- Merkle Root
- A single hash derived from all transactions in the block via a binary Merkle tree; any change to any transaction changes the root value.
- Nonce
- A 32-bit number that miners increment repeatedly until the resulting hash meets the network's current difficulty target.
- Transaction List
- The ordered set of validated transactions included in the block, each represented as a leaf node in the Merkle tree.
Consensus Mechanisms: How Nodes Agree
A consensus mechanism is the protocol rule set by which nodes that do not trust each other agree on which candidate block gets appended to the chain, without delegating that decision to any central arbiter. The two dominant designs differ in what they require participants to expend, and they produce different transaction verification speeds and security properties.
Proof of work requires miners to compete in a computation race. Each miner repeatedly hashes a candidate block header with different nonce values until the resulting block identifier falls below a numeric threshold, expressed as a required number of leading zeroes. The first miner to find a valid nonce broadcasts the block; the network validates the block fingerprint digest instantly (hashing is cheap to verify, expensive to produce) and appends it to the chain. The miner receives a block reward denominated in the chain's native token. Bitcoin uses proof of work. The security property follows from the computational cost: reversing any historical block requires re-doing all the proof of work from that point forward at a rate faster than the rest of the network extends the chain, which becomes prohibitively expensive as network hash rate grows.
Proof of stake replaces computation with capital at risk. Validators lock up (stake) the chain's native token as collateral before they are eligible to propose or attest to blocks. The protocol selects validators proportional to their stake weighting. A validator that proposes an invalid block or signs conflicting blocks faces slashing: the protocol burns a portion of their staked collateral. Ethereum adopted proof of stake with The Merge in September 2022. The Ethereum Foundation documentation on proof-of-stake details the validator lifecycle, slashing conditions, and the finality mechanism. The IEEE survey on blockchain consensus mechanisms provides formal analysis of both models and their security assumptions.
| Property | Proof of Work | Proof of Stake |
|---|---|---|
| Resource expended | Computational energy (electricity + hardware) | Capital (staked tokens) |
| Participant role | Miner | Validator |
| Penalty for dishonesty | Wasted electricity; no block reward | Slashing (token loss) |
| Transaction verification speed | Slower; block intervals average 10 min (Bitcoin) | Faster; Ethereum finalizes in ~12 seconds per slot |
| Primary example | Bitcoin | Ethereum (post-Merge) |
Public, Private, and Consortium Blockchains
Public blockchain networks let anyone read the ledger, submit transactions, and run a validator node. Bitcoin and Ethereum are the canonical examples: both are permissionless, their ledger data is openly auditable, and neither requires identity verification to participate. The tradeoff is throughput; a fully open peer-to-peer network prioritizes decentralization over raw speed.
Private blockchains restrict validator access to a single organization. A company running a Hyperledger Fabric deployment controls which nodes can join, endorse, and commit transactions. This preserves the append-only audit trail while delivering throughputs far closer to a conventional database. Consortium blockchains distribute validator rights across a fixed group of known organizations rather than opening them to the public or locking them to one entity. R3 Corda, used widely in banking consortia, and supply-chain networks built on Hyperledger Fabric follow this model.
Self-executing code deployed directly on the chain, called a smart contract, operates across all three permission models. Smart contracts on Ethereum are the most widely deployed example on a public blockchain, automating settlement, token issuance, and governance without an intermediary.
| Attribute | Public Blockchain | Private Blockchain | Consortium Blockchain |
|---|---|---|---|
| Who can read | Anyone | Authorized users only | Member organizations |
| Who can write | Anyone | Authorized participants | Member organizations |
| Who validates | Open validator set | Single organization | Fixed member group |
| Relative latency | Higher | Lower | Low to medium |
| Example | Bitcoin, Ethereum | Hyperledger Fabric (enterprise) | R3 Corda (banking) |
What Blockchain Is Actually Used For
Blockchain's application scope extends well beyond cryptocurrency, though that remains its highest-volume use case. The digital ledger's core properties, including append-only writes, cryptographically verifiable history, and decentralized network validation, apply wherever multiple parties need a shared record they cannot each unilaterally alter. Readers looking for a deeper treatment of production deployments will find them covered as part of broader emerging technology infrastructure.
- Cryptocurrency. Bitcoin stores value-transfer records as immutable records on a public blockchain; every transaction since January 2009 remains auditable by anyone.
- Supply chain traceability. Walmart Food Safety uses a Hyperledger Fabric deployment to trace produce provenance in seconds rather than the days a paper-based system required.
- Digital identity and credentials. Self-sovereign identity systems anchor verifiable credentials on a distributed ledger, letting individuals prove attributes (age, qualification, citizenship status) without exposing the underlying document to a third party.
- Healthcare data sharing. Patient record access can be logged as permissioned ledger entries with auditable trails, reducing unauthorized access and simplifying compliance reporting.
- Tokenized assets. Real-world assets including bonds, real estate, and commodities are represented as smart contracts on public chains, enabling fractional ownership and automated settlement without a central clearinghouse.
What Blockchain Cannot Do
Blockchain advocates frequently omit the structural constraints of the protocol. Three limitations are worth understanding before evaluating any blockchain proposal.
- The scalability trilemma. Vitalik Buterin's theorem holds that a blockchain cannot simultaneously optimize for decentralization, security, and throughput. Bitcoin processes roughly 7 transactions per second; Visa handles approximately 24,000. Ethereum's proof of stake consensus mechanism raised throughput compared to its proof of work predecessor, but layer-1 capacity remains constrained by the need to keep node participation accessible. Layer-2 solutions like Lightning Network and Ethereum rollups address this externally, but they reintroduce trust assumptions of their own.
- Immutability as a compliance problem. The same property that makes blockchain records tamper-evident creates a GDPR conflict. Once data is written to a public blockchain it cannot be deleted. GDPR's right to erasure (Article 17) requires that personal data be removable on request. Storing personal data directly on a public chain makes compliance with erasure requests technically infeasible. Architects route around this by anchoring only cryptographic hashes of personal data on-chain while storing the underlying data off-chain, but that design introduces its own complexity. A digital ledger built on immutable records is not a neutral data store for every use case.
- The oracle problem. A smart contract executes code against data that lives on-chain. When that contract needs real-world information (a commodity price, a shipment status, a weather reading) it must receive that data from an off-chain source called an oracle. The decentralized network's transaction verification guarantees apply only to what is on-chain; they cannot extend to the accuracy of oracle inputs. A supply-chain contract that relies on a single compromised oracle produces immutable records of false data. Projects like Chainlink attempt to mitigate this with decentralized oracle networks, but the oracle problem is a fundamental boundary of the consensus mechanism, not an implementation bug.
Further reading
- Bitcoin: A Peer-to-Peer Electronic Cash System (Satoshi Nakamoto, 2008): primary source for the block structure, Merkle tree design, and PoW consensus consensus.
- NIST Interagency Report 8202: Blockchain Technology Overview: authoritative definitional reference for distributed ledger technology and consensus types from the U.S. government standards body.
- IEEE survey on blockchain consensus mechanisms: peer-reviewed analysis of PoW, proof of stake, and alternative consensus protocols with formal security analysis.
- Ethereum Foundation documentation on proof-of-stake: primary reference for validator mechanics, slashing conditions, and finality as implemented post-Merge.
- Smart contracts on Ethereum: procedural guide covering how to write and deploy smart contracts on Ethereum's public chain.
- OpenAI Alternatives: Best Large Language Models for Developers
- Quantum Computing in Drug Discovery: VQE, QAOA, and the Pharma Pipeline
- Quantum Computing in Finance: Portfolio Optimization, Risk, and Pricing
Additional refs: IETF RFC 6920 Naming Hashes.
Frequently Asked Questions
Is blockchain only used for cryptocurrencies?
No. Blockchain is a general-purpose distributed ledger technology that records any kind of transaction or data, not only financial transfers. Cryptocurrency was the first large-scale application, but the same protocol is used in supply chain traceability, digital identity systems, healthcare data sharing, and tokenized real-world assets. The underlying mechanism (cryptographically linked blocks validated by a peer-to-peer network) is application-agnostic.
Can data stored on a blockchain be deleted or changed?
No, not without the consensus of the network. Each block contains the cryptographic hash of the block before it, so altering any historical record invalidates every block that follows. On a public chain like Bitcoin or Ethereum, overwriting history would require controlling more than 50% of the network's computational or staking power simultaneously. This makes blockchain records tamper-evident by design, but also means personal data governed by GDPR's right to erasure cannot be stored directly on a public chain.
What is the difference between a blockchain and a regular database?
A conventional database is controlled by a single administrator who can update, delete, or restrict access to any record. A blockchain is a distributed ledger replicated across many independent nodes; no single party controls it. Records are append-only: once a block is confirmed, the data cannot be revised. The tradeoff is throughput: a traditional database can process millions of writes per second, while most permissionless chains handle tens to thousands of transactions per second.
