What Is a Blockchain? A Plain-English Explanation

Start with the problem it solves

Any shared record has the same weakness: someone has to maintain it, and everyone else has to trust that they are maintaining it honestly. Your bank balance is a number in the bank's database. You trust the bank not to change it arbitrarily, and mostly that trust is well placed because the bank is regulated and audited.

A blockchain is an attempt to build a shared record where that trust is not required — where the honesty of the record comes from its structure rather than from the reputation of whoever holds it.

What a block is

A block is a batch of transactions bundled together, plus a timestamp and a reference to the block before it. Nothing exotic: a list of "A sent 3 to B" entries with some metadata.

Why the chain matters more

Each block contains a cryptographic hash of the previous block. A hash is a fixed-length fingerprint of some data, with two useful properties: the same input always produces the same output, and changing the input even slightly produces a completely different output.

Because block 500 contains the hash of block 499, altering anything in block 499 changes its hash, which means block 500's reference no longer matches. To make the tampering consistent you would have to recompute block 500 as well — and 501, and every block after it.

That is the mechanism. Not that the data cannot be changed, but that changing it requires redoing all the work built on top of it, publicly, while everyone else continues building on the original.

Who keeps the copies

The chain is replicated across many independent computers, called nodes. Each holds a full copy and each validates new blocks against the network's rules. There is no master copy to attack, and a node proposing an invalid block is simply rejected by the others.

How new blocks get added

Someone must decide which transactions go into the next block, and that decision has to be resistant to abuse. The two dominant approaches:

Mechanism	How it works	Used by
Proof of Work	Participants expend computing power to earn the right to propose a block	Bitcoin
Proof of Stake	Participants lock up capital, which is forfeit if they behave dishonestly	Ethereum and most newer chains

Both make dishonesty expensive rather than impossible. That is the real design principle behind the whole system.

What blockchains are bad at

Being honest about the limitations is more useful than the usual enthusiasm:

Speed and cost. Having thousands of computers independently verify everything is inherently slower and more expensive than one database doing it once.
Privacy. Most blockchains are public. Anyone can see every transaction associated with an address, forever.
Reversibility. There is no support line. A transaction sent to the wrong address is generally gone.
Off-chain truth. A blockchain can prove a token moved. It cannot verify that a real-world thing the token supposedly represents actually exists.

The technology is genuinely useful where the absence of a trusted central party is the point. Where a trusted party already exists and works, a normal database is usually better — and recognising that distinction is what separates a considered view from a promotional one.

Where to go next

Once the record-keeping makes sense, the next practical question is how you control what is recorded as yours — which comes down to public and private keys and the seed phrase that backs them up.

Public, private and permissioned chains

Not every blockchain is open. A public chain like Bitcoin or Ethereum lets anyone run a node, read the data and submit transactions. A permissioned chain restricts who can participate, and is used by consortia that want shared record-keeping among known parties.

The distinction matters because most of the properties people attribute to blockchains — censorship resistance, verifiability by anyone — come specifically from being public and permissionless. A permissioned chain run by one organisation has most of the costs of a blockchain and few of the benefits, which is why many corporate blockchain projects quietly became ordinary databases.

Forks: what happens when the rules change

Because the network is a set of independently run nodes following shared rules, changing the rules requires agreement. Where nodes disagree, the chain can split.

Soft fork — a tightening of rules that older nodes still accept. Backwards compatible.
Hard fork — a change older nodes reject, producing two chains if a meaningful group declines to upgrade.

Several well-known assets exist because of hard forks. For holders, a fork can mean receiving tokens on both chains, which has its own tax consequences in India.

What "decentralised" actually measures

Decentralisation is not binary and is worth examining rather than accepting as a label. Useful questions: how many independent nodes exist, how concentrated is block production, who controls the main client software, and how much of the supply sits with a small number of holders?

Many networks described as decentralised score poorly on at least one of these. That does not make them useless, but it does mean the trust assumptions are different from the marketing, and understanding where the actual dependencies lie is part of evaluating any chain honestly.