On-chain scaling vs layer 2 scaling: What are the differences?

You can find out a lot about a person by asking them whether they believe Bitcoin can scale on-chain. Even at a conference hyper-focused on highly scalable blockchains and the utility they enable, a poll of the audience is likely to reveal some strongly opposed opinions.

Some will say yes; of course, Bitcoin can scale on-chain. Others will say that it is impossible (or desirable) and that scaling on Bitcoin needs a layer-two solution to work.

But regardless of your opinion on this question in 2023, you cannot escape the fact that Satoshi Nakamoto’s ambition was to have Bitcoin scale natively: not only was this an explicit goal of Bitcoin, it necessarily follows from its ambition to serve as peer-to-peer electronic cash.

No doubt this was the spirit in which Kurt Wuckert Jr., moderator of the “On-chain vs. Layer 2 Scaling: Exploring the differences” panel at the London Blockchain Conference, asked the assembled audience for a show of hands:

“Who thinks Bitcoin is slow?” a smattering of hands went up before the follow-up: “Who likes being lied to?” None did.

Joined on stage by Connor Murray, director of stewardship at the BSV Blockchain Association; Seamus Andrew, lawyer and partner at Velitor Law; and Jack Davies, senior strategy manager at nChain, Wuckert toured the arguments for and against on-chain scaling and tried to make sense of the persistent belief that it can’t—or shouldn’t—be done.

He began by asking the panel to take a crack at defining what makes something ‘Layer 2’. Boldly (though maybe not surprisingly), it was the only lawyer on stage who put his hand up:

“A Layer 1 transaction has to manifest itself on the blockchain,” said Andrew. “Although there are all sorts of ways in which layer 2 purports to manifest on the blockchain, they don’t really on analysis, in my view.”

He referenced BTC’s Lightning Network, the long-promised (“It’s been 18 months away since about 2016,” joked Wuckert) Layer-2 solution that is supposedly necessary for Bitcoin to scale on chain.

“Lightning would be a possible example of Layer 2, where you take Bitcoin off the blockchain, play around it, these transactions are recorded on lightning, and they go back to the blockchain in the end”—but crucially, this is not the same as a transaction manifesting on the blockchain.

Perhaps miraculously, the technical minds on stage broadly agreed with Andrew’s definition.

“The distinction is the directness of how you’re using the blockchain,” elaborated nChain’s Davies.

Davies explained that when you move to Layer 2 solutions, you begin playing by a fundamentally different ruleset than in Bitcoin. On Layer 2, says Davies, you’re dealing with Bitcoin plus something else. He says this increases the “surface area of risk” instead of keeping everything on Layer 1.

That all seems pretty straightforward, so why dedicate an entire panel discussion to this distinction?

Murray said this could be boiled down to two points.

“The first is scalability: this myth that Bitcoin couldn’t scale.”

The myth Murray refers to is almost archaic. It’d be pre-historic if it wasn’t for the fact that it, by definition, arose long after the release of the Bitcoin white paper, which specified a scaling system and said nothing about the need for an additional layer for scaling. The myth is especially absurd when discussed on-stage at a conference packed with applications built on Bitcoin that rely on exactly the kind of scalability that isn’t supposed to exist.

“The other is that there is a desire for anonymous transactions and the obfuscation of the identities behind those transactions.”

In other words, keeping transactions on-chain is disastrous to the wrong kind of person.

But as the panel discussed, taking them off-chain can be disastrous, too. Wuckert asked his panelists what the real trade-offs are when adding the complexity of Layer 2 to the Bitcoin system.

“From a topological perspective, the nodes in the network form a densely connected core, and once you send a transaction to Layer 1, it’s reflected instantaneously,” said Murray. “In layer 2, it’s a mesh network—to get the transaction from a to b, you have to go from hop to hop. In Bitcoin, you’re talking 2 hops or less.”

Each additional layer, therefore, introduces another series of hops that must occur before a transaction is reflected on-chain.

“If you go back to a question one could ask oneself in any difficult situation—what would Nakamoto have done?” said Andrew.

“If you go back to the white paper, he was specifically trying to eliminate the requirement for trusted third parties. One way of looking at all these layers is it’s a series of trusted third parties introduced into the system. In that sense, rather than eliminating the trusted third party, you’re introducing new trusted third parties, like the lightning network. You have to trust them to do their bit.”

Seizing the opportunity to ask a clearly well-informed legal mind about the practical consequences of this, Murray asked Andrew what some of the legal implications might be.

“I think that it does create more legal complexity,’ said Andrew. “The blockchain is an elegantly simple system. One of the purest objections I feel about moving from layer 1 is you lose some of that purity. It dilutes the original Satoshi idea. Legally that has its ramifications—you create a whole new network of contractual relationships and several layers of jurisdictional complexity if they are based in different jurisdictions, which they always are.”

“The big elephant in the room,” adds Davies, “is when you add new third parties, they leech from the underlying system, so you get this zero-sum game between the actual fees nodes can collect and what these other providers are doing, because the end users have to split their funds between them.”

It’s clear there are layers of absurdity to the on-chain vs. Layer 2 philosophy. Like the fact that, pointed out by Murray, despite being held up as the solution to a problem that arguably doesn’t exist, the scaling powered by Layer 2 pales in comparison to what has been shown to be possible with on-chain scaling protocols like BSV—which offers the ability to process millions of transactions a second—as a start.

Or the fact that on-chain scaling wasn’t abandoned because this imagined scaling limit was reached by early forms of Bitcoin.

“One of the first things to say would be that lots of the bottlenecks… are bottlenecks that were theoretical. We didn’t move to layer 2 because we were hitting the bandwith bottlenecks or signature bottlenecks. It was more abandoned from a philosophical perspective: ‘we don’t want big blocks for reasons nothing to do with scalability,'” says Davies.

Why, then, is the desire for Layer 2 solutions so persistent?

In addition to the ideological (and criminal reasons) alluded to above, Murray suggests part of it is a sunk cost fallacy: “we’re at the point now where it hasn’t worked, so they keep trying to add. You’re so invested in this narrative and this idea; you’ve rejected the other side: maybe that’s part of it too.”

However it is we got to this point, there’s one thing that on-chain scaling will always have on Layer 2 solutions: the truth.

“Layer 1 can be framed as a concept of truth. Only by going to a blockchain that is run on a consensus mechanism can you establish the verifiability of all the facts,” said Andrew. “What we’re looking at on layer 1 is the truth. The blockchain is important philosophically for that reason alone.”