We Need Better Terms for Everything in Cryptocurrency (including “Cryptocurrency”)
“The beginning of wisdom is to call things by their proper name.”
-Confucius
Crypto discourse has gotten really angry.
The good news, however, is that I don’t think this problem is necessarily reflective of personalities in the crypto community. I mean sure, crypto-folk can be hostile, argumentative, closed-minded, and tribalistic, but so can everybody else on the internet. I suspect, rather, that the root of the current state of argumentative gridlock is a limitation not within ourselves, but in the words we have available to us.
Derived from a variety of sources — computer science, economics, social theory, and the like — we get an olio of appropriations, portmanteaus, and neologisms that comprise the current crypto lexicon.
TLDR: all the terms in it are bad. Let’s do better.
Most if not all of the following linguistic objections and suggestions aren’t new. In fact, that’s kind of my point; I’m sick of seeing people talk past each other and needing to have these things spelled out again and again, and if you partake in crypto twitter debates (god help you), there’s a good chance that you are too. Re-re-hashing these things ad infinitum is a frustrating waste of precious time and energy, time which could be better spent with friends and loved ones and energy which could be better spent mining Bitcoin.
If you’re in the crypto space, think of this as my small (but valiant) effort to move things in the right direction; if you’re outside of the crypto space peering in, maybe this can serve as some sort of lexicographical guide, or at least provide some context for why it is that you’re so confused (preview: it’s not your fault!)
Consensus Mechanism
The notion of consensus in distributed systems was (roughly speaking) introduced and formalized in Leslie Lamport’s “Byzantine Generals Problem,” a brilliant paper with insights that have been both productively applied and blatantly butchered ever since. A consensus mechanism, in the classical sense that Lamport established, is an algorithm by which a fixed, known number of actors pass messages to each other until they come to agreement about *something*.
Proof of Work and Proof and Stake are not consensus mechanisms. They may be cool, and they may very well even be more interesting, but they’re something else entirely. What POW/POS provide is a provable and unforgivable economic barrier (and reward, presumably) for participation and compliance within a system. This lets our system derive its security from purely economic properties all the way down, instead of ultimately resting on things like trust, identity, and provenance.
What POW and POS do not have anything to say about is how the parties involved communicate with each other and how they interpret said communications. For that you need… well, a consensus mechanism.
So what are POW and POS then? Emin Gun Sirer suggests Sybil Defence Mechanism, which I quite like. I’d also be fond of something along the lines of Incentivization Scheme, though admittedly that doesn’t quite have the same ring.
Here’s the tricky bit: POW/POS are indeed necessary features for enabling consensus to be achieved in a truly open, unpermissioned cryptocurrency system; thus, it’s tempting to resort to some half-baked term like “Nakamoto Consensus.” But the temptation to include the word “consensus” at all here should really be resisted, I claim: it only adds to the confusion and can yield dangerous results, namely, projects framing themselves as offering new alternatives to POW/POS when in fact all they have is some far less sexy variation on some old consensus mechanism, thereby bilking users and/or investors out of millions of dollars.
Buyer beware.
Smart contract
Smart Contracts are dumb. This is isn’t a criticism; the economic guarantee of bone-headed enforcement of publicly verifiable code without recourse to the social sphere is in fact what makes them useful (I know, I know, but that’s besides the point right now).
As for “contract”… I mean sure, you can think of them as agreements between parties I suppose, but the defining feature of smart contracts isn’t ossification of an agreement, but rather the ability to actually enforce and carry out consequences.
So what to call them then? Autonomous Program is one you’ll hear floated around, which, modulo any necessary unpacking of “autonomous,” seems, to my ears, to do a much better job of capturing their functionality. Let’s go with that.
Stablecoin
This can refer to a coin that tries to achieve value stability via some unorthodox economic mechanism, like decentralized over-collateralized loans or algorithmic supply adjustment to retarget price (this is one is still, errr… theoretical).
…it can also refer to a coin that simply comes with a guarantee of direct redemption for some other asset (one that, presumably, has a more reliably stable value than cryptocurrencies do).
These two things couldn’t be more different.
I nominate that we grant the “stablecoin” neologism to the more novel approaches (the former). As for the later, there’s an argument to be made that we don’t need a new term at all; Gemini’s GUSD, for example, is ultimately, just a new way to transfer dollars, and thus, in the same way that we didn’t feel the need to invent terms like “Paypal dollar credits” or “Venmo digi-USA-bucks,” maybe we ought to just call them dollars too. If, however, we must have some way to show off how blockchain-savy we are to our friends: I’ve been trying to make fiat proxy a thing for a while, but nobody seems to care, so… fine. Maybe ___ backed token (dollar, yen, gold, etc) is the best option. It’s boring terminology, but I’m fine with that, because frankly, it should be.
Layer 2
To some, “layer 2” suggests things like payment channels and plasma chains; i.e., mechanisms by which transactions can achieve finality and maintain the main-chain’s security guarantees without ever actually having to touch the chain itself. To others, “layer 2” can refer to literally any medium for transferring crypto-wealth that doesn’t require an on-chain transaction, i.e., channels, plasma chains, federated sidechains, coinbase, IOUs on post-it notes passed to your classmates, etc.
That we have created a confusion-vector by which some folk will assume the later necessarily implies the former is really, really bad.
I nominate that we stick to off-chain for the superset of all “non-layer 1” transactions, and reserve layer 2 for the more interesting sub-set (bias mine).
Decentralized
There’s no shortage of philippics out there about the problems with this word, and I think I basically agree with all of them. Consider the following:
- Non-custodial
- Distributed
- Unpermissioned
- Fault tolerant
- Democratic
- Open source
- Non-hierarchical
Odds are, your intent when you say “decentralized” is better signified by one (or perhaps two or three) of the those terms. By sticking to the D-word, you’re making people think you’re implying things you aren’t and you’re making them mad at you. Stop saying it, and make the world happier.
Sidechain
Everyone knows what a sidechain is: it’s when you lock coins in a multi-sig transaction on your mainchain, and then give a federation control over pegging those to coins in a separate, permissioned blockchain environment. Although maybe to really be considered a sidechain, mining hashrate has to be in control of the peg. Okay, but either way, the defining feature is that it isn’t technically a non-custodial system. Unless, of course, it is. Also, just FYI, “altcoins” might actually be Bitcoin sidechains. And Jorge Stoff’s living room sofa might be too.
I have no suggestions for cleaning up this mess. It’s hopeless. You’re on your own.
Blockchain
I’ve heard unsubstantiated rumors that there was once a time where this term had meaning. If true, those days are long gone.
The policy of “It is a blockchain if it has blocks and it has chains” sounds reasonable on its surface, until you realize the wide net of essentially unrelated systems captured by this definition, many of which have none of the properties blockchain proponents are wont to rave about. So we could prescriptively enforce the rule that to earn the blockchain descriptor, a system must meet certain blockchain-y requirements. But this leaves us with the task of sorting out whether a protocol meets your particular standards of distributed-ness, trustless-ness, immutability, etc., leaving us right back into the semantic wasteland territory we started in. So we either have a term so general as to be meaningless, or a term so loaded with baggage that it will inevitably spread more confusion than anything else.
I’m afraid this term has to go the way of “sidechains”; get rid of it. I know that sounds scary, but we’ll be better off without it, I promise. It’ll all be okay.
Hard Fork
So let’s recap: in the crypto world, we have a general notion of forks, which we get from the normie software engineering world: an open source project’s code base is copied and used as the starting point for a new project. The “soft fork” / “hard fork” terminology refers to an update in the chain validation rules themselves, soft/hard signifying whether or not the change maintains backwards compatibility with the nodes that don’t accept the update. Often, however, the usage of hard fork is assumed to signify an update that results in the single chain splitting into two new persistent new ones (i.e., Ethereum spawning Ethereum Classic or Bitcoin spawning Bitcoin cash).
If you think this isn’t confusing, you’re deep in a bubble and you should step outside.
The terms chain split, or sustained chain split are handy to have in your arsenal if that’s indeed what it is you’re trying to say. If you really are just talking about upgrades to the consensus rules, unfortunately, the term “fork” inevitably elicits images of one thing bifurcating into two; saying soft/hard upgrades may take some getting used to, but I think it’ll do just fine.
Cryptocurrency
How did we let this one slip through?
You’ll often hear it claimed that cryptocurrency is “money secured by cryptography,” or the variation “secured by math.” While this is technically true, I guess, “secured by cryptographically & math” also applies to credit card payments, text messages, password verification, DNS validation, and basically everything in your life that doesn’t involve parchment and paper (and sometimes even that!).
So what is it that differentiates cryptocurrency? Is it the digitial-ness? Well… no. We’ve had a long history of (mostly failed) digital currencies before the recent crypto boom. Hell, even the dollar itself is 90% digital.
So if digital-ness is necessary but not sufficient, and cryptography is only a means to an end — what is, in fact, this end? Well, it’s… all the terms listed in the decentralized section, so really, the term we pick should reflect that: Immutable Digicoins? P2P e-money? Electro-Freedom Credits? (Open to suggestions here.)
Onward
…All that having been said, I will most likely continue to use all of the above problematic terms. There are three reasons for this:
- Language tends to evolve organically even with explicit attempts to guide and coordinate it, so if one wants to effectively communicate one must, to some degree, conform to linguistic community standards.
- All of my suggestions have more syllables than their corresponding original terms.
- I’m a hypocrite.
But still though. If we want to get any closer to these dreams of mainstream adoption, closed-loop cryptocurrency (!) markets, “institutional investors”, etc. etc., we’re going to have to start being more coherent. After all, if we can’t even decipher what we’re yelling at each other, what hope do we have of bringing Average Joe and Jenny No-coiner on board? Here’s to making better sense.
Words matter.
Honorable Mentions:
ICO, Altcoin, Full Node, Peer to Peer, Scam, Sound Money, Anonymous, Token, Cryptoeconomics
