On the Principles of Financial Transactions Secured by Witnesses
Satoshi Nakomoto notes in the first paragraph of his famous paper (Bitcoin: a Peer-to-Peer Electronic Cash System) that financial transactions over the internet are too expensive to enable micro-payments. He then presents Bitcoin as a solution. Nine years later however, that particular problem is hardly solved.
Much has been written about high energy costs, low transaction speeds and concentration of power that ridicule the so important decentralisation. In this article I present another, inherent reason that principally prevents Bitcoin from enabling micro-payments. To understand, one must look afresh at the concepts ‘value’, ‘information’ and ‘money’. The last two especially are core concepts of the ideology of our time. It is hard to question one’s own ideology. What would a fish ask about water?
Value (and storing it)
An object has value if people are willing to part with something, or to expend some effort, in order to acquire it. Those who are thirsty, will pay for water. Such people need water. People go to lengths to acquire gold, too. But no one needs gold like they need water (mental syndromes aside). One desires gold because it is valuable. That closes the loop: gold has value because people will give a lot for it and they do so because it is valuable.
Gold is valuable in our society, but it hasn’t been always and everywhere like that. The value of gold is culturally and historically determined. In the parlance of monetary goods (the general name for gold, diamonds, art, etc.) it is about the network effect and path dependency. The more people value gold, the more valuable it becomes. That is the network effect. And the price of gold is strongly determined by what has been paid for it recently. For example, a rising price is likely to continue rising. That is the path-dependency. In the course of a society’s history something like gold can, over time, become a store of value.
The metaphor, ‘store’, is interesting. It suggests that the medium has no intrinsic value, like a bottle doesn’t have value but the wine in it does. It suggests, too, that different amounts of value can be stored in it (the intriguing metaphysical question being, of course, how value can be stored in something anyway!).
Not every object lends itself as a store of value. Rare objects are more suitable. This disqualifies water and air, just to name two. An artefact that can be easily constructed is not suitable, either. Gold is rare and cannot be created (well, not easily). Obviously, if the construction cost of a type of object is larger than the value assigned to it, creating more of it is not useful and that contributes to its suitability as a store of value. It helps if an object can change of owner easily. Price is established in trade, so ownership must be flexible. Obviously, storage must be sustained over long periods of time. And finally, one should be able to establish that the object is the real thing. Pyrite is a metallic substance that resembles gold, but it is a lot more common — and thus, cheaper. Pyrite is called fools gold, for obvious reasons.
Bitcoin develops as a store of value. The number of Bitcoins is limited; more and more people desire Bitcoins; it can be traded easily and cannot be forged. As a bit of information it is very durable and available. The price of Bitcoins has (averaged out over longer periods) gone up for nine years and probably will continue doing so.
(Our) money is debt
Suppose you want to build a house and lend from the bank. It provides half a million euros in new notes that bear your name on the backside (not really, but it helps to understand how it works). Moreover, there is this magical machine that changes a note with an arbitrary name on it for one that bears yours. Using your brand new notes you pay ten construction workers who build your house. Your money spreads through society and it is your job to bring it back to the bank (it wrote down how much you owe). So you get a job and make money. Each month you set aside some notes. In general these will not be your notes, but using the magical machine you change them into notes with your name on it and bring them back to the bank. The bank burns these notes and updates your debt administration. Thirty years pass — you have to live, too — before you’ve returned the very last note.
This is how money operates.
The magical machine? It operates quite simply. It has an administration of all personalised notes in circulation (how many for whom), a money printing press and an oven. A note with John’s name on it that you return, is burnt. The number of notes bearing John’s name is decreased by one. The press prints a note for you and this is duly noted. The total number of notes in circulation is not changed. Now, if John inserts a note bearing your name, the inverse happens. Attention: this administration is not the same as the administration of debt kept by the bank! It may happen that you’ve paid off your debt entirely, while there are still notes with your name on it in circulation. But these will eventually be inserted by someone else.
In other words: it doesn’t matter whose name is on a particular note (that is, obviously, why the bank prints standard money without names…).
So this kind of money really is an administration of debt. A note is like a written declaration of debt. But, anonymously. The bank knows how much you’ve got to return, but you cannot derive that from the notes in circulation.
The value of money
People do things for money. Why? Because they can have other people do things for that money. Money has, just like gold or Bitcoins, a value. But there are differences.
The value of money is calibrated time and again by the creation and destruction — lending — of money. A mortgage is only granted if it is likely that the receiver will pay off the last penny. That ties money to real life, with realistic efforts of real people. That is what gives money a relatively stable value. This is important. Imagine that the value of money would increase tenfold during the lifetime of a mortgage — and Bitcoin value fluctuations exceed that in much shorter timespans! — then it is very likely that mortgages would never be payed off. Inheritors would be saddled with debt, generations would be enslaved to the bank. Reversely, if money loses its value really quickly, another disaster happens: people lose their pensions, investments disappear into thin air, etc. To sum up: the value of money should be stable, changing only slowly.
The value of Bitcoins and gold arises from trade in Bitcoins and gold. The value of money arises from trade with money, stabilised by continuous calibration of creation and destruction of money.
Having written that, this is how it should be. One might reasonably doubt whether it still is, in these times of ‘quantitative easing’ practiced by central banks.
The Latin verb ‘informare’ can be translated with ‘giving shape’. ‘Information’ might be understood as ‘a thought that has acquired a shape’. One should understand that as a physical shape, like a clay tablet or a letter. ‘Information storage medium’ is a tautology because the physical carrying is already implied in ‘information’. ‘Thought storage medium’ would be better.
An aside: what form lends itself to carrying thoughts? Usually a community of people agrees on a convention. An alphabet and written language are examples and so is spoken language. The chosen sounds are arbitrary and carry their meaning by convention. Artists, however, spend their creative lives looking for the right form to convey a thought.
Why do we give thoughts a shape? In order to communicate. One can send a letter. A sentence can be spoken. We want to communicate and as long as we cannot read thoughts, we must make the effort to write thoughts. ‘Writing’ in the most general meaning of the word: ‘expressing’, ‘giving a shape to’.
Money as information
A bank note obviously is information. In the paragraph before last I argued that the thought that is carried in this case, has to do with debt. The note with your name on it that is in circulation expresses ‘I should pay this loan’.
‘I should pay this loan’ is a different kind of thought than ‘I should buy peanut butter’. The former is typically found on shopping lists and constitutes information that you send to your future self (when you are in the supermarket).
But ‘I should pay this loan’ or, better, ‘Joop Ringelberg should pay this loan’ is a statement. A statement is declared to others. There is little sense to declare something that nobody can hear (psychological effects aside).
Statements, promises, requests, etcetera are called ‘speech acts’. They are acts between people, executed by speech.
Bank notes are a written form of the speech act ‘to acknowledge debt’, ‘to enter a commitment to repay’.
This sheds light on forgery. Remember how bank notes could bear your name. A forgery of a note is like accusing someone wrongly of debt! Small wonder we don’t like counterfeiters.
By far the biggest part of the money in circulation doesn’t have a physical form like notes or coins. Many authors speak of ‘scriptural money’ or ‘virtual money’. The latter form especially suggests an almost meta-physical quality. But even virtual money is a thought expressed in a physical form. But what form, exactly?
Consider electronic payments. An account essentially is a number. Withdraw some money and the number correspondingly decreases. Each account is a series of increases and decreases, debet- and credit acts. What is credited to one account, is debited from another, thus preserving a balance. Crediting and debiting happens in a single, unbreakable transaction — meaning one cannot be done without the other.
When you borrow money (from the bank), your account is credited, but no other is debited. Money is created. Reversely, when you pay off a debt, your account is debited but no other is credited. Money is destroyed. Here we break the law of balanced transactions. Borrowing from the bank is not a transaction between accounts!
This administration is the ‘shape of the shape’. Long ago the physical form was a written ledger, now it is a database.
Banks record these transactions. As that is just elementary math, why shouldn’t account holders do it themselves? It would be far cheaper. One just needs a small ledger to register money coming in and flowing out. Even better, we could have an App doing the administration.
If A sells something to B for 100 euro’s, they would contact each other through the App. Each App keeps a record of expenses etc. on the owners’ phone. A’s App adds a 100 euro’s to A’s account and B’s App subtracts the same amount from B’s account. A little bit more accurate: A’s App receives the height of B’s account and checks whether a hundred euro’s can be subtracted. Only then the transaction is carried out.
It might seem odd that no ‘real’ money is involved, but that doesn’t happen at the bank either — transferring money is just doing the math. It is manipulating information.
We could transfer money this way. But cheating would be very easy. Hack the App, increase your account.
This is precisely why and how the bank adds value and to what it owes its existence: it keeps the ledger, customers cannot manipulate it directly. And not just anybody can start a bank. Governments license banks. The government has the strong arm of the law. Altogether a quite reliable system — even if you don’t like it.
Satoshi Nakomoto famously writes in the first paragraph of his famous paper (Bitcoin: a Peer-to-Peer Electronic Cash System) that electronic payments are not completely irreversible. That is a friendly way of saying that cheating is possible. This leads to high transaction costs. It also makes companies should not trust their clients, on principle. His next paragraph introduces his solution:
“What is needed is an electronic payment system based on cryptographic proof instead of trust, allowing any two willing parties to transact directly with each other without the need for a trusted third party.”
Nakomoto wrote before the crisis of 2008. In 2018, all the wiser because of the astronomically expensive saving of banks that are to big to fail, it is widely held that we should turn in a new direction. Nakomoto’s brilliant idea is world famous. The average lay person knows the problem and the solution: Bitcoin.
Nakomoto looked for a way to break the bank’s monopoly. His solution was to discard ‘trust’ entirely.
Unforgeable money is reliable information. Bitcoins are information and have to this day proven to be unforgeable. Thus, a Bitcoin reliably transfers a thought. What thought? The most essential or minimal of all: ‘I exist’, or: ‘I am a Bitcoin’.
But the value of Bitcoin is not tied to the world of human effort and realistic commitments of working people. Bitcoin develops into a storage of value, like gold. Being unforgeable is essential. But it clearly is not enough to turn a medium into useful, reliable money that only slowly changes its value.
What we need is a reliable, stable money that we can spend in arbitrary small amounts. In other words, what we need is a reliable way to communicate speech acts like ‘I credit your account with X’ over the internet. How can we make speech acts like ‘Joop Ringelberg owes X euro’ undeniable?
It so happens that there is another way of doing just that. A way that is as old as humanity.
We marry in the presence of witnesses. This is because marriage is an act in front of a community of people, however much married life is a private affair. The witnesses listen in name of the entire community to the yes-I-do — a speech act if there ever was one!
Nowadays we also have a marriage act. It is like a contract that the witnesses, too, sign. Thus we put security on security. Not without reason: we may not be able to reach a witness when we need his testimony, the witness may lose his memory, turn out to be evil or even die.
The marriage act is information. It is the speech act in the form of a written statement that two people vowed to keep their matrimonial obligations. Their signatures supposedly make the marriage act non-repudiable and undeniable.
So the marriage act and the bank note are very similar. Both are information (carriers of a thought). Both give shape and form to speech acts. Both represent speech acts addressed to a larger community of people. The big difference, of course, is in the nature of the speech acts themselves. A bank note is a declaration of indebtedness, a promise to repay and thereby to end the loan. A marriage act is a declaration of reciprocal obligation without the intention to end it. Moreover it is an obligation of support and help, mutual accountability, etc. These obligations are anchored in law. Many people also understand the obligation to be one of loyalty and personal care and, last but not least, as a statement of love.
Many more relations between people arise as the consequence of a formal (= having a form, persisted as information!) speech act. Think of custody of children, acknowledging parenthood, the oath of a medical professional, of a political servant such as a mayor or alderman, etc.
Many if not all of these relations require witnesses to be valid.
We could apply this to money.
Let’s return to our imaginary payment App. Assume that a third party C witnesses the financial transaction between A and B. This party receives information about the transaction and holds on to it: the parties involved, the date and the amount of money.
Again, let us put B in the position of buying something off A. Now A’s App will not only receive information about B’s account from B himself, but also from the independent witness C! So if B hacked his account, this will come to light.
One may ask: how do we know that C is not deceived by B as well? This is because C witnessed B’s previous transaction — knowing that the transaction before that was sure because it was witnessed by another independent observer, etc.
This is why the independent witness is so important.
It is a social construction that may be as old as humanity — probably by definition.
Moreover, any party can act as witness. This will prevent each of them to get into the position of knowing too much about anyone’s financial position. Each organisation with a reputation to lose, can act as witness and facilitate reliable financial traffic between citizens. Such transactions create a realistic market for witnessing, allowing for a lot of competition (so much different from the current monopolistic situation!). The price of a transaction could lower substantially compared to tariffs payed currently to the banks. The actual effort is almost futile and can be fully automated while the number of potential suppliers is enormous.
We coin the name WitCoin for this form of money, for ‘Witness Coin’.
Last, but not least, this approach is fully scalable. Each partner brings his own witness to a transaction and pays his own energy bill and computing infrastructure. This is because the infrastructure can be completely distributed. Contrast this with Blockchain’s distributed database: the former means each participant keeps his own stuff, while the latter means the information collection is duplicated a number of times. WitCoin is the ultimate decentralisation. From the perspective of the bank-as-database, WitCoin is a form of ‘extreme sharding’.
Is WitCoin based on debt, like our current money? It could be. Imagine a lossless exchange where euro’s and other currencies could be exchanged for their WitCoin equivalent (as ‘coloured coins’: a WitCoin Euro, a WitCoin Dollar, etc.). Those WithCoins would have exactly the same characteristics as the original currency — but allowing far, far cheaper transactions.
Money based on debt is not without its own problems, interest probably being the biggest one of all. WitCoin is neutral with respect to that problem.
The bank, revisited
We can think of the bank as the reliable witness of a financial transaction. A witness in name of the community of people.
This aspect of being-a-witness has, however, been relegated to the background, in favour of the unforgeableness of money. As if the signature on the marriage act is more important than witnessing!
The introduction of scriptural money represented a further step and gave banks the monopoly to keep the ledger of debts and accounts.
Banks are extremely powerful companies that can force even the most powerful countries to solve their problems, to the detriment of its citizens — as we have seen during the 2008 crisis.
It’s about time that this form of witnessing is democratised!
Could we construct WitCoin using Blockchain?
A transaction with a witness sounds like a ‘smart contract’. Ethereum’s Solidity is, without doubt, suitable to program such a contract. WitCoin can then be based on Blockchain.
However, this would be putting the cart before the horse. The concept of witness is much more elementary than the unbreakable database of Blockchain. In a future paper I will discuss a conflict between two of the design decisions underlying Bitcoin. Witnessing will turn out to be a way of resolving that problem.
Nakomoto correctly notes that the grain of trust on internet is too big and especially that trust is too expensive. Micropayments are just impossible with transaction costs of twenty to thirty cent! In the meantime, today’s intensive cooperation and communication of the internet really needs micropayments of fractions of a cent.
The inherent characteristics of Bitcoin make it very suitable as store of value, but not as a coin for financial transactions in arbitrary small amounts of value. The fact that there is no mechanism tying Bitcoins value to the work and life of real people importantly contributes to that fact.
Our current, practical money is the specific shape given to a statement to repay a loan. As a form carrying a thought, money is information. Such information should be extremely reliable because no one wants to be saddled with a repayment obligation for a loan he did not subscribe to.
To this day, Bitcoin has a very good reputation of reliability. However, structural high energy requirements and limited scalability rule Bitcoin out as a practical system for large scale financial transactions. On top of that the built-in convergence to centralisation due to the system of miner rewards compromises, perversely, the reliability of the entire system in terms of non-censorship. Last, but not least, Blockchain is overly complex when it comes to constructing a system of reliable witnessing.
Testimonies are a historically proven method to make speech acts non-repudiable. Much of society’s fabric consists of such testimonies.
A distributed system of witnessing financial transactions would scale in direct proportion to the number of users.
What we need is not, like Satoshi Nakomoto writes, “an electronic payment system based on cryptographic proof instead of trust, allowing any two willing parties to transact directly with each other without the need for a trusted third party.”
He chose the wrong path.
We should not use technology to abandon trust in others, but on the contrary, to be able to work with it on the internet, too. At arbitrary scale.
It is about trust, dummy!