The importance of trade and what we ambiguously call “money” demonstrates a simple fact: humans are natural cooperators. Evolutionarily, this is no surprise. Agents gain more if they manage to coordinate on mutual contribution than if they act alone.
Any subset of a population that succeeds in coordinating on such cooperation will tend to outcompete those that do not. The evolutionary problem is therefore not why agents would want to help one another, but how such agents can find one another across time and space.
The specific difficulty is this: deferred reciprocity requires that contributors be identifiable after the act of contribution and in the absence of memory.
Animal signals
The essence of the problem is signaling.
Consider a population of agents who wish to help one another whenever possible. Suppose agent j helps agent i at time t₁. At time t₂, agent j encounters agent k and needs help.
For deferred reciprocity to function, j must be identifiable by k (who might be the same as i).
Ordinary signals — such as marks, gestures, or speech — cannot solve this problem. If agent j emits a signal that agent i can also emit, the signal cannot uniquely identify agent j. Agent i may emit the same signal and be helped by k at time t₂, even if only agent j has contributed.
This is why promises, declarations, reputational claims, and symbolic badges and tokens fail without a prior rule. What agents require is not mere communication, but a signal that only the contributing agent can emit.
Transferability and objecthood
Transferability is a necessary condition for such exclusivity.
Consider the difference between a balance on an account and a bearer instrument. A balance is a signal that can be re-emitted — anyone can replenish an account. A bearer instrument irrevocably transfers signaling capacity when passed on. The sender can no longer signal.
This is an essential feature: the ability to transfer signaling capacity from one agent to another, disabling the sender from signaling again. The signal thus acquires the character of an object — a structure whose control confers signaling capacity and whose transfer reallocates that capacity.
This object enables a simple decision:
Agent j helps agent i if and only if agent i transfers the object to j.
The holder of the object is thereby identifiable. No assumptions about trust, value, or prior rules are required. Identification is carried by the object.
Uniqueness
Transferability is necessary but not sufficient.
If multiple agents can hold equivalent objects, two agents would be able to signal the same contribution even if only one had contributed.
This is why mere material concreteness is insufficient. A coin or a note may be transferable and recognizable, but if it belongs to a class with more than one member, it does not uniquely identify a contributor by its intrinsic properties.
Therefore, the object must be unique: the only member of its class, either as a whole or as a verifiable part of a divisible, unique whole. The scarcity of a substance such as gold matters only insofar as it preserves this uniqueness.
Money and credit
At this point, money can be defined minimally, without reference to social or cultural criteria:
Money is a unique object whose transfer secures reciprocity.
An object is money if its intrinsic properties enable reciprocity between agents who possess no memory or persistent identity. Everything commonly attributed to money — value, medium of exchange, store of value, unit of account — follows from this more basic role. Terms like “hardness” or “final settlement” are approximations to the same underlying signaling properties: uniqueness and transferability.
But what about objects that lack uniqueness? A non-unique object cannot secure reciprocity by its intrinsic properties. It can still circulate — but only if agents follow a prior rule: accept the object in return for contribution. This is credit.
The distinction is fundamental. Under money, the agent values the object for what it is — the only member of its class. Under credit, the agent values the object because others accept it.
Calling non-unique items “money” conflates two structurally distinct mechanisms. Existing accounts of money describe credit systems: they model environments in which reciprocity is sustained by a rule, not secured by the object itself.
Bitcoin and ideal money
The origin of money is often explained by reference to cultural or conventional criteria, such as state authority or a commodity’s “saleability.” These are accounts of how credit systems arise, not of what makes something money.
According to the present argument, an object is valued because its uniqueness allows agents to realize the benefits of cooperation without presupposing any rule or institution.
Across history and across environments, humans have repeatedly converged on such objects:
- salt and grain in constrained economies
- cattle and livestock in pastoral societies
- cigarettes and alcohol in prisons and closed systems
- shells, beads, and ornaments in intergroup exchange
- metal tools, bars, and coins
- gold as a divisible, globally recognized stock
- and now, bitcoins on the internet
What these objects share is not intrinsic utility or cultural meaning, but contextual uniqueness and transferability. Political or commodity use is contingent, not foundational.
Bitcoin makes this explicit. Bitcoins have neither government backing nor material substance. They consist entirely of digital signals — ones and zeros transmitted over a network — yet they are called “digital gold”. The Bitcoin protocol achieves uniqueness and transferability at a global scale, with a fixed limit of 21 million units, each divisible into 100 million satoshis.
Bitcoins approximate the scientific definition of money more closely than any previous monetary object. For the first time, money has been reduced to its primitive signaling properties.
References
Maanmieli, Jose. Money as a Signal (2026). Available at SSRN: https://ssrn.com/abstract=6400919
Maanmieli, Jose. Money is a Token of Cooperation: A Theory of Exchange in Animals (2026). Available at SSRN: https://ssrn.com/abstract=6500722
