© Philipp Paech 2017
A distributed record as described in my previous post is only the base component of a blockchain network. In particular, additional mechanisms are needed to guarantee that the updates of records kept by nodes reflect the truth since practically any node would be in a position to propose updates to the other nodes, including fraudulent ones.
No trusted third party
Traditionally, the truthfulness of records in financial markets is ensured through a mechanism involving trust (in the everyday sense of the word[1] and responsibility. Clients trust their intermediaries to keep records diligently so that they reflect the true state of holdings at any given time. Reputation may be the original bedrock of this trust, but more importantly today it is a question of regulation: clients typically trust financial institutions because they know they are authorised and supervised.[2] Clients expect intermediaries to be able to correct erroneous records, and to do so either voluntarily or compelled by the judiciary.[3] In other words, regardless of the outcome of the technical process of record keeping, intermediaries and, ultimately, the courts have the last word as to whether rights such as securities or cash in accounts have been acquired or lost and, hence, whether the relevant record entries correspond to the truth.
By contrast, the inventors of blockchain relinquished the current model for ensuring truthful outcomes built on ex-ante regulation-induced trust and ex-post review by the courts.[4] Instead, blockchain technology relies entirely on a technology-based solution giving nodes the certainty that transactions are correctly executed and accurately recorded. In addition to the idea of a distributed record (see preceding section), the concepts builds, first, on a process to establish consensus amongst nodes regarding the correctness of an update of a record on the basis of a mathematical-probabilistic approach (this process is called the ‘proof of work’ in the Bitcoin context) and, secondly, on a process by which all processed transactions are locked in a chain of sequential, logically intertwined sets, or ‘blocks’, that cannot be changed – in principle[5] – once a new block of transactions has been validated by the nodes (this latter feature is the origin of the term ‘blockchain’).[6]
Permissioned and unpermissioned networks
For such a system to work, however, it is imperative that no person or group be in a position to take control of the majority of nodes and thus of the validating process. This goal is achieved by conceiving the network as ‘permission-less’,[7] ie, as an open network. Anyone with the necessary (freely available) hardware and software can join Bitcoin, Etherium and other networks as a node following this strict logic. Even though this openness may allow fraudsters to join, the idea is that the well-nigh unlimited reservoir of computing power spread across the globe can theoretically be made available to the network and will always be greater than the computing power of a potential attacker, thus rendering the network tamper-proof and censorship-resistant.[8] Newer blockchain networks, in particular those set up amongst financial institutions, depart from this logic and restrict access to their networks, for instance to members in a specific consortium.[9] However, this is possible only because these networks imply some level of trust amongst nodes.[10] Hence, the ‘permissioned’ model of blockchain networks is different not only in that it requires permission to access. In actual fact, these networks are based on fundamental assumptions different from those of the original blockchain technology.[11]
Notes:
[1] Though this ‘is one of those “I know it when I see it”’, Werbach, n 11, 8, see for a discussion of ‘trust’ ibid, 8-15.
[2] See ibid, 15-16.
[3] Ortolani, n 36, 607.
[4] Nakamoto, n 4, 1; M. Raskin, ‘The Law of Smart Contracts’ (Working Paper 22 September 2016), 7 at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2842258, visited 30 Nov. 2016.
[5] See text to footnotes 67-68 and 85.
[6] See Nakamoto, n 4, 1-4; Wall and Malm, n 4, 5-23.
[7] Nakamoto, n 4, 8.
[8] Nakamoto, n 4, 3; Wall and Malm, n 4, 7; for a critical assessment see De Filippi and Loveluck, n 28, 14-17.
[9] Peters and Panayi, n 3, 6.
[10] ibid.
[11] See ibid, 7.