Blockchain is a memory transfer sytsem. These memories are moving through space and time. This is something that I recognized. As a type of record, a ledger is a type of record traditionally. There are a number of international standards. One of them is ISO 15489, international records amangemetn. You will see this idea of memory of transactions as evidence. You will see something like "proofs on the blockchain". This is information received and created and maintained. In pursuance of legal obligations or the transaction of business.

As an archival scientist concerned with record keeping, I am fascinated by blockchain.

• ISO 14721
• ISO 15489:2001
• ISO 16175-1
• ISO 16175-2
• ISO 16175-3
• ISO 16363:2012
• ISO 17068
• ISO 18128
• ISO 23081
• ISO 26122

Requirements for proof: reliability, authenticity, archival bond, persistence (semantic, representational, technological). There is a historical reliability nad documentational reliability. How well does the representation handle the real-world facts? Then there's authenticity. That talks about how reliant, how much reliance we can place on establishing and preserving the identity and integrity of the document. This notion of authenticity and non-repudiability and also the persistent integrity of the record from the point of creation to the time thereafter.

Something I'm not hearing a lot about. There's the concept of an archival bond. From our tables earlier, I heard something about not storing actual records, but just storing hashes on the blockchain and having links or mappings. In archival science, we call this archival bonds. Those also have to be persistent. One of the attack vectors or weaknesses could be to rupture the link between the hash and the originating record. You need bonds between the real-world thing and the record and the hash representation. Graph theory comes into play there.

And then persistence, like semantic, representational and technological. Some of the standards are encapsulated in these ISO standards I mentioned, like hte open archival information systems framework for example.

What are the risks? The risks that I see are that if we don't think about record keeping standards, we don't make use of the principles of archival standards, we risk authenticity and reliability and long-term utility and persistence. We want to avoid those pitfalls and take advantage of the opportunities here. We are researching some of thes eproblems. We are looking at http://interparestrust.org/ which is an international research partnership. It's a project called TRUSTER where we investigate long-term preservation of timestamped, digitally signed or sealed digital records in various jurisdictions.

Basically what I'm trying to convey here is that there's an archival science that underpins record keepings. Because blockchain is fundamentally not only a value transfer system and memory transfer system, that record keeping theory (or archival science) is highly relevant to conversations about blockchain, like conversations around identity and provenance that we will be discussing today.