Blockchain as translation’s new foundational technology

January 13, 2018 Localization Stuff

There is sustained excitement in the translation space about deep learning and neural machine translation. Yet through this lopsided focus we risk ignoring an equally momentous emerging technology, blockchain.[1] While the less tech-savvy press broadly associates blockchain with cryptocurrencies such as Bitcoin, I argue that it has the potential to redefine the entire value chain of linguistic effort, with far-reaching consequences for compensation, compliance, workflows, and tools.


Watching TV by candlelight: ICOs

Characteristic of the short-sightedness of “innovators” looking for an effortless way to disrupt established industries is the recurring noise emanating from Initial Coin Offerings, or ICOs.

These are the cryptocurrency equivalents of Initial Public Offerings, or IPOs. Two recent examples from the translation industry are MiTio and Langpie, both reported by Slator.[2][3]

An ICO, however, takes a painfully narrow view of blockchain, seeing nothing else in it but a currency vehicle. Without a vision, the funds raised through an ICO would get mobilized in businesses with an appallingly conservative business model. Or, in the worst case, the entire ICO turns out to be a scam, as it happened with Langpie.[4]

Blockchain fundamentals

To arrive at a more substantial of understanding of the potential of blockchain, let’s take a look at this technology’s fundamentals.

A blockchain is effectively an open, distributed ledger that contains a cryptographically solid record of transactions. Its power lies in the particular ways it achieves this. Often, these ways go against the very grain of our intuition about digital information.

There is no central authority to verify authenticity. The distributed ledger exists simultaneously on an arbitrary number of nodes, which achieve distributed consensus via mass collaboration powered by collective self-interest. The traditional model of trust is replaced by a peer-to-peer network with high Byzantine fault tolerance.[5]

Although the distributed ledger contains a trace of every transaction in the system, participants can choose to retain their anonymity. These privacy guarantees are underpinned by the use of formally verified public-key cryptography.[6] To compensate for the accretion of data from a massive number of transactions, closed chapters of the past are ingeniously compacted into Merkle-trees.[7]

Most intriguingly, blockchains achieve digital scarcity. All other types of digital information can be copied any number of times at virtually no cost. Blockchains, in contrast, remove infinite reproducibility from a digital asset. This is the very property which enables them to support a currency, by solving the problem of double spending.

A plea for a global translation ledger

The characteristics outlined above open up exciting avenues for the application of blockchain technology in translation. Blockchain holds the promise of solving many of the industry’s pivotal problems.

Intellectual property and compensation. The status of translations as intellectual property is woefully unclear. While we have learned to live with this issue by basically ignoring it, the uncertainty continues to hamper our ability to assert the value-add of linguistic effort.

On a trivial level, it impedes knowledge pooling: no attempt at establishing large crowdsourced TMX repositories has been commercially successful.

More importantly, it has created a loophole through which intellectual property leaks to external players on a massive scale, without any compensation for the creators. The machine translation services offered by the largest providers all recycle the effort of human translators, gleaned from the internet by robotic crawlers.

Blockchain has the ability to assign title rights. Other industries are already discovering this, as evidenced by DJ sets[8] that pay artists immediately when a track is played. A similar example is Cryptokitties,[9] a blockchain-based virtual game that allows players to adopt, raise, and trade virtual cats.

Embedding translations in a blockchain opens the possibility to claim intellectual property rights in perpetuity. Because of blockchain’s cryptographic guarantees of privacy and anonymity, this can be decoupled from the source content itself, whose copyright belongs to its creators. Ultimately, individual translators can be compensated in the global translation ledger for every translation synthesized by a public MT engine trained on a corpus that contains a sentence they translated.

Quality consensus. Since time immemorial, the translation industry’s focus on quality has bordered on the obsessive. Yet the nature of the problem is such that no single truth has emerged. In practice, quality continues to be defined in terms of consensus.

That is precisely where blockchain yields itself as an obvious facilitator. The technology is effectively a large-scale implementation of multiversion concurrency control.[10] Decentralized social networking[11] is a shining proof of blockchain’s potential in this area.

Intermediation. If you rephrase the role of the distributed ledger, it is a record that compels offer and acceptance.[12] From that perspective, it is a perfect mirror of the services provided throughout the translation value chain, and of the corresponding compensation sequence. This suggests groundbreaking changes in the way supply chains are structured, and promises major exploitable efficiencies in the way jobs are assigned, delivered and accepted. Additionally, blockchain removes much of the complexity incumbent in today’s cascading project management systems and compensation schemes.

Compliance. A large part of translation work happens in regulated industries such as pharma, legal, or medical device manufacturing. Existing protocols for change tracking, authorization and auditing are extremely complex. They rely on decade-old technology such as change-tracked Word documents, and they add a massive overhead in terms of administration and verification.

Blockchain allows the integration of off-chain oracles[13] as well as validators vetted by the network owner.[14] To ensure the integrity of the audit trail, a solution can integrate centralized timestamping servers operated by the regulating authority. If the authority’s jurisdiction spans multiple time zones, the cryptographic hashes can easily be normalized to UTC.

Open issues

Admittedly, we cannot adopt blockchain technology in the translation space before resolving a few outstanding questions. These stem from the idiosyncrasies of the language industry. Since blockchain has not been considered for this domain in the past, there are no tried-and-tested answers.

The most obvious dilemma is the choice of consensus method. We must explore proof of work, proof of stake, proof of authority and even proof of burn as options. The latter might sound like an unlikely choice, but if we consider the likely influx of venture capital, proof of burn has great appeal. It would facilitate burn rate management and runway forecasting for the coming generation of VC-funded translation startups while they pivot towards product-market fit.

Many of the benefits from the previous section obviously imply a blockchain implementation that supports smart contracts. While existing frameworks like Ethereum Solidity already incorporate a Turing-complete scripting language, the fact remains that due to lack of widespread use, the legal status of smart contracts is unclear.

Finally, great care must be taken in mining the translation blockchain’s genesis block. This is the block that will serve as the global translation ledger’s starting point. While orphan chains are by definition resolved through ad-hoc message passing, multiple initiatives in the past have been faced with the dilemma of a hard fork.

A foundational technology

Iansiti and Lakhani assert in the January 2017 issue of the Harvard Business Review[15] that “it will take decades for blockchain to seep into our economic and social infrastructure.” They also add, however, that “it has the potential to create new foundations for our economic and social systems.”

I share the second assertion, but I also believe that the time needed for all-out adoption depends largely on how forward-looking the stakeholders are and how ready we are, as a community, to embrace an unprecedented opportunity.

As I showed above, the translation industry is uniquely positioned to benefit from blockchain. On top of the industry-specific synergies, blockchain’s universal advantages also hold true: there are no centralized points of vulnerability, and there is no centralized point of failure. What else can a profession hope for that is as geographically and culturally diverse as translation?

Make no mistake: as a foundational technology, blockchain is not meant to disrupt an entire industry overnight. But for those of us who are in it for the long haul, being among the first to embrace it holds the promise of exponential return in the future.


This entire text is made up of 80% baloney and 10% cohesive material. I recently read the Wikipedia article about blockchain[16] and was baffled by the density of jargon, naïveté and starry-eyed idealism there. I went on to read the original Satoshi paper,[17] then returned to Wikipedia. I was still just as baffled as the first time.

All the outlandish phrases come straight from the Wikipedia article, and so do most of the references. I mixed in a bit of translation industry jargon for couleur locale, added some @Managerspeak,[18] dumped it all in a blender, and decanted the goo into a makeshift story arc.

For every fancy phrase here, there exists some context in which it makes perfect sense. With near certainty, such contexts are not to be found in this here text.

And yet, who knows, maybe by pure chance there is an unintended grain of truth somewhere in here. I guess that a similar article 25 years ago about how you could use TCP/IP for business transactions would have had the reader believing it was all obviously in jest.

And sorry about mixing those cute Cryptokitties into this business.






[5] Lamport, L.; Shostak, R.; Pease, M. (1982). "The Byzantine Generals Problem". ACM Transactions on Programming Languages and Systems. 4 (3): 382–401.

[6] Alfred J. Menezes, Paul C. van Oorschot, and Scott A. Vanstone (October 1996). "11: Digital Signatures". Handbook of Applied Cryptography. CRC Press. ISBN 0-8493-8523-7. Retrieved 2016-11-14.

[7] R.C. Merkle, "Protocols for public key cryptosystems," In Proc. 1980 Symposium on Security and Privacy, IEEE Computer Society, pages 122-133, April 1980.

[8] Kastelein, Richard (24 August 2017). "The World's First DJ Mix That Pays Artists in Seconds Using Blockchain Technology". Blockchain News. Retrieved 2017-09-03.

[9] CryptoKitties craze slows down transactions on Ethereum. BBC News, May 12, 2017.

[10] Greenspan, Gideon (19 July 2015). "Ending the bitcoin vs blockchain debate".

[11] Decentralized social networking? Forbes, August 7, 2017.

[12] "Blockchains: The great chain of being sure about things". The Economist, October 31, 2015.

[13] Project Bletchley Whitepaper, Microsoft, 2016-09-19.

[14] Prisco, Giulio: "Sandia National Laboratories Joins the War on Bitcoin Anonymity". Bitcoin Magazine. August 25, 2016.

[15] Iansiti, Marco; Lakhani, Karim R.: "The Truth About Blockchain". Harvard Business Review, January 2017.