Data is not Oil or Gold or Labour or anything else…
Words are a symbolic linguistic invention. When we use words we invoke the larger concepts and meanings they represent – they are a shortcut to more detailed explanations. However, this means that they often lack context and relationships. Words are units of data that require the addition of meaning derived from context to inform the listener and become information.
The average word can mean, or be interpreted to mean, many different things depending on context and relationship. The 2019 update to the New Oxford Dictionary adds the words ‘agender’ and ‘intersexual’ to enable more nuanced conversations about sexuality and gender identity. Increased subtleties in terms and definitions help to avoid conflict and confrontation in our society.
Words might allow us to explore and debate wider and deeper concepts, but their misunderstanding leads to disagreement and turmoil as much as innovation, problem solving and creativity. When we don’t have a word for something we need to spend a lot of time using metaphors to add context and relationships. For example, how did we describe ‘competition’ before the word ‘competition’? In 1996 Nicholas Negroponte wrote a book called Being Digital – he spends an entire chapter explaining Broadband, another on Social Media, and yet another to explain what ecommerce is.
All fields create their own language to explain the order of things, from economics, biology, physics, psychology, maths and beyond. However in each field, words have certain limitations and bring with them assumed context and relationships, which means when they are applied to new areas they may not adequately describe the matter in hand.
Speaking of data
We lack words to describe many new activities, models and functions in a data driven digital world. Our current word-set may constrain us and slow us down because of ambiguities inherent in the words we currently use. As an example, when we talk about the internet, we say there are ‘sites’, ‘domains’ and ‘locations’ that we visit, thus framing the internet as real estate. Yet when we speak of pages that we ‘author’, ‘publish’ and ‘syndicate’, we are framing the web as a publishing system. When we speak of ‘content’ comprised of ‘packets’ that we move, upload, download and ‘store’ with ‘addresses’, we’re framing the underlying infrastructure as freight forwarding between storage facilities. Analogies such as these inevitably have their limitations.
The word ‘data’ is a particular problem. This is a word that we want to constrain by context and relationships, but ‘data’ does not comply to the same boundaries, field, domain, graph, market or constraints as anything else.
As much as we would like to explain data and its functions with a metaphor or analogy – it is unique. The term is comparable to the discovery of a new core element for the periodic table, a new energy concept for quantum, or a new model for dark matter.
Every model we use to explain data fails in some way. Data is not oil, data is not gold, data is not labour, it does not pass with time. You cannot declare ownership of data (though many people try), you cannot control it, you lose nothing when you copy it. Generically data is not a commodity. Commodities, at least of the sort that get bought and sold in stores and in commodities markets, are both rivalrous and excludable by nature. Data is, by its nature, non-rivalrous and non-excludable.
Metaphors and a new terminology
While ‘time is money’ is perhaps not literally true, this metaphorical frame makes sense to us because our experiences of time and money situate them as valuable commodities. As explored in The Mind is Flat by Nick Chater, our brains are built to create and make sense of things, but we need these metaphors to make the jump and move us forward.
Our digital world is too radically new and different to be fully conceptualised, understood, or explained by the metaphors we apply to them. The words and metaphors we use for the internet, web and data are not sufficient – and that’s a problem.
Creating value from data requires an entirely new set of words. The concept “data is oil” breaks down – as do the likes of data storage, data consent and data analysis – when put into context and relationships. Data storage is not the same as it was when we had an economic model for the storage of documents in 1980. In 2019, digital data storage has a relationship and context with security, access, rights, liability, control, sharing, conflicting national compliance laws and privacy changes. In spite of this, we continue to use old economic framing, thinking and words to describe these new data functions that then fail.
As Doc Searls puts it:
“We are now digital as well as physical beings. Our habitat as digital beings is very new, strange and has no history so we are forming new human experiences, even though we live in a digital world almost as much as we live in the natural world.”
Data ownership – learning from history?
From the 10th to the 21st century our thinking was built on words based on an economic model physically limited by space and time. During this period relationships could be discovered, explained and modelled. In our new data world the word sets that we previously used to describe the constrained physical are holding us back. The world of data is not limited in the same way as the vocabulary we have developed to understand it. Since the objective value of a word is to create a shortcut, our new data world needs new words to describe its new functions. These should reflect its status as messy, interconnected, interdependent, relational, immediate and feedback driven.
One area that needs discussion and a better term is ‘data ownership’. If we had a better word to describe the context and relationship of this concept we could save pages of debate.
Can you actually own data? It would be good if the answer was ‘yes’, however in reality it seems to be ‘no’.
This being said, you can own the machines and software that store data, as well as the mechanisms that allow different players different rights of access to it.
In fact, the non-rivalrous nature of data plays havoc with modern notions of ownership. The Romans had a much subtler understanding of the nuances of ‘ownership,’ when they created separate legal rights and processes for ‘usus’, ‘fructus’ and ‘abusus’.
Usus (use) was the right to use or enjoy a thing directly, without altering it. For example, to walk on a piece of land or eat a fig off a fig tree. Fructus (fruit, in a figurative sense) was the right to derive profit from a thing possessed: for instance, by selling crops (but not the land on which they were produced), taxing for entry, etc. And abusus: (literally abuse) was the right to alienate the thing possessed, either by consuming or destroying it or by transferring it to someone else (e.g. sale, exchange, gift). These notions when applied to territory, not ‘private property,’ show that different rights apply inside and outside clearly delineated boundaries.
When the 18th century constitutionalist William Blackstone observed that an Englishman’s home was his castle, he wasn’t talking about absolute rights of private property. Rather, he was talking about Englishmen defending a piece of territory where they were safe. Englishmen didn’t just have their castles, they also shared the fruits and benefits of commons, public rights of way and so on. Each of these different territories had different rules and rights associated with them. In contrast to this subtle ecosystem of rights and responsibilities, Blackstone characterised modern notions of private property as “the sole and despotic dominion, which one man claims and exercises over the external things of the world, in total exclusion of the right to any other individual in the universe.”
A new data age
In a new data age we need to establish a new concept in digital and data that builds appropriate boundaries each with its own rules, rights and responsibilities. Critically, individuals’ rights to ‘usus’, ‘fructus’ as well as ‘abusus’ in relation to their own data need to be clearly delineated. This is very different to current debates about ‘control’, virtually all of which relate to individuals trying to control what other parties do with their data rather having the right and ability to use their own data for their own purposes.
Data today has many different definitions, each with individual labels and biases. However, if we are specific, we can be clear about they type of data to which we refer – flat, big, meta, real-time, old, static, new, current, statistical, empirical, computer, binary, or linked, for example. That noted, not all data is created equal and as such data is contextual to where value may lie. As yet we have not been able to add context to types of data such as rights, ownership, providence, trust, privacy, security, faithfulness, or accuracy. But, when we can, data will become more valuable than ever before.