Language is perhaps the most useful social convention, followed by law. Words are exchanged by humans to coordinate action, express thoughts and feelings, and to entertain. It is difficult to imagine much human progress would have occurred in its absence.
Despite the obviously immense value of language, words have no private or physical value, although representations of words can be stored.
The closest economics comes to understanding the value of language is the idea of a ‘network externality’. To my knowledge, Hume is the first to identify this, although his terminology is somewhat archaic. The concept was popularised by the advent of telephony. A telephone’s usefulness depends on its use by others, something distinct to most other goods and services. The classic example is a fax machine. If I alone have a fax machine, it is worthless. If everyone has one, it is extremely useful.
A network externality is a very specific property, with important implications. I can speak and read two languages, English and Italian. Use of these languages helps me to achieve other things, and I also enjoy the use language itself. But a private language would not have these properties. Even a private language with a wider vocabulary and greater efficiency of use than Italian or English. The value of a language is directly related to how many people use that language. It’s value to an individual resides in its acceptance by others. There is a network of users and the value to any individual user is contingent on the use by the network of others. The more users, the more useful. These properties may be partially true of many other things, but they are inherent properties of tools of communication and coordination.
There could easily be ‘better’ languages, with more consistent rules of grammar and spelling. Languages which are easier to read, learn and communicate with.
How do networks become established? This is important because people often confuse how a network is established with the properties of the network. Once established, networks are extremely hard to break – that is why ‘better’ languages face an uphill struggle. Typically the dominant network is established by the state.
The official language is usually taught in schools and required for official documentation. But once established, it will thrive and prove resilient. There is usually a hierarchy of language and subsidiary, coexisting ‘quasi-languages’, with regional variations and slang. The state usually continues to play a role, with literacy targets and standards of official language.
Language is very clearly a social asset. It is often a national asset, but its use spreads beyond national borders. The English language is an asset to many people globally. Is it possible to put a number on the value of this asset? Language has no meaningful ‘book value’ (pun ignored). It should go without saying that this does not render it valueless. Theoretically we could pose the question like this: how much would a society commit to paying if it did not have a language, in order to obtain one?
The other way to see this value is to consider a privately-owned ‘quasi-language’: Facebook. Facebook earns a revenue off the symbolic language it has created. It is a curious business. Linguistic networks are stronger the greater the number of users, so requiring users to pay for usage creates a tension – ease of access strengthens the franchise, paying for access may open the door to competitors. Facebook ‘wants’ you to use its language ‘for free’. So privately-owned languages must generate revenue tangentially – in this instance, by advertising. (The fact that someone else is also ‘paying’ for our access is one of the main reasons the diffusion of social media has been so rapid.)
It should be clear that Facebook trades at a huge multiple of its book value. Book value is an accounting device to proxy market value. Market value is what we collectively believe things are actually worth. In theory, market value is a present value of future cash flows.
Valuing networks is curious because externalities by definition aren’t usually captured by private ownership – it would probably be more accurate to say that Facebook owns the network over which its language is used, rather than owning the language per se.
Either way, it is clear that Facebook has immense value, the internet as a whole must have even greater value – analogous to the printing press. This doesn’t appear in the national accounts, neither does the English language. This suggests that assets are often harder to measure than liabilities. No one even puts a value on our greatest social assets. It also suggests that the value created by the internet is not being measured. A social asset of huge value is nowhere to be found in book values, and the only market value which is measured is that of tangential revenues. For the same reason, ‘secular stagnation’ is almost certainly a monstrous error of measurement. Language never appeared in measured output, neither did the printing press, nor does the internet.
Language and money have something else in common – they are unique phenomena.
A thought exeriment: what happens if you replace the word “language” in this blog with “money“?
After I originally wrote on this subject, the ever-rigorous and thought-provoking Nick Rowe sent me to an article he had written on the difference between an ‘externality’ and ‘strategic complementarity’. Where an externality means if you do something it affects me (pollution is the textbook example), a ‘strategic complement’ is when you doing something raises the marginal utility of me doing the same thing. It strikes me that both of these are properties of language and money. So I will use ‘network externality’ to capture both the externalities and strategic complements inherent in communication and coordination networks.
Thank you for this