This is an expansion of thoughts on terminology and why we need new words. The following is a dialogue on the original consensus problem: agreeing on what we mean. You can view the full artifact here.
Characters
- W: a wordsmith, a poet, a pluralist
- P: a pedantic
P: W, have you ever wondered we need so many words?
W: I can’t say I have. The whole point of language is to have shared mental models to enable us to communicate complex ideas more easily. We have complex ideas so having the right materials to express those thoughts feels important.
P: Well, don’t we have enough words to express all we want to express already? There are innumerable ways to combine the words we already do have.
W: There is a form of epistemic injustice known as hermeneutical injustice where one has no labels or common terminology to describe or explain experiences to others. Clearly, there are not enough words to express the human experience.
P: Okay, well I can admit that having more words can be helpful. But who is to say what each word means? Who is to say that the word ‘apple’ even refers to the fruit I just ate, or that ‘metaverse’ is even a real word?
W: Language is logically decentralized, is it not? There is no governing body that determines what a word means. The meaning of a word lies in its use1, and that is decided by those who speak the language.
P: That is not necessarily true though. The language of the law sets out centralized definitions for words so that definitions cannot be swayed or morphed to fit the needs of its wielders. Similarly, at the start of a mathematical proof, one should always define the axioms or givens to be agreed upon before attempting to use them in any capacity.
W: There seems to be a flaw in your argument. Yes, I agree that a consistent set of definitions is required for any sort of productive knowledge sharing. But you fail to account for plurality. Context modifies meaning.
P: Plurality as in a multitude of definitions?
W: Precisely.
P: Well that is almost certainly problematic as well, is it not? Too many competing definitions cannot be a good thing. We have dictionaries for a reason. How would anyone new to learning English be able to grasp the nuances of all of the meaning and history behind each term? It would get overwhelming incredibly quickly.
W: I concede that you have a good point. However, we can take a leaf out of Karl Popper’s 2 book: the lie-to-children. We can create simple glossaries and terminological definitions as some abstraction for the larger, more nuanced concepts of the real world.
P: Wait… lying is categorically bad is it not?
W: Hold on and let me finish my thought. According to Pratchett3, “a lie-to-children is a statement that is false, but which nevertheless leads the child’s mind towards a more accurate explanation.” I think what I’m trying to get at is this concept of verisimilitude: that there is no binary true or false, some propositions are more true than others, especially in context.
P: Okay, I accept that taking a toy-model approach to new terminology makes sense. But I still don’t understand why we couldn’t just have a single definition for each word in this model.
W: My concern with a ‘standard’ library of definitions is that interpretations of concepts are not allowed to adapt to usage over time. How does a dictionary or thesaurus adapt to the historical meaning of racial slurs, for example? Those who hold the ability to rewrite the past are the ones with power. Again, this is why having centrally controlled language is a bad idea. This was quite clear in Orwell’s work.
Of course, this is not to say that definitions are not important. In fact, quite the opposite. Definitions hold immense power in shaping how we talk and think about the world writ large. What I am saying is that language needs the ability to evolve on its own. Individuals and groups should have the agency to ‘reclaim’ harmful and outdated definitions.
P: Isn’t this exactly why we need centralization? This is like Plato’s Ship of State4. Any large vessel by their very nature needs to be steered firmly. Those aboard must yield to their captain’s commands; no reasonable person believes that a ship can be run democratically.
W: Would sailors want to obey a captain who takes them somewhere they don’t want to be? No. I imagine language like multiple small ships, each one with its own crew (crew of course, meaning an agreed meaning for a set of definitions). Let me reiterate without the ship metaphor: language requires localized consensus for it to function.
P: Ah, now I see that you don’t mean complete decentralization, that makes a lot more sense. I can agree on this. I am, however, curious about how local agreement should propagate to become widely accepted. After all, almost all English speakers can agree and what an ‘apple’ refers to, yet nobody seems to have a good definition for relatively newer terminology like the ‘metaverse’.
W: This is an interesting question to think about. Not only do we need to think about spatial locality, but also temporal locality as well. I’d like to think about this in terms of metallurgical annealing if you’d let me.
P: As in the process of heating metal to make it more malleable?
W: Yes, exactly! Let us imagine definitions of words as metal nodules. After a nodule is heated to a high temperature, it is workable. Similarly, when new terminology is coined, it has a period where it is malleable and adaptive and achieving local consensus is relatively easy.
Of course, over the course of the annealing process, meaning can drift. Centralizing then, is a form of metal hardening and shaping. It anchors meaning and prevents it from being easily modified. Just as one should not have metals be load bearing until they’ve hardened and been shaped for their specific use case, semantically fluid terms should not be epistemologically load bearing.
We imagine new words as small nodules. The size correlates with the number of individuals that agree with the definition. Smaller nodules reach the right temperature for annealing more easily. As the definition becomes more widely accepted, it gains mass and becomes harder to work.
P: Sorry, I don’t follow this metaphor very well. Could you rephrase?
W: I hope you excuse my thinking out loud, but this metaphor has given me the clarity to better express this idea in terms of regular language.
Early on, achieving local consensus on a definition is rather easy; there are only a handful of people who then know about the term, let alone use it regularly. As the term grows more popular, it becomes increasingly difficult to manoeuvre and adapt in such a way that the meaning of the term cannot shift very much without causing fracturing.
P: Ah that makes a lot more sense. Hmm. This covers the initial semantic definitions for a word but is there any way to change the meaning of a definition after it’s been concretized?
W: This seems quite difficult for terminology already ingrained within society. Especially as meaning is not dictated by some central organization, any old definitions need to be collectively forgotten. As we concluded earlier, this is incredibly difficult for widely used terms.
As we enter an age of digital permanency, we should normalize the right to be forgotten for terminology. It should be normal for terminology to be forgotten — for it to slip through the hands of time like sand.
P: This ignores a lot of history, does it not? “By changing what we were, you change what we are and what we are going to be.”5 This is a form of erasure through terminology change. By ‘forgetting’ terminology, you deny its existence.
W: I may have phrased my words poorly, that was not my intention. Maybe abandoned is a better word? I want to create dictionaries and glossaries that keep terminological history. A sort of ‘append-only’ record of how terminology has split, died, and evolved over time.
P: What do you mean by append only?
W: The only way to overwrite an existing definition is to either hard-fork it within a subcommunity or to create entirely new words. Hard-forking involves an agreement from a subcommunity to use an alternative definition for specific terms.
P: Well, what words need to start anew? Clearly there are words that could benefit from redefinition.
W: The ‘metaverse’ for one could use a lot of redefinition. Meta has put a lot of effort into claiming and defining this term for their own benefit — a closed, profit-driven, and attention-farming dystopia. A few groups have cropped up around reclaiming some of this terminology, including one forking John Perry Barlow’s 1996 Declaration for the Independence of Cyberspace as a reaction to Facebook’s recent rebranding as Meta.
Likewise, many groups across the world have tried to redefine and hard-fork the definition of ‘hackers’ for quite some time. Hacking, as known by the general public, generally refers to gaining unauthorized access to technology. Yet within this local hacker subcommunity, a hacker is widely defined as one who builds and creates for the sake of creating.
P: I am increasingly convinced by your argument against centralized definitions. I think, as a society, we need to think more critically about language and terminology and how they carry power. We want to enable evolution and creation of new terminology to enable others to have the language to speak of their lived experiences and complex ideas.
W: Yes! In fact, I think we need a new approach to building glossaries and dictionaries. At the core of it all, we are hoping to better solve the original consensus problem: agreeing on what each other mean. To abolish the existing centralized dictionaries and glossaries without suggesting an alternative would be in bad faith.
P: What sort of alternative are you proposing?
W: A collectively curated glossary of sorts. One which involves a rich history and context of terms, pluralist in nature, and always ongoing. If meaning is a negotiation, then the history of that negotiation needs to be a crucial part of achieving consensus on the meaning.
Let us create a new glossary of terminology. May it be a source of truth rather than the source of truth.