How do we come to know the world around us? Language. From the moment we are born, despite not understanding it, we are spoken to in a language. Humans are constantly engaged with language, through communication with others and with our own thoughts (just try thinking not in a language). There are other forms of expression, but almost always these are interpreted, understood and reformulated in terms of language. No other means of communication comes close to language in terms of efficacy.

The centrality of language has become an imposing notion within the last 150 years. Language has become such an important topic that the shift of attention was given its own name – the linguistic turn. Starting with Gottlieb Frege in the late 1800’s (although there are notable philosophers of language prior to this), the focus shifted from the concepts supposedly referred to by language, to the language itself. Instead of focusing on ‘why’ questions like ‘why are properties universal’, many philosophers focused on what it means to say this collection of terms. In other words, how does the language we use relate to the world we live in? What do our words mean? This approach focuses more on the means of communication than the content being communicated.

From Frege, a number of very different, often conflicting, theories about how the language we use relates to the world we live in have surfaced; there are the logical empiricists, who (very roughly) believed that they world stood in a 1:1 relation to the words we use (words are labels for things), there are the possible world theorists who (very roughly) believe that words correspond to a certain way in which the world could be, there are the constructivists, the Griceans, etc.

In this article I will focus on one sliver of the philosophy of language known as Quinean constructivism. William Van Ormand Quine was a mathematician, who later turned to the philosophy of language. He is credited as one of the most influential philosophers of the 21st century and his ideas on the relationship between words and their referents directly attacked the ideas of the prior logical empiricists through the denial of the idea of concrete reference. His work is vast, thorough and sophisticated - if you find any of this interesting – I highly recommend.

However, in the spirit of short-form, accessible articles, (as well as a relatively vertiginous understanding on my own part), I will lay out his ideas in as plain terms as possible. This article is about the relationship between a Quinean-constructivist account of the world and creativity. Specifically, I will highlight two occasions where markedly creative thinkers of recent ‘pop’ culture and technological history, who have no prior philosophical grounding, have formulated the life work of philosophers of language like Quine in beautifully simple terms. I hope to explain the link between the two and explore how a Quinean constructivist account of the world provides some answers to the puzzle of creativity and its associated dangers.


How do words come to get meanings? How do we decide that ‘car’ refers to motorised, four wheeled vehicles? Or ‘pen’ to cylindrical, ink-filled writing instruments? Do we always know exactly what our words mean? Of course not.

Therefore, it’s not surprising when the meanings of terms are not concrete. For example, ‘unjust’, ‘good’ or ‘evil’ are all words that we tend to realise as unknown in meaning – not everyone has the same idea as to what they mean. This is the foundation to normative theory, and something we all encounter in everyday life. The point to be made here is conceptual; what does this ‘unknown-ness’ tell us about meaning. I think that there are two possible answers. 1) maintain that there is a concrete meaning to the terms, it is simply unknown, or, 2) accept that the meaning of these words are fundamentally non-concrete, there is no way to know the concrete meaning because a concrete meaning does not exist.

For the philosophers amongst you, this is instantly recognisable as the debate between subjective and objective morality. I am using it as an example because it contains nice similarities to the next example which may be less easily recognisable.

It may be surprising to find that the meanings of even the most ‘definite’ terms are also not concrete. Terms like ‘car’, or ‘pen’, or ‘purple’. Surely the meaning of pen is something like ‘a cylindrical, ink-filled writing instrument’, car means ‘a four-wheeled, motorised vehicle’, purple as ‘that mix of the colours red and blue’. Although these definitions could certainly be sharpened, there must be a definition that eventually captures all things that we call pens, cars, and purple. Right?

Wrong. In the mid-20th century, there was an intense debate as to what counts as a car due to disputes over road safety in the United States. One farmer put a sofa on four wheels with a generator as a motor and argued that he need not abide by road laws (or pay tax) because it’s not a car, it’s a sofa. The same goes with pens, if I dip a car in ink, and use a crane to sign a comically large cheque, is the car a pen? Is red a property shared between all red things or do all things that look red have a different kind of red-ness.

These are somewhat sillier examples than the ethical one – but it shows that even the most certain of referents can be debated. The problem is a problem of definition. We can always paraphrase what we mean by a certain term, and this process of defining and paraphrasing cannot terminate in a concrete answer. There will always be a better/different definition, another paraphrase that will, supposedly, capture the meaning of the term fully. And after that, someone will come along and raise an issue that challenges it, and a new definition will appear (this is what Susan Stebbing was getting at when she distinguished same-level and directional analysis).

The constructivist has a great answer to this. The meaning of our terms is derived from behavioural dispositions to certain observable criteria, much in the same way that science proceeds. Meaning exists only insofar as it is a shared assumption that the sense experience that you label with the term ‘pen’, is the same as the sense experience that I have when I use the term ‘pen’.

Quine’s radical translation thesis captures this well. If I have the sense experience of something and I label that ‘rabbit’, and you give an affirmative response to what I assume is the same sense experience, then we can both proceed with that sense experience now denoted as ‘rabbit’. This is the example Quine used when talking about a ‘field linguist’ attempting to translate the language of a previously uncontacted tribe. The tribes people say ‘gavagai’ when they see a rabbit bouncing along the plains. Through several occasions in which the rabbit is observed, the field linguist asks ‘gavagai?’ and the tribe responds in the affirmative. The field linguist now rights down ‘gavagai=rabbit’.

But what if gavagai refers simply to the rabbit ears, or the act of a small mammal bouncing across the plains. How could we ever be sure that the translation of gavagai relates to the same thing we mean when we say rabbit? Quine applies this indeterminacy to our own language. When I say rabbit and you say rabbit – are we referring to the same thing?

It's entirely possible that we aren’t. It depends on your theory of properties for one. Does rabbit refer to anything with rabbit-ness, or does it refer to a certain biological composition? It also depends on your theory of material objects. Do objects persist over time? Does rabbit refer to all time-slices of the rabbit or just the present rabbit? We can only define rabbits insofar as they are certain experiences that we label with a term, we cannot be sure that our meanings of the term rabbit are the same.

(This may seem like a philosophical mind game, and for all intents and purposes we can be almost certain/certain enough that we mean the same thing. But the fact we cannot be sure reveals something about the nature of language and its relation to the world).

We are back at the problem of definition. There is no concrete reference, even in our own language.

So how do we come to know the world? Well, we do so exactly through the process above, and it works very well. There are no pens, rabbits or cars outside of language. We simply experience certain stimuli and label them with a certain grunt. From here, we build unbelievably complex worlds of different grunts that we assume refer to the same thing. We can also be as certain of this as we can be of anything, but it does not change the fact that reference is indeterminate. When we start to deal in incredibly abstract concepts and the resulting terms, the insufficiency of language as a concrete referring device becomes clear.

Avoiding Scepticism

Constructivists have been charged with simply repackaging cartesian, external world scepticism in a modern style. ‘Why are you doubting the existence of pens and cars?!’. Constructivists are not doubting the existence of pens and cars, they are doubting that there is concrete meaning to these terms. Pens and cars are just linguistic entities, and therefore they are indeterminate. The sense experiences of a pen and a car are not indeterminate, but they can be labelled with different words (see my car-pen example earlier).

Quine spoke specifically about this. As an empiricist and mathematician, I’m sure he was irritated by charges of scepticism. He was concerned with what there is (ontology), and he asserted that linguistic entities do not refer to definite pieces of matter. This does not mean that we should reject the common-sense ideas about reference. For example, we should not start calling cars pens, or pens windmills, or windmills aeroplanes. This would be absurd and would basically spit in the face of the complexity and effectiveness of language as a way of organising the world.

However, as unsettling as it may be, that is what language is - a way of organising the world. Fortunately, much more often than not, it does more than a good enough job, and so we need not slip into scepticism about the external world.

Constructivism and Creativity

A constructivist account of language, and consequentially the world, can provide some answers as to the nature of creativity/invention. At this point, we can say the following: humans, through a process of sharing and recognising behavioural dispositions to sense experiences, construct a system of words that represents the world around them. We posit coffee tables as coffee tables because when someone says coffee table and points at, well, a coffee table, we both nod in agreement. Therefore, there is nothing about the collection of matter that the coffee table represents that makes it a coffee table beyond our naming it so.

It’s important to understand how we come to know the world we live in. We do so through a socialised, conditioned use of language, and so our world is shaped by the words we use. It really is shaped by language. We only think of coffee tables as coffee tables because we have learnt what coffee tables are. For most of the time, this is nothing but a benefit to humans. Without it, we would be in an ocean of confusion; this constructed worldview resulting from language is likely the most important element of human life. It is quite impossible to think of life without it.

But creativity occurs when this conventional worldview is challenged. When a person formulates the world in a way that is different to the world constructed by language, they construct a new world in which things do not mean what they normally do. While this may be brilliant for creating art, inventing new products, or revolutionising the way we live, there is an inherent danger. If you spend all your time trying to think outside of the world constructed by language – then you risk losing the positive benefits of such a world.

For a long time, people have wondered why the best artists ‘go crazy’, why are creative people seemingly more prone to mental illness? One can imagine that if you are constantly rejecting the constructed worldview in the search of creativity and invention, then you may start to lose the fundamental, guiding elements of the constructed world. Although the world is constructed by language, it is important to remember that this construct serves a vital purpose. To reject it entirely is to lose yourself in the ocean of confusion, and to unsurprisingly find yourself in very murky waters.

A two-year-old jumps on a coffee table and someone says, “Don’t jump on that – that’s a coffee table”. By the time you’re 40 years old you have a world full of coffee tables.

(Ye, formerly Kanye West, 2019)

When you grow up you tend to get told the world is the way it is, and your life is to live your life inside this world and not bash into the walls too much.

(Steve Jobs, 1994)

I’ll admit I was a bit nervous to include a quote from the musician Ye. I do not want to imply that I support any of his bigotry or harmful ideas. Regardless, his quote captures, quite incredibly, the subject of this article and the concept of constructivism. Further, you will see that his character fits well with the point I make towards the end.