The chicken of language and egg of reality

Words are powerful, but not all-powerful in shaping our thinking

Part 1: Language warps reality

Language can affect how we perceive the world. But where is the limit?

People judge a car in a traffic accident to be about 10mph faster if they are asked at what speed they ‘smashed’ vs. ‘contacted’ each other. Famously, Russian speakers are better at telling apart different shades of blue because they have two separate colour words for light and dark shades of blue that don’t exist in English.

No doubt these effects are fascinating. They show how intimately our perception of the world is warped by how we use language.

These insights have also energised cultural movements in the past decades to fight battles over the use of language. In the past years, words such as ‘blacklist’ or ‘master/slave branches’ were systematically purged from the collective vocabulary, in an attempt to remove their perceived damaging connotations from public discourse.

These efforts almost admit language to almost have a runic quality. As if the utterance of words themselves can act as incantations to uphold or destroy cultural realities. It’s no coincidence that we picture God to have created the universe by speaking it into existence.

That is why ownership of language is the holy grail of every power-hungry totalitarian regime. This has been famously branded into our cultural consciousness by George Orwell, who painted the devastating effects of Newspeak on free thought among the citizens of Oceania in 1984.

It is equally no surprise that in the world of marketing and branding, there is an obsession with using the right words and names, too. When you are ‘googling’ or ‘zooming’, these brands have achieved the highest form of brand awareness by becoming part of our everyday language.

Owning language seems like owning real estate in our cultural consciousness.

Part 2: Reality warps language

But there is also a different way of looking at the power of language. What if language dovetails with cultural truths only then, when these are already known to be true another way?

In an experiment, Spanish and German speakers were asked to imagine that inanimate objects, like keys, apples, or chairs could speak. Then they were asked if the voice was male or female.

Spanish speakers imagined objects with a feminine grammatical gender to mostly have feminine voices and vice versa. “La pluma” (feather) was thought to have a feminine voice, and “el tornillo” (screw ) was imagined to have a male voice. This finding plays perfectly into our growing cultural suspicion that the gender coding of language deeply influences our perception of gender in the world.

But surprisingly enough, this was not true for German speakers. Although the German language is much more strongly and consistently gendered than English or even Spanish, German speakers assigned male and female voices to objects that had nothing to do with their grammatical gender.

But here is the most surprising finding: The genders that German speakers imagined the objects to have almost perfectly matched the object’s grammatical gender in Spanish. In fact, French and Italian speakers equally ‘agreed’ with the Spanish language on the gender assignment for inanimate objects.

The most likely explanation is that German is using grammatical gender almost arbitrarily. ‘Woman’ (Frau) is feminine, ‘Girl’ (Mädchen) is neuter and the three items of cutlery (knife fork and spoon/ Messer, Gabel, Löffel) all have different gender without any obvious orthographic reasons for these differences. Spanish, on the other hand, seems to have developed a much more semantic relationship with the world it describes. The actual ‘femininity’ or ‘masculinity’ of objects influenced the grammatical gender that was assigned to them in language.

A guitar, with its curvy, gentle features strikes us as more feminine than masculine, regardless of what language we speak. That’s why it’s ‘la guitarra’ and not ‘el guitar’. Most people would agree on this, regardless of the language they speak.

This makes the relationship between features in language and features in the world more complicated. The power seems not to be in the word itself, but in how we use it. German speakers seem to experience grammatical gender as a mere linguistic technicality, precisely because it does not meaningfully map onto any perception of gender in the world. This is different for Spanish speakers. Their language developed to harmonise with the semantic qualities of the objects it describes.

This shows that language indeed sometimes reflects reality, but is not able to overwrite it beyond intuitions that we share that go beyond language, even for something as elusive as the gender of non-living objects.

Part 3: Intentions matter

Somewhere in between the chicken and egg of our thinking and language, there seems to be a less obvious force at play. Language is an imperfect tool in a much more connected and rich toolkit of human communication. There is more meaning in each sentence than just the words it’s made up off. Not because words are so powerful, but because our human minds are so good at creating a rich model of whoever it is that expresses the sentence.

This can sometimes create the impression that simple manipulations of the words we use can drastically change our perception when in reality these changes are perfectly rational if we consider how language is pragmatically used for real communication.

We can see how pragmatic language use explains the power but also the limitations of framing effects. Most people wouldn’t choose a surgery that has a 90% chance of death but would be ok with a 10% chance of success. And meat that is 80% lean seems healthier than if it has 20% fat. Framing is the bread and butter of each advertiser, marketer, politician, and spin doctor because it seems to warp reality just by clever use of language.

But it’s not that simple.

When we think about language, and study it experimentally to come up with noteworthy effects, we are pretending that the words we say are all the information there is. The effect of saying 90% chance of death vs. 10% chance of survival seems so compelling because we believe that they convey exactly the same information, just said in a different way. It is that assumed symmetry that makes us conclude that framing is a language effect.

But the information in these two sentences is not symmetrical. It turns out that most people reading these types of sentences are reading a tacit ‘at least’ before the percentage number. ‘At least’ 10% chance of survival and ‘at least’ 90% chance of death.

In real life, we know that a doctor would be careful with managing expectations and we are sensitive to that nuance, even when presented with seemingly objective information. If we instead make it explicit that the likelihoods are exactly 90% and 10% respectively (by using the word ‘exactly’), the framing effect almost disappears.

When we are using words, we are not just conveying the meaning of that word. We are simultaneously conveying information about not having chosen an alternative word, and about what kind of people we are in doing that. That is why words can take so many different meanings based on who utters them and in what context. It is not the word itself that carries the meaning. It is the explosion of context around it through which we can gauge someone’s intention.

Language is only one way of knowing the world. It is powerful, but not all-powerful. Our pragmatic use of language and the intentions we harbour when we are communicating are much more important than any power that we presume words and language itself to have.

Next time we find ourselves in the trenches of the culture wars, we have to remember: It is not the words that matter. Language on its own has no magical power. Context matters. Reality matters. Intentions matter.