How language structures reality

For many, aliens only exist in the realm of science fiction. But regardless of whether aliens exist or not, the potential existence of extraterrestrial species throws into question our entire metaphysical framework, which has long gone unchallenged. In particular, we would need to rethink our understanding of language, which currently determines how we experience our world. But what if there were other languages out there that better captured reality? For theoretical philosopher Matti Eklund, these are the questions philosophy needs to be asking.

Suppose we find out how every human language works, down to the last detail. Do we then know how language works, full stop? Obviously not. There can be languages that are not human languages. Maybe some animals use other languages. Maybe extraterrestrials do. Maybe God does. Suppose we find out the truth about every language used, by every language-user. Even so, we don’t necessarily know how language works full stop. What about other merely possible languages could there be, not used by anyone?

There are many kinds of possible differences between languages. Some differences concern surface matters like format. Where we use sounds and writing, other possible language users could use electric impulses or emit odors. But possible differences in format are relatively trivial. Other kinds of possible differences are semantic: they concern meaning. We use certain kinds of symbols with certain kinds of meanings. But might there be languages with symbols with different kinds of meanings? Speculating about this presents certain challenges. In his famous short story “Tlön, Uqbar, Orbis Tertius”, Jorge Luis Borges brings up possible languages without nouns, using verbs and adjectives in their stead. One language he describes has “no word corresponding to the word ‘moon’, but there is a verb which in English would be ‘to moon’ or ‘to moonate’”. The sentence corresponding to the English “The moon rose above the river” (“hlör u fang axaxaxas mlo”) corresponds to the English “upward behind the onstreaming it mooned”. This may be a strange language. But even Borges’ language is still firmly within the realm of known linguistic categories. It is verbs that are central. What might more radically different languages be like?

___

Languages present the world as being in different ways depending on what kinds of meanings their expressions have.

___

In my recent book Alien Structure: Language and Reality (Oxford University Press, 2024), I discuss what possible languages there can be, focusing on the possibility of languages whose expressions have different kinds of meaning.  This has consequences for broader philosophical issues. Languages present the world as being in different ways depending on what kinds of meanings their expressions have. Already the kind of language that Borges describes arguably illustrates this. Some philosophers, prominent examples are Alfred North Whitehead and Wilfrid Sellars,  have held that we should favor process ontologies – according to which the world fundamentally consists of processes – over ontologies that put things at center stage. Some such philosophers, Sellars among them, think languages much like that envisaged by Borges present the world the most faithfully. If there are more radically unfamiliar languages, then they present the world as being in different ways still. We are accustomed to thinking of the world as consisting of things and processes, corresponding to nouns and verbs. It is hard for us to see how the world could be different in kind from this. But maybe that is due only to our lack of imagination.

My own interest in the issue of what I call alien languages is unabashedly theoretical. It relates to these metaphysical issues. What is the structure of the world? Can it be faithfully presented by a familiar language? Or only by a more alien language? Or is the structure of the world ineffable – beyond the reach of all languages, both alien and familiar? There are other possibilities too. Maybe the world does not have a privileged structure, and familiar and alien languages describe the world equally well. But considering the possibility of alien languages also has other kinds of applications.

Consider comparative linguistics. What sort of diversity is there already among human natural languages? One view is that they are all fundamentally similar semantically, with only minor variations. Some linguists have argued that some natural languages, among them Riau Indonesian and Salishan languages, lack a noun/verb distinction. But even such descriptions take familiar linguistic categories as primary: those languages are described as lacking a distinction drawn in familiar Western languages, and not described in positive terms as having something that these familiar Western languages lack. Might there be more radical diversity, going beyond what tends to get envisaged?

Going beyond humans, some theorists speculate about what communication with extraterrestrials would be like. Some such work is found in the recent collection Xenolinguistics: Towards a Science of Extraterrestrial Language (Routledge, 2024), edited by Douglas A. Vakoch and Jeffrey Punske. Could we and extraterrestrials understand each other? Even if the extraterrestrials in fact do use a language not semantically different in kind from ours – they too have nouns, verbs and the rest – achieving understanding could be very difficult. How do we identify what their words mean? But another layer of difficulty is added when we take into account that maybe the extraterrestrials use an alien language. How do we even know they use nouns, verbs and the rest in the first place?

___

It is reasonable to speculate that AI systems could come to develop languages of their own.

___

Returning to Earth, consider the development of AI systems. Such systems as they actually exist are trained on human language and their outputs are strings of human languages. Or that’s the way it appears. Maybe they just look to be. But let me not get into that. However, the “minds” (if one can use that term) of these systems do seem importantly different from human minds. It is reasonable to speculate that AI systems could come to develop languages of their own. If they do, those languages would be very different from human languages, including semantically different. (Maybe some AI systems already do develop their own languages? See here and here.)

The different questions and speculations just presented assume that semantically alien languages are possible. That is also what I believe. But some might want to resist this. Even agreeing – as everyone certainly should – that some languages are very different from familiar human languages in superficial ways, they could insist that deep down, all possible languages work the same way as far as semantics is concerned.  No matter how different from us some extraterrestrials or some AI systems may be, such theorists would say that if these things speak a language at all, they speak a language with familiar semantic structure, for there are no other kinds of languages.

___

What kinds of meanings can different representations whether linguistic or not have?

___

However, the underlying issues go beyond language specifically. Even creatures that don’t use a “language” might use some other kind of system of representation. We could imagine creatures just using pictures and maps to communicate and to represent the world, and depending on how one uses the label “language”, they might then do these things without using a “language”. We can therefore raise the questions I have raised about language to also regarding these kinds of representations, and concerning both format and semantics. What kinds of meanings can different representations whether linguistic or not have? What unites the familiar kind of representations, and what other – alien – kinds of representations can there be?

Philosophy often aims to ask radical questions and to shake things up. One way to do this is by considering alternatives to familiar languages. Certainly philosophers have considered some such alternatives, both recently and throughout history. But there has not been anything like a general, systematic discussion of what other, semantically different kinds of languages there can be, and the philosophical consequences of this. If reality has a certain structure, it would be a miracle if familiar languages contain all the resources to capture this structure. Investigating what alien languages there can be is a natural and fruitful way to investigate different ways for the world to be.