THE LANGUAGE PARADOX
As the story goes, the German organic chemist August Kekulé described the eureka moment when he realized the structure of benzene, after he saw a vision of Ouroboros, the snake that bites its own tail. None of this would have been possible without the use of language. While there is no doubt that the world, like benzene, exists without language, there is equally no doubt that all we, as humans, know about the world, benzene included, is mediated by language. The question is not so much ‘why only us’ as elaborated by Berwick and Chomsky (2016) but what language is. While there have been innumerable attempts to answer the latter, be it by Chomsky (ibid.) or Christiansen and Chater (2018), there cannot be an outcome similar to Kekulé’s discovery. The reason being, as will be argued here briefly, is that language is the only natural phenomenon that has no meta-language, like chemistry has, to describe itself. So, when language describes herself, we have the proverbial snake biting its tail. Sure, one can discern and extrapolate all sorts of rules and conventions that describe language and then try to explain – as elegantly as possible – how such constructs arise from human brains; brains that in all other ways subscribe to the natural sciences as we know them, hence one attempt is appropriately called bio-linguistics. Still, here lurks the mainly philosophical doubt that biology and evolution are mere stories, destined to never come to an end, never to discover the formula that gave rise to language. Traversing dictums by Panini, Plato, Aristotle, Descartes, Humboldt, Russell, Wittgenstein and Orwell, to name but a few, one cannot dismiss Nietzsche’s contention that science too is but a language construct through which we see the world, telling a vitally important story, it seems but not necessarily the only one. Given the present human condition, one must admit that the language of science is losing out to the language of fakery, pointing to the fact that language use is only one aspect of language per se and that language is not only a scientific distillation of human biology. Indeed one can blame science for many of the ills besetting the world today, from climate change to species extinction – all caused by scientific and technological advances dreamt up by latter-day Frankensteins. One commentator recently suggested that anti-vaccers should go back to the Middle Ages – well, at least in the Middle Ages our planet’s ecology was still OK even though human conditions could be quite brutal compared to medical advances of today.
Be this as it may, one thing is for sure, language must be the end all and be all of the human condition. Perhaps the only biological, if not mathematical underpinning, would be, that there is no end of it, that infinite thought, as expressed through language alone, equates with what we assume the universe to be – a convolution of space and time, difficult to imagine even for people who have no limits in their imagination. Contrast this with the other reality of the human condition, one of very limited time and space, a humdrum existence that lowers language to statistical probabilities, matching every utterance to the one recorded before, betraying every thought process as an algorithm to be exploited by the fascists in the know. The everyday struggle to survive - to make a living as the metaphor goes (Lakoff and Johnsen, 2003) – reduces language to the same level, and even then there are escape clauses, like the lyrics of the blues and the songs of Auschwitz.
The current obsession with the evolution of language thereby misses the point, namely that whatever gave rise to language, there is simply no way to explain what language is today by what it was yesterday and eons before. Even the very tenets of biological evolution fail to explain what is going on today: there are no more finches to adapt to new environments because the spring is silent (Carson, 1962). There is something endearingly naive in Christiansen and Chater (2018:43) when they say that language is something like a ‘beneficial parasite’ – when otherwise they are wedded to the theory of cultural transmission and bits of biology – perhaps subconsciously indicating that metaphors are all we have to describe the un-describable. If language were a parasite one would have to conclude that by and large it is not of the beneficial but of the pathological kind. Given the extreme ends to which language is put to, one can only conclude that language, as a cognitive tool, is as neutral as the adze in our hand: it can kill as much as it can create art. Sure, the more we kill the more we will refine the adze. Same with language. To claim then that language arose because we like to kill (Lorenz, 1963) is patently absurd. The current great communicator (the one before him was Ronald Reagan), Donald Trump, uses language that is a great delight to late night show comedians because twittered chunks of gobbledygook masterfully disguise a neo-feudal enterprise. If natural selection is a predictor of evolutionary changes, how come the instinct of the proverbial cave man has survived to this very day, making language useless?
The endless contradictions (Mao Zedong, 1937) that are served up in politics as much as in philosophy and linguistics (Harris, 1993) are testimony to humans putting their thoughts into words, never knowing where both their thoughts and words came from but nevertheless being convinced of their intelligence that somehow resides inside their brains. The possibility that Mao was right and Harris is wrong is the next step that a truly sentient being can contemplate, and even then we may end up with what Shakespeare called ‘signifying nothing’. Really, the best we can do for language is, according to Chomsky (Barsky, 1997), along the lines of the old joke that the drunk looks for his lost keys under the streetlamp because that’s where the light is. And this from a man who does not usually tell jokes. There is a sense that despite all the progress that science and technology has brought us, via language – a progress that evolution cannot explain – the basic contradictions remain: we can send man to the moon and take pictures of a black hole 55m light years from earth but we cannot solve the devastating human problems on earth. As those clever climate change protesters say on their posters: there is no planet B. Language used like this is truly enlightened and enlightening, so that even the drunk can see the key that unlocks Pandora’s box.
So why is all this scientific gravitas employed when arguing who is right and who is wrong, as for example Christiansen and Chater (ibid.) do when trying hard to push Chomsky to the sidelines? Are the cave men fighting for academic territory – àla Lorenz (ibid.)? Neuro-linguistic research may well have progressed to seeing a black hole but, like astro-physicists, they have no idea what is inside. Anumanchipalli, Chartier & Chang (2019) were able to demonstrate neural decoding of spoken sentences, which is no mean feat, but essentially it is reverse engineering. I read a sentence which I cannot speak (due a pathology) and a brain scan at that time can be decoded to produce speech artificially. All of AI is predicated on reverse engineering: matching new data with existing data (constantly updated by new data) by way of clever algorithms allows for fairly reliable translation machines if the data bases for individual languages is sufficiently large – if not, gobbledygook results, or shall we say Double Dutch so as not to implicate Trump in this equation. Corpus linguistics has brought us many new insights about the use of language as well as statistical analyses of syntactic structures but again, like the black hole metaphor, we do not learn anything about language itself. Language, like a black hole, swallows all data, reducing all to a singularity, and if we may speculate a bit, let us employ Chomsky’s (1995) singularity in the form of MERGE, the big bang of linguistics. However, if one insists on the absolute equivalence of language and thought, as I do, the question remains how MERGE relates to the biological construction of thought. If one assumes, in simplistic terms that some linguistic item X merges with another item Y and then further simplifies this idea by substituting X and Y with lexical items like ‘John died’, one can envisage that a thought was created: but how can ‘John’ and ‘die’ be created as atomic thoughts in the first place? If we support the supremacy of syntax, i.e. in the beginning was MERGE, we could propose (as does Fujita, 2016) that the lexicon evolved by virtue of syntax creating categories that trigger the creation and subsequent insertion of lexical items. This seems like an elegant explanation of an intractable problem, i.e. a sort of double big bang for thinking (MERGE) and what to think about (LEXICON). Whilst critics may argue that this is putting the cart before the horse, one can equally put the metaphor on its head – as it were – and argue that once it had been discovered that the horse has a pulling ability/capacity, an immediate/concurrent consequence was the question as what can be usefully be pulled along, like a cart. Hence the capacity for language and thought in terms of MERGE (or framed as the earlier Chomskyian Language Acquisition Device) goes in tandem with the development of the lexicon. Another metaphor employed in this thought experiment is that for building a house we first need a structural plan (SYNTAX) and accordingly we select the bricks and mortar (LEXICON) to actually build the house (SENTENCE). In this vein we can also revisit the other great issue in linguistics: the Saussurian distinction between language (langue) and language use (parole), similar to Chomsky’s distinction between competence and performance. Perhaps we can solve this dichotomy in line with Fujita’s (ibid.) proposal: syntax = competence and lexicon = performance. As such the question of what we are thinking about when we are thinking becomes more interesting to answer: while the lexicon is limited, the creation of sentences seems unlimited, largely due to the iterative qualities of MERGE. Whether or not such potential output is really infinite or just an astronomical number may be an interesting question for AI, as it is clearly banking on the latter. When language is compared to the chess game (as does Saussure and Wittgenstein) then AI is laughing all the way to the bank because, as with chess, huge data bases and deep learning based on statistical matching allow AI to win games against most grandmasters of chess – and imagine if and when this comes true for language. Luckily perhaps, language and thought may be infinite in scope and as such defeat AI in any creative field of endeavour. Of course, functional linguistics provides a platform for AI as long as we can guarantee that language arises from its purported function, i.e. communication. Moreover, if we can strictly limit this function for 99% of the human population, we have the perfect fascist state. Again the good news seems to be that, in theory at least, language and thought cannot be constrained by language and thought – exactly because language cannot explain itself, lest it bites its own tail.
Even so, this freedom of unlimited language of thought, apart from iteration, must have some computational processing constraints, such as the Chomskyian principles and parameters, and government and binding. Descriptive thoughts/sentences invariably revolve around some categorical action (VERB) and actor (NOUN), giving rise to complex rules of verb and noun phrases where the relationships between lexical categories in space and time are elaborated. If it is true that narrow neurological pathways become major cyber highways due to frequent use, then one can imagine that a main (perhaps egocentric) protagonist (NOUN) need not be spelled out all the time once we know who he/she/it in gendered English is. In other words we have invented PRONOUNS (also known as anaphors and cataphors in terms of reference). Invariably pronouns increase the processing speed and thus facilitate thoughts around a main subject or object. Obviously pronouns need to be bound to a particular noun, either within or outside a sentence (the latter is called context). This becomes even more interesting when we consider the binding rules for reflexive pronouns, which must be bound within a sentence. Different languages handle this in different ways. English mainly follows the local binding syntax rule while Mandarin follows both local and distant binding rules, depending on the pragmatics of the semantic content. Some linguists (Levinson, 1983, Huang, 2000) interpret this as proof of linguistic relativism, like consigning some languages to being ruled by syntax and others by pragmatics. This can be further extended to the Sapir-Whorf hypothesis that different language shape different world-views and cognition. Since I advocate the equation of language = cognition (as far as cognition refers to Descartes’ famous cogito, ergo sum) I have to say that linguistic relativism is a fabrication, simply because all natural languages operate on the same principle of MERGE and subsequent rule-based processing. That languages differ in surface features is of course well known. What should also be well known is that all languages can be learned and translated amongst each other, which makes linguistic relativism a non sequitur, for when taken to its logical conclusion, we end up with the Tower of Babel story, which in terms of linguistics makes no sense at all. At this stage I also usually trot out another extreme version, actually put forth by a modern linguist (Kramsch, 2009):
The more people speak English around the world the less people understand one another. So it’s this irony that we’re moving into an era where more and more people speak English and yet less and less do they understand one another because through English they are thinking, they speak English but they think French, or they speak English and they think Hindi. And so it becomes an invisible multilingualism behind the English that they speak and I think applied linguistics has a lot to contribute to that understanding of what it means to have a multilingual mentality, a multilingual competence.
I am fond of then saying that, therefore, what I am writing here in English is a weird translation from my German thinking (my native language), hence can be dismissed as Double Dutch (one cannot resist of also trotting out the silly joke of the German coast-guard intern responding to a mayday call “we are sinking” by saying in a heavy German accent “what are you thinking about?” – purported linguistic relativism makes for great comedic entertainment, as also put to great effect in Fawlty Towers what with Manuel’s ‘I come from Barcelona’ antics). On a more literary note – and whilst I do not want to compare myself to Joseph Conrad or Vladimir Nabokov - one may therefore dismiss Kramsch’ argument (and linguistic relativism generally) as ludicrous. Quite obviously the lexica of different languages are shaped by various environmental contingencies and cultural practices but language per seis not. In fact I am only reasserting what Lenneberg made abundantly clear in his seminal book, famously entitled Biological Foundations of Language(1967) - and whom Chomsky credits with the erstwhile foundation of biolinguistics.
What is also interesting, is Wittgensteins’s erstwhile idea that language ought to be described in ways that thought is often defined, namely that thinking is, or at least ought be, based on logic. As such we have to applaud Wittgenstein as more or less proposing the equation of language = thought but with the additional thought = logic. As we know, try as he might, Wittgenstein could then not square the circle of language = logic. This seems odd if we assume that language/thought arises from a biological brain that is totally constrained by the very logical laws of nature, or is it? So why is language NOT a purely logical construct, even though in pure science it can be, and sometimes is (cf. Einstein’s scientific treatises)? This in turn can be linked to the dual problems often cited by Chomsky, namely Plato’s problem and Orwell’s problem. How come that in the face of so much logic (as in the natural sciences) we know so little in terms of organising our lives around peace and prosperity for all – and not just for the 1%? And how come we know and create so much without any logical thinking? The latter is often couched in terms of other intelligences, like emotional or kinaesthetic intelligences, which can dispense with logic – artistic expression often seems like that. Wittgenstein’s escape clause – to the disappointment of Russell – was that language is a game (like chess perhaps) in which there are some initial rules (often based on fuzzy logic) on how to play the game but there is no telling on what happens next. Well, at least we got game theory, which is yet another metaphor some people live by, i.e. winner takes all. Russell on the other hand stuck to his axiomatic mathematics and logic, and conducted himself as a bit of a progressive political activist on the sidelines (one cannot resist to tell a much better joke at this stage: when Russell and his wife had their little Beacon Hill School, what with children running around naked, and the vicar knocking on the door, and the naked kids opening it, causing the vicar to shriek “Oh my God”, and the kids responding “but there is no God!”).
One other thing that Wittgenstein deserves credit for is his refusal to cite references, whilst acknowledging that he too stands on the shoulders of giants. Most teachers these days will be aware, painfully or otherwise, of the obsessive instructions for assessment tasks that students MUST use their OWN words, lest evil plagiarism condemns them to academic oblivion. So who ‘owns’ the words? Surely nobody or everybody! Barthes (find the reference yourself, if you insist) famously intoned ‘language is the author’. The assumption, that the words I write now are somehow my own, has only grown in economic strength since intellectual property- and copy-rights were invented by capitalist corporations. If I look up a word in a dictionary published by so and so, should I have to pay a royalty? Obviously it is the ultimate dream of the corporate sector that language becomes copyright: let’s take Microsoft as an example: when I buy my Microsoft Office software package (or more likely my on-line license) which comes bundled with dictionaries and grammar checking applications, I actually pay for the words I use. Since the competitors still have free access to dictionaries and grammar handbooks that are not subject to copyright anymore, the playing field is still level to a certain degree but watch for the monopoly move that will spell the end of free language use. While what I write on my MS-Word application is still my ‘own’ intellectual property, similar writing applications on various on-line social media platforms use all the data as fodder for further large data analysis, like functional grammar patterns. Whenever you use on-line Google Translate, the translated items are fed back to their databases, providing ever expanding grounds for statistical matching algorithms. In the meantime, please go ahead and copy my words, my text, my thoughts, and make them your very own without having to reference me at all. Actually, please do not make my thoughts your own lest you also pass them on to the next person who wants them. Remember that authorship is only a recent invention – before that, all language and thought was a free and shared commodity. Wittgenstein’s Tractatus Logico-Philosophicus is a fine academic text without a single reference – contrast that to the vast ‘publish or perish’ academe of today where every 10-page article has at least 15 pages of references to make the cut (my apologies for my references below).
Languages and their speakers are therefore in some ways analogous to societies and their respective individual members, inasmuch no individual has a claim to represent the whole of society. Even capitalists like Kennedy intoned that ‘do not ask what society can do for you but ask what you can do for society’, so the collective good ought to rule over the individual good. Marx and Engels pushed this idea to the next level by suggesting the ‘leap from quantity to quality’ whereby a universal socialist society is the embodiment of a just quality of life. The interesting question is how language intersects with these ideas, i.e. how does language facilitate such thoughts in the first place, and why is it that some individuals come up with these ideas but not others? Obviously the human species is not pre-programmed like ants or other animal societies, where the biological aims of the ‘colony of ants’ propel the individuals to only perform tasks that serve the colony. This is not a triumph of quality over quantity but it is, when we consider the human species. Here, paradoxically perhaps, it is the individual who leapt from quantity to quality by way of a biological transformation in the brain, via MERGE perhaps, and voilà, language and thought is created. Now, while all humans have this Language Acquisition Device (Chomsky’s famous LAD), we can readily account for the considerable differences on how individuals operate this uniquely human device. The rest of human biology is more or less the same across all individuals – although there are individuals who, to make a point in language and thought alone, argue for fundamental differences according to race, gender and what-have-you. These very differences in the output of language and thought propel us into actions, that range from the sane to the insane.
So, why has language and thought not elevated individuals to a societal quality that socialists advocate, given that language itself is the biological outcome of a leap from quantity (e.g. the neurons of the brain) to quality (e.g. the LAD)? Why are we not collectively operating like some sort of enlightened ant colony? Has this leap from individual quantity to quality actually been an evolutionary disaster that will lead to the extinction of the human species? Current evidence seems to support such a scenario rather than the ever-optimistic Marxist idea that a socialist paradise will be the inevitable outcome of human history. The idea that this individual language and thought capacity will eventually lead to yet another leap in human quality is of course a very beguiling one, precisely because the biological foundation of our collective language and thought capacity will override the many pathologies that were associated with the early stages – until now – of language and thought. By ‘pathology’ I mean incorrect thinking, thinking that violates the laws of nature that gave rise to thinking in the first place. This includes a wide range of pathologies, from thoughts that are translated into actions like genocide down to accusing students of plagiarism. How laws of nature can shoot themselves in the foot is of course the great paradox that we are discussing here. Not that we can claim, that all we know about the universe and its laws of nature, is somehow predicated on yet another law of nature that says that all of creation is of mutual benefit to all who live in it. Natural disasters – although increasingly aided by human folly - seem to be inexplicable in that sense. If ultimately there is no rhyme and reason and that the human species with its languages and thoughts is a random phenomenon dreamt up by evolution, then of course we can pronounce the language paradox as the highest form of this phenomenon.
However, it seems that language and thought has elevated us above the laws of nature inasmuch we can imagine all sorts of things that contravene them, like gods, ghosts and the proverbial science fiction. The latter is doubly interesting because language and thought have also propelled us to the heights of science and technology, from space exploration to smart phone technology. Why is it that we can leave science and technology behind, and imagine robotic humanoids like the homunculi of old? Why metaphysics, alchemy, astrology, religion and superstition? Did Engels not consider what this ‘quality’ really is that leap-frogged quantity? One paradox – that language can conceive of a paradox is a paradox in itself – is of course, as mentioned before, that language as a totally new and still unique phenomenon lacks the meta-language to describe itself, like all other systems that cannot describe themselves lest there is a meta-language. Therefore language may well have conferred the type of human freedom we always seem to celebrate in contrast to the biological determinism we seem to observe in all non-human domains. That we are still prisoners to many aspects of biological determinism is of course viewed as yet another paradox we would dearly like to resolve, like achieving immortality and almighty power. In the past this has been called the mind-body problem, having elevated the mind above the biology of the body, but as Chomsky keeps pointing out, there cannot be such a problem because the mind, whatever it is, must arise from the brain and as such cannot escape its biology. So, while one cannot argue with such devastating logical thinking, language as a biological device in the brains of humans remains as inaccessible as finding out what is inside a black hole: it is a self-defeating exercise. Sure, neuro-linguistics is making all sorts of advances like telling us where language operates in the brain by observing brain activity when engaged in language and thought. Finding out where something happens does not tell us how and what is happening. From the scientific point of view this is yet another paradox, especially if we subscribe to Feynman’s famous dictum ‘if I cannot create it I cannot understand it’, for Feynman, like everyone else, creates language and thoughts every other minute. Of course one might argue that a super-scientist like Feynman, especially in the form of one Noam Chomsky, does fully understand language and thought, Q.E.D., and as such is at liberty to create whatever he wants. There may well be some truth in it, even though Chomsky proclaims that even he is only at the level of a Galileo when it comes to understanding language and thought but at least there is a scientific principle operating here that says that eventually we – or at least a future Einstein of linguistics – will crack the code. In the meantime we can only hope that common sense prevails, putting to shame and oblivion all those who use language and thought outside the norms of civilized behaviour.
Some references:
Chomsky N. (1995). The Minimalist Program. Cambridge, MA: The MIT Press.
Fujita, K. (2016). On the parallel evolution of syntax and lexicon: A Merge-only view. Journal of Neurolinguistics43 · June 2016.
Kramsch (2009). http://pterodactilo.com/numero6/?p=541
Lenneberg, E. (1967). Biological Foundations of Language. New York: John Wiley & Sons.
