On our third epistemological walk, we tackle the Big One: Immanuel Kant argued that the mind has an active role in shaping our knowledge of the external world. We then look at the role of language in determining the limits of knowledge. Finally, we conclude that what counts as knowing depends very much on the context, and return to the useful notion of ‘fuzziness’.
In the event, it was a few days before we got our next decent walk. I had a couple of out of town meetings, and Mrs McG took charge of the dog care. She brought him to the vet to see about his hip, and other problems. It wasn’t the best news. The kind of news you don’t really want to think about. So, for a treat, I took him on the bus to Primrose Hill, one sunny morning. He always liked it there, and would sit and gaze out across the vacuous glass towers of the City, as well as the gentler shapes of the old churches. We lay together on a blanket on the slope near the top of the hill, out of the wind.
‘You remember where we got up to? It was a while ago…’
Kant, I think. You left it hanging over us, like a threat.
‘OK, let’s summarize. The rationalists believed that the human mind was the source of all real knowledge, and they distrusted and dismissed the evidence of the senses. The empiricists allowed that some knowledge, such as that of mathematics, may be arrived at by the mind working independently (although Locke denied even that), but for the most part they believed that we know what we know because our senses have transmitted information about the world to our minds, which were otherwise blank slates. The rationalists offered us sure knowledge of what is unbelievable, and the empiricists offered us only doubts about what we thought we knew.
What was needed was, of course, a way of combining the two views, giving due weight to each. The person who achieved this great act of synthesis was Immanuel Kant. Kant’s account of the way the human mind comes to know the world, and what the limits of that knowledge are, is one of the greatest achievements of philosophy. The work in which he published his theory – Critique of Pure Reason (1781) – also happens to be one of the most difficult books ever published, the profundity and complexity of the thought mirrored by the knots and contortions of the language. (Reading it in German doesn’t help – many German scholars have claimed it is easier to understand in English.)’
You’re really selling it…
‘One problem with Kant is the technical language he uses. Kant was trying to be as precise and accurate as possible, using words in highly specific and novel ways to avoid the possibility of misunderstanding. But in trying to avoid misunderstanding, he often entirely bypassed understanding. He was working within an established tradition of metaphysical thinking, taking its language from Aristotle, Descartes and Leibniz, to which he added his own terms, usually defined within the text, but often defined in ways that seem to take you further away from, rather than closer to, comprehension. All these things make reading Kant an extraordinary challenge for any non-professional philosopher. However, it is just about possible to present Kant’s key ideas in a way that simplifies but does not distort them, although it must leave out much of the richness, as well as the mind-boggling trickiness.
Kant claimed that he had been awakened from his “dogmatic slumbers” by reading Hume. Hume’s scepticism challenged him to find a way to precisely delineate what could be known. As well as Hume’s scepticism, Kant was stimulated by the role that Hume gave to the mind in making sense of the data supplied by the senses. But Kant also wanted to avoid the idealism of Leibniz and Berkeley: there was a real world out there. The question was, how do we grasp it? So Kant’s project was both to bring certainty to what could be known, and to properly theorize the role of human consciousness in the process of knowing.
I’ve tried to avoid most of Kant’s most difficult technical language, using terms that I hope are familiar, but we’ll have to begin with a few definitions.
As we’ve seen, before Kant, it had been assumed that there were two types of possible knowledge. There were those things that you could know from having experienced them – Kant terms this type of knowledge a posteriori (meaning “from the later”). And there were those things that you could know without experience, by reasoning, which he calls a priori (“from the earlier”).
Most a priori truths are analytic – the truth is already in the propositions, and the reasoning process simply brings it out. The syllogisms we looked at earlier are examples of this. Though a priori knowledge can help clarify what you know, it is not really “new”. So, once you define what a triangle is, all the other interesting things you can say about it – the square on the hypotenuse type stuff – is already locked up in there just waiting for you to unpick it. A posteriori truths come from experience. They are synthetic, in Kant’s term, meaning that they create new knowledge – they bring before the understanding things that were not there before.
In Critique of Pure Reason, Kant set himself the challenge of demonstrating that there can be such a thing as synthetic a priori knowledge.
Hey, you said no jargon!
‘I said I’d try to keep it to a minimum… But this is pretty straightforward, if you focus your little brain. Synthetic – meaning containing new stuff; a priori – meaning not learned from experience.’
And is there, er, any of that stuff?
‘Does he pull it off? I think he does, yes!
For Kant, the mind has two faculties involved in the production of knowledge. Sensibility is the mind’s receptivity, its ability to take in sensory information from the outside world. Understanding is the mind’s capacity to organize and manipulate the ideas that originate with the sensibility. Both of these sound similar to the way Hume envisaged knowledge to be generated. However, Kant’s intention was to go far beyond Hume’s relatively superficial account of the relationship between the mind and the world.
Kant’s first step is to see that the acts of perception and understanding are precisely that – acts. Sensibility and understanding are not matters of passive reception, but active processes. Together they turn the chaotic jumble of sensory impressions, which Kant calls the manifold, into the logical and comprehensible world of our experience.
Beginning with sensibility, Kant never disputes that the sensory data that enters the mind relates to something real – he argues that there is simply no meaning to the idea of an appearance without having something real that appears. So he has no truck with the idealism of Berkeley, who maintained that perceptions have no reality outside the mind of the perceiver. The colours and sounds and smells of the world are representative of something out there, just as they were for Locke. For both Locke and Hume, our minds work on the incoming data, combining and interpreting it in various ways. Kant argues, however, that this act of organizing by the mind happens not passively, once the images have been received, but actively. Our perceptions have a structure from the outset, and that structure is imposed upon them by our consciousness in the very act of perception.
So what is this structure? Kant strips away both the raw data, the flashes of colour and flurry of action, and the later, more complex work of arranging and organizing done by the understanding. What he finds left are two concepts that had previously always been placed “out there” in the world, but which Kant now moves back into human consciousness: space and time.’
Whoa! So space and time are in our heads?
‘Exactly. We don’t observe space or time. They are not empirical concepts derived from our experiences. These are the preconditions, Kant argues, of any kind of perceptual experience. Kant’s way of putting this is that space and time are a priori intuitions. To see an object is already to have the idea of space and time written into them. It is impossible to imagine something that isn’t already organized spatially and temporally. We cannot think of objects other than as occupying some space, and as having qualities such as separateness, or connectedness.
This is what Kant means by that idea of the synthetic a priori. Usually something before experience can only be analytically, or tautologically, true. But here we have something that comes before experience, and yet contains real knowledge.
To the person who argues that we do simply perceive space “out there”, Kant replies that even the idea that space exists outside us presumes the existence of the spatial concepts of inside and outside, so the argument is circular. Similarly, the idea that things exist either in a chronological sequence or simultaneously is a structural principle that helps us to make sense of the world, not a thing we actually find out there.’
I’m trying here, but I’ve only got a little doggy brain.
‘It’s not just you. It’s genuinely challenging to force the mind to think of space and time as being projected from consciousness rather than being passively perceived by them. It might help to think about them as being the rules of a game. If you watch a football match or two people playing chess, what would otherwise be utterly baffling and random movements are, if you know the rules, entirely understandable. And it’s not that we passively receive the images of a person kicking a ball or moving their queen, and then subsequently interpret these movements: the game itself is structured by our consciousness – we read the rules into the match.
And this active making sense of the world is very much in accord with modern psychology, which fully endorses the view that the brain works hard to give the chaotic muddle of sensory perception a form and meaning. Just one very basic example. When we look at something, the image of that thing hits the retina. Because of the structure of the eye, the image on the retina has been flipped upside down. One of the first jobs the brain has to do is to flick it the right way up again. But that whole flipping business implies an innate knowledge of spatial relations.
It’s impossible for us to think of a world without time or space. That moment before the Big Bang, when space and time come into existence? The human mind literally cannot conceive of it. We have the words “before the Big Bang”, which seem to mean something, but there is nothing, no image in the mind that corresponds to them. To think of any object is to place it in space. To imagine it moving involves not only an idea of space, but also of time, which passes as it moves.
So, even sensibility, the first stage of acquiring knowledge through the absorption of sensory perceptions, is structured by qualities of the mind. The next stage in the process of knowing now takes over: the understanding. Intuitions – Kant’s word for this raw data, organized in time and space – have to be further interpreted by the understanding for them to mean anything to us.
So, the sensibility begins the process of turning raw sense data into something we can know, by arranging it in space and time. However, the real heavy lifting is done by the understanding, during which the mind sorts, organizes, combines and judges. To do this the mind uses what Kant calls categories, different conceptual entities that the mind employs to make sense of the incoming sensory data.
The idea that things in the world “out there” could be analysed by means of categories was first used by Aristotle (well, it had to be, really, didn’t it…). The original Greek term, kategoria, meant an accusation or a charge that could be levelled against someone in a court of law. Aristotle used the term to mean something like a predicate – a thing you could say, or a question you could ask, about something.
Aristotle has ten categories, which he thinks exhaust all the possibilities for describing an object. There is substance, i.e. what kind of thing it is, a man or a dog, say, or a tree. Then quantity – how many of them are there? Then quality – this is a little vaguer but means what qualities it has, such as colour, texture, smell, etc. Then relation – the way things relate to each other, so, Monty, you’re smellier than me, and I’m taller than you. Then, place – we’re here on this blanket. Then time – it’s today! Then posture – I’m sitting here, and you’re lying on my chest. Then state – another slightly trickier one, by which Aristotle means that you’re in a condition of having had something done to you – the examples Aristotle gives are having your shoes or your armour on. So, Monty has on his little cute coat… Then action – what is the thing up to? We are thinking and talking. And last, passion, which is the passive version of acting – having things done to you, being kicked, or shouted at.’
That was a little boring.
‘Sorry. I get that, but it’s kind of important. We now know everything that Aristotle thought you could say about something! If I analysed you using the categories, there would be literally nothing left to say about you. You’ve been fully categorized!’
Yippee!
‘The key thing to remember is that these things, for Aristotle, are the properties of the subject – they are qualities that belong to the objective reality of whatever the thing is we’re talking about. But Kant flips this round 180 degrees. All the ways of categorizing a thing are moved from the world to the mind. In place of Aristotle’s ten, Kant has twelve categories, arranged into four subgroups, which he calls moments. These moments are Quantity, Quality, Relation and Modality. We don’t want to get too bogged down in Kant’s categories, but roughly the group under Quantity deals with how many of a thing there are; Quality is to do with a thing being present or not, or limited in some way; Relation is about how things in the world are related to each other, for example by cause and effect; and Modality is about how an object might exist, for example does it actually exist, does it possibly exist or does it necessarily exist. On perceiving anything in the world, the mind instantly imposes the categories on it in an act Kant calls synthesis, and this synthesis, this sudden grasping of a thing in relation to the categories, is what understanding, or knowing, truly means.
It’s all a bit, er, confusing. In fact, that was pretty well just noise. Can you put it into dog language?
‘I’ll try. So, for the moment that he calls Quantity, Kant has noticed that I could say, for example, Monty barks, or some dogs bark, or all dogs bark. These are all the possible ways to think about number or Quantity – one, some or all. So, we have the categories Unity, Plurality and Totality, under the moment of Quantity. In the moment of Quality, the possible categories are Reality (Monty’s here), Negation (Monty isn’t here) and Limitation (Monty’s here until we leave).’
I’m still not really… Those pills the vet gave me… they can make it hard to concentrate.
I stroked his nose.
‘When I look at you, my mind automatically does all that work, applying all the different categories – insofar as they are relevant – to you, and my understanding of the thing in the world called Monty is the sum of all those subconscious acts. Better?’
I think so. Sorta. Maybe. No, not really.
‘The thing to cling to is that idea that my mind is actively engaged in creating the thing that I understand as you. It does it by imposing certain forms on the crazy mad flurry of sense impressions I pick up. First it imposes time and space, and then it fixes you into the various categories. And the space and time and the categories are all in my head, rather than in the world.’
Hang on, does this relate back to the a priori synthetic thingamabob?
‘Yes! The categories and so on are a priori, and yet they are also true knowledge, reaching beyond themselves to conquer new cognitive land – in fact, they are the preconditions of me having any knowledge at all. Furthermore, these categories are a basic part of being rational, so any human being will possess them. This means that the world as grasped by the human mind will be a common one, one shared between us, and mutually comprehensible. We’re not locked into private worlds, like Leibniz’s monads.’
Well, that’s a relief. Can you tickle my… aaaaahhhhh yes, that’s the spot.
‘What we’ve left out, so far, is the world behind the appearances. I said Kant never denied that something is out there, initiating the chain of events that concludes in me seeing and understanding the thing called Monty. Locke had posited that there are primary and secondary qualities, and he thought that the primary qualities really belonged to the thing, and were, in a sense, what the thing is. Berkeley (and Leibniz) thought there is no thing out there, just the idea in your head—’
Don’t tell me, Kant takes some perfect middle path – what did Aristotle call it, the golden mean?
‘Not quite. Kant calls the thing in itself the noumenon. And it is among the most mysterious entities in all philosophy, this something, out there, about which we can know literally nothing. Other than that it exists. It has no shape, no weight, no texture. All of those qualities are generated by our mind’s active powers of creation. So, yes, there is a presence there, but Kant says that human knowledge and reason is ultimately limited, and it never reaches the noumenon. The categories only work on what is empirically available – and the noumenon, standing behind the empirical, is unreachable. The same, by the way, applies to God.’
Oh, another atheist?
‘No one can quite decide about Kant and God. He often talks of God, and of the necessity of believing in him, as a way of grounding morality, but he also says we can never know that God exists. He takes God out of the field of rational or scientific proof, and leaves him entirely in the world of faith.’
Do you basically think that that’s it, then, that Kant is right, and that most of what we perceive and understand in the world is imposed on it by the mind?
‘Broadly, yeah.’
So that’s it, then, we’ve done epistemology.
‘Not quite. Can you cope with a bit more?’
As long as you keep tickling me just there, yep, I can cope.
‘Epistemology didn’t stop with Kant, but I think that from then on, all epistemology was a matter of wrestling with Kant. And he usually wins. But there’s one more theory I’d like to quickly mention, and then I’ll tell you what I really think.
Pragmatism is one of the very few new philosophies invented since the Greeks. There’s nothing quite like it in the Ancient world. It’s associated with a handful of mainly American thinkers, active in the second half of the nineteenth century and first half of the twentieth, including Charles Sanders Peirce, William James and John Dewey. The pragmatists entirely broke away from the idea that there is a truth to be discovered, either empirically out there in the “real” world, or derived by pure rationality inside my head. For the pragmatists, knowledge wasn’t a matter of some correspondence between what I think and what the truth is. The truth is simply whatever works best in a given situation. It’s essentially an evolutionary hypothesis: we are animals attempting to survive and reproduce in a world that is full of challenges. Some beliefs that we hold will help us to prosper. Those that don’t help us will be weeded out, like sickly organisms. This is all that truth can ever mean.
Let’s say that I am a mammoth-hunter fifty thousand years ago. I know that at certain times of the year, a herd of mammoths wander down a particular valley. I believe that this is because the spirits of my forefathers drove the mammoths down the valley, on their way to celebrating a feast held in their honour at this time of year. Because I know when the ancestors will be arriving, I know to be ready with my spear. The effect of my belief is that I get a good feed of roast mammoth.
For the pragmatist, my belief in the ancestors is “true”. They would not say that the ancestors are really driving the mammoths – that isn’t what truth means. Truth is never about what’s really there – there is no “really” – it’s about what gets me through the day.
Or, I have to catch a bus. I believe the bus is always late on Tuesdays, so I delay my walk to the bus stop. I miss the bus. My belief about the bus was wrong, but only because I missed it. Does believing in God make you happy, and improve the quality of your life? If so, then God is “true”.
Pragmatism solved many of the traditional problems of metaphysics and epistemology. It solves them by saying that they are irrelevant. We no longer need to enquire into the ultimate nature of reality, or theorize about the mind’s categories or concepts. All that matters is: did I catch the bus? If I caught the bus, then my beliefs about it were true in the only sense that true can ever mean anything.
I’ve a bit of a soft spot for pragmatism, and my own view borrows a little from pragmatist thinking. However, the philosopher Bertrand Russell (1872–1970) gave pragmatism such a thorough beating that it has never recovered. Russell says that if a belief is “true” only if its effects are good, then we need to know (a) what is good, and (b) what are the effects of the belief. Without discovering these things, we can’t ever know what is true. But rather than making life simpler, this process makes it much more difficult. He gives the example of trying to find out if it’s true that Christopher Columbus discovered America in 1492. Rather than going to some authority, checking the internet or reading up on the history, the pragmatist has to somehow ascertain the effects of this belief. Will believing Columbus crossed the Atlantic in 1492 be more beneficial to me than believing he crossed it in 1493?
A pragmatist would point out that there are circumstances when giving the answer 1492 would give you a definite benefit, for example if you were sitting an exam. Then it would become “true”. But what if on passing the exam, I was in such a good mood that I walked into the road without properly checking for traffic, and was run down by a bus (possibly the same one I had missed earlier on)? Would it now be false that Columbus sailed the Atlantic in 1492 because “knowing” it had killed me?
If I believe that something is true only if it has good consequences, then my belief in this, i.e. pragmatism, must in turn have good consequences if it is to be true. And my belief that my belief in the good consequences of pragmatism must also have good consequences if it is to be true. And this belief in the belief of the good consequences of the belief in the good consequences of the belief in good consequences must also be—’
ENOUGH!
‘OK, you get it. The argument goes on forever. The problem is that truth has been defined solely in relation to my subjective experiences, with nothing to tether it to anything outside them. How do we know that Father Christmas exists? Because he makes more people happy than he makes sad.’
But you were saying there was still something in it…
‘Time for my take on truth and knowledge. You won’t be surprised that it’s got something to do with that concept of fuzziness. I think that much of the trouble we get into when talking about knowledge is because we’re confusing different types of discourse, which each have their own criteria for knowledge, some demanding a high degree of certainty and some not. So what we have to do is look closely at each knowledge-requiring situation, and decide what counts as true or not in those circumstances. In effect this is saying that “true” means nothing unless you specify the conditions.
Imagine that I’m waiting at the bus stop. The bus is due at 11.55. The man next to me in the queue is a little worried about having missed the bus. He sees that I’m wearing a watch, and asks me if I know what time it is. I could give several answers. I might say, “Hold on a sec, and I’ll check.” I actually know someone at CERN (the European Organization for Nuclear Research) and they have the most accurate clock in the world, and so I phone them up, reach the switchboard, get put on hold, and finally, as the man boards the bus, I say, “It’s 11.54 a.m. and 17.21345621 seconds!”’
Annoying.
‘Or I could say, “No.”’
Huh? Has your watch stopped?
‘No, it’s working. But no clock, not even the sort of atomic clock used in scientific labs, keeps exactly the right time. My wristwatch might be reasonably accurate, but it’ll be off by several seconds. The CERN clock might be out by a nanosecond or two. So, no, I don’t know exactly what time it is.’
Don’t expect me to protect you if he hits you with his umbrella.
‘Or when the man asks me, I say, “It’s about twelve noon.” But that’s also no use, because it’s so vague he might or might not have missed his bus. In this situation there’s a right answer, and it’s not the most accurate (“I don’t know”) and not the most specific (“11.54 a.m. and 17.21345621 seconds”), and certainly not the vaguest (“about twelve noon”), but the one with the right degree of accuracy (“It’s just coming up to five-to”).
This view, that truth is, in some ways, dependent on context, is a little akin to pragmatism, in that truth is tied to usefulness. But it also acknowledges that there is a real time, which isn’t real because it makes me happy, but because of certain objective facts about the world.
There are other situations in which the fuzziness is more palpable, where truth is genuinely hard to get at. Often this is because of the nature of language. Language probably evolved as a pragmatic tool for achieving certain things – enabling our ancestors to better co-ordinate hunts, helping to secure the social bonds between individuals; possibly assisting individuals to gain an upper hand over others in the competition for mates and food. But language soon started to be used for other purposes.
It clearly has some ability to convey realities about the world (“Tiger! Run for it!”), but there are times when it is a crude and inadequate tool. You might be asked how you feel about something. That involves turning strange biological and psychological processes into words, which, even if you are extraordinarily eloquent and determined to be truthful, might be impossible.’
Feelings are always a bit tricky. But usually people manage to get their meaning across, don’t they? Walkies, din-dins. It’s not rocket science.
‘It may not be rocket science, but language is a trickier, stranger, more elusive thing than you think. We tend to think of language in a simplistic way, words having a straightforward, unproblematic relationship to objects out there in the world. The word “bus” is just like a finger, pointing at that big red thing with four wheels, full of passengers. Language amounts, in this view, to a sort of picture of the world, and the more closely the picture resembles the world, like a hyper-real computer-generated portrait, the closer we get to truth.
Ludwig Wittgenstein (1889–1951), who we met way back, developed a more sophisticated version of this in the Tractatus Logico-Philosophicus (1921). The Tractatus is a little like Spinoza’s Ethics, in that it consists of numbered paragraphs, arranged into sections in a quasi-mathematical way. Each section begins with a bold assertion that, like Spinoza’s propositions and Euclid’s axioms, is assumed to be self-evident. Each proposition is then illustrated and expanded on in numbered sub-paragraphs. So, it begins:
1. The world is all that is the case.
1.1. The world is the totality of facts, not of things.
Reality consists of a series of facts, or states of affairs. The job of language is straightforwardly to picture these facts. Imagine a car accident. You have to describe what happened. You are given small scale models of the cars involved, and the roads and buildings.’
And a bus?
‘And a bus. Each model stands for a thing in the world. You accurately recreate the accident, using the models, carefully placing the toy car and bus in just the right place. The “truth” of the model depends on how closely it corresponded to the facts in the world, as they occurred.
Now substitute words and statements for the models. That’s how language works to picture the world. Rather than using the models, I might say The car that was driving south on Cannon Hill crashed into the bus that was driving north. Each phrase in the sentence corresponds to a fact in the world. There’s a suggestion that perhaps language had an even more literally pictorial origin. In the sentence The car is to the left of the bus, the word “car” is physically to the left of the word “bus”. And, of course the earliest written forms of language used pictograms to represent objects.
However, there are many things you can’t talk about if you see language as having a pictographic relationship with the “facts” in the world. Wittgenstein acknowledges this in the famous last words of the Tractatus – it’s one of the few lines that every philosopher knows off by heart: “Whereof one cannot speak, thereof one must be silent.”
Those things of which one cannot speak are not unimportant – to either Wittgenstein or us. He means religion, ethics, beauty, the meaning of existence. These are all, for Wittgenstein, things that language can never reach, as they are not matters of fact to which words and propositions can be allocated. Wittgenstein believed he had set a limit to what could be said. Within that limit, anything that could be said could be clearly understood. Beyond it was only meaningless waffle – by which Wittgenstein meant most talk about art, philosophy and religion.’
And is this true? Is this what language is, a list of facts about stuff in the world?
‘It’s an alluring idea, in some ways, that language is simply a matter of linking a proposition to a fact “out there”. Truth and lies become readily detectable. Ordinary language suddenly starts to function like mathematics, where there is never any ambiguity about the meaning of a symbol. What the Wittgenstein of the Tractatus offers is a cure for fuzziness. If language really was as he projected, then everything that could be said could be said with absolute, mathematical clarity. The price is that much of what we might want to say couldn’t be said at all.
Wittgenstein really thought that philosophy had come to an end with his book.’
I’m guessing he was wrong…
‘Yep, he was wrong. There are many problems with this view of language, even if we ignore the aching chasm of what it necessarily leaves out. The picture theory might appear to work for nouns, verbs, adverbs and adjectives – the red bus drove quickly makes sense as a picture of something that happened. But what about words like “and”, “or”, “but” and “why”? What are they pictures of?
Eventually Wittgenstein came to see that his early view of language was inadequate. It embodied a reductively atomic view of the world and of language: experience is divided up into tiny, independent units, which are mapped onto a single word or phrase. This oversimplifies both the world, which is made not of isolated units of being, but of complex interconnected networks, and language, where words are woven together, and meaning emerges from their interplay.
Wittgenstein’s later philosophy, put forward in Philosophical Investigations (published posthumously in 1953), looks in detail at some of the many ways we use language. The ability of language to generate meaning comes not from the word and the object in the world being skewered on the same kebab, but by subtle patterns of use, and complex rules, embedded in the texture of our lives.
However, although this later view of language is much richer, more fecund and, I think, truer, Wittgenstein proposed it before certain developments in linguistics changed completely how we view the relationship between the world, our thoughts and language.
Drat, just when I was getting the hang of things.
‘Traditionally, linguistics had studied how words change over time, and the meaning of a word was thought to be wrapped up in this historically evolving relationship between the word and the thing it represented. This etymological view is often fascinating in its own right. So, for example, in one of his lectures on metaphor, the Argentinian writer Jorge Luis Borges points out how the English word “threat” comes from the Anglo-Saxon ðreatt, meaning crowd. It’s quite easy to see how the modern meaning evolved from the earlier one, crowds being places of danger.
But the Swiss linguist Ferdinand de Saussure (1857–1913) took a very different approach to analysing how language works. He thought that rather than tracing how the meanings of words change over time (which is called diachronic linguistics), we have to analyse language as it functions now, as a signifying system (synchronic linguistics).
Knowing that threat comes from ðreatt is fascinating, and perhaps illuminating about Anglo-Saxon society, but it tells us nothing about how the word is used now. If I warn you about the threat of a snake in the grass, you need to know that there’s a danger, not a crowd.’
I could take a snake. It’s basically just a sausage with a tongue, from what I can see.
‘The basic unit of structuralist linguistics is the sign. A sign is made up of two components, the material part, which could be marks on paper or the sounds of spoken language, and the mental part, the idea to which the physical part refers. The material part is called the signifier, and the mental component is the signified. The two are united in the sign. The word DOG is a sign made up of the letters D-O-G, and the idea of a dog. You hit a red traffic light. The red is the signifier, the signified is STOP; the two together are the sign. You woof. The woof is the signifier, the signified is—’
Give me a sausage?
‘I give you a sausage. The sausage is the signifier, the signified is “I love you”.’
That’s sweet. And this sausage is where, exactly…?
‘Just an illustration, for now. Signs never work alone, but function together in a language, their meaning always defined by their relation to other signs. Red only means stop in relation to green (go) and amber (wait).
The relationship between the signifier and the signified is usually arbitrary – the English word “dog” and the French word chien both refer to the same idea; and as long as we all agreed, we could use any sound or written squiggle to stand for dog. This arbitrary link between the signifier and signified means that over time meanings have a tendency to slip and change in unpredictable ways, adding an element of instability to the system of meaning.
The study of signs is called semiology, and remains a fruitful and fascinating way of investigating human culture. I really like a way of looking at signs developed by the pragmatist Charles Sanders Peirce (1839–1914). Peirce described three types of sign, the Icon, the Index and the Sign proper, defined by the different relationships between the signifier and the signified. With an icon, the relationship is one of resemblance between the two – a photograph of you, Monty, is an icon. As would be a portrait of a face, or a drawing of a tree, or an onomatopoeic word like “bang” or “crash”. With an index, the connection is less clear-cut, but there is still a physical relationship: a blush would be an index of embarrassment, or a black cloud an index that a storm is coming. Third is the sign proper, in which the connection is purely conventional, as in most spoken and written language. It’s quite fun to wander around mentally allocating different types of sign to the right category. Soon you start to see everything in the world around you as signifying something: cars as an index of wealth, the subtle differences indicated by the icons on toilet doors—’
Weirdo.
‘It takes all sorts. For Wittgenstein, the structure of the world dictates the structure of language – states of affairs out there call for a word or phrase to describe them. In structuralist linguistics, this relationship is reversed, or at least greatly complicated. Language shapes how we see and understand the world. Just as the signifiers – the terms we use for ideas – are arbitrary, the way language chops the world up is also arbitrary, or, rather, something that varies subjectively with cultures, rather than being an objective fact.
There’s a playful example of this in an essay by Borges called “The Analytical Language of John Wilkins” (1952). Borges describes “a certain Chinese Encyclopaedia”, Celestial Emporium of Benevolent Knowledge, in which it is written that animals are divided into various fanciful categories, such as “those that belong to the Emperor”, “embalmed ones”, “those that tremble as if they were mad”, “those that have just broken a flower vase” and “those that from a long way off look like flies”.’
I like that. Tickles me.
‘The point is that rather than using our system of animal classification based on morphological resemblance and shared evolutionary history, here the criteria for categorization are to do with use, or whimsy.
The most famous example of this idea that language is responsible for how we see the world, not merely a reflection of it, was the view put forward by the anthropologist Franz Boas, and later popularized as part of the Sapir–Whorf hypothesis, that the Inuit had fifty words for snow. Because they had the fine-grained language, they were able to “see” different types of snow that would be indistinguishable to us. This view became a little discredited, but the latest research suggests that Boas may have actually underestimated the Inuit’s ability to discriminate. For example, in the Inuit dialect spoken in Canada’s Nunavik region there are at least fifty-three words for snow, including matsaaruti, for wet snow that can be used to ice a sleigh’s runners, and pukak, for the sort of powdery snow that looks like salt.
Even such apparently objective things as the colour spectrum are chopped up in different ways. The number of basic colour terms employed in different cultures varies from two to eleven, with those societies with few colour terms being unable to see the colours for which they have no names. Similarly, remote tribal groups in the Amazon, and other parts of the world, have never needed to develop number systems that go beyond four (or some other relatively small number). Confronted with assemblages of objects greater than four – including their own children – they resort to “many”. Without the language for enumerating, they simply cannot make their minds grasp bigger numbers.
Feminists have argued persuasively that male power and cultural control has ensured that maleness is inscribed in our language, forcing us to see women as inferior, reinforcing negative stereotypes of women and buttressing their subservient position. Racist language serves a similar function in othering and demeaning different racial groups.’
This is all moderately interesting, but I’ve sort of forgotten what it is we’re trying to do here.
‘Oh, sorry. I’m trying to show that in these ways language becomes not a simple tool for knowing, but a way in which what is known is controlled and shaped.
Structuralism was an important movement in European thought for much of the twentieth century, influencing everything from anthropology to film studies. The central idea was that all meaningful human behaviour can only be understood as taking part within a structure, which functions in some ways like a language, with a grammar and vocabulary, and in which each individual element only comes to mean something in relation to the others.
Although structuralism allows for some play and slippage of meanings, there is a fundamental stability, which Saussure and the other structuralists thought gave a scientific, objective value to their theories. This confidence was shattered by Jacques Derrida (1930–2004), who brilliantly undermined the scientific confidence of structuralism. Derrida argued that underlying structuralist linguistics – indeed underlying almost all theories that try to pin down an objective idea of truth, going right back to Plato – is a certain notion about the relationship between truth and language. The truth is a simple, unitary thing inside me. The purpose of language is to convey that simple unitary truth to another person, passing it on, like one person lighting a candle from another’s flame. The most reliable and direct method of transmitting the truth is by speech. Speech guarantees the authenticity of the truth, because the speaker is present. Truth and presence are closely allied. When I speak my truth to you, I can control it, and ensure that your understanding of it accords with my own. Once speech becomes written language, then it escapes that close control.
Plato bewails this move from speech to writing in the Phaedrus, and the same idea – that speech is truth and writing is lies – crops up again and again in the history of philosophy, right down to Saussure, who looked to speech and presence as the best way of preventing that dangerous slippage between the signifier and the signified.
Derrida regards this notion of communication as a myth. In its place he has a view of language in which truth, rather than being a flame inside each mind, kindled and passed on, becomes an elusive property of language itself. In its simplest form, this view of language emphasizes that words always point to other words. Definitions rely on words, which must in turn be defined in words, ensuring that meaning remains always “in-house”. Language is an endless chain, and we never reach the end, never reach that final truth inside someone’s head, and outside language.
He illustrates this with the Greek word pharmakon, used in the Phaedrus. The word means poison. Plato describes writing as being a pharmakon. However, pharmakon can also mean cure. In this sense it’s a little like our word “drug” – which can mean a medicine like penicillin, or a recreational substance like heroin. Whatever Plato might intend, as soon as language starts to work, the meanings are uncontrollable. It is impossible not to keep both meanings alive, even if Plato wanted to kill the pharmakon- as-cure definition.
Insofar as all attempts to understand the world are linguistic (and it’s hard to think how they would not be), we will never have a final answer to the What do we know? question. Truth will always squirm out of our grip, like a greased piglet—’
Wait, we’ve come all this way, only for you to tell me that the sceptics were right, and we can’t know anything?
‘No, not quite. I think that Derrida is right, in that a final answer will never be found to any question framed in normal language. That’s one of the ways in which normal language is different to the language of mathematics. But a “final answer” doesn’t mean no answer and, going back to our discussion about bus timetables, it doesn’t mean that we can’t have a satisfactory answer to the bus question, just because the meaning of “bus” and “time” will be slightly different for every passenger.’
So what do we know?
‘I think that there’s a sort of rough consensus. Kant’s assumption that there is a real noumenal world out there isn’t really challenged much. Apart from mystics and religious thinkers, there aren’t many idealists left who think the external world is a phantom. And it genuinely seems as though we’re getting closer to seeing or at least understanding it, that elusive noumenon. Quantum physics gives us an insight into the ultimate nature of reality that would have delighted Kant. But the noumenon’s strangeness and otherness will mean that it remains always just outside the grip of knowledge.
What remains, the phenomenal world of objects with colours and weight and smells and tastes, is shaped by the extraordinary creative human mind. We can “know it” as long as we define “know” appropriately, applying the right standard of specificity in each case, and keeping ourselves aware that the very nature of language means that, in everything other than mathematical knowledge, the truth is like the light in the fridge: you “know” it goes off when you shut the door, but knowing isn’t knowing…
So that’s epistemology. There are sceptics who think we can’t know anything. There are rationalists who think that we can know everything, but that everything is the universe inside my head. There are empiricists who think our senses give us solid evidence on which to form reliable theories about the world beyond our minds. And then there are those who follow Kant, in thinking that knowledge is an active product of the human mind, which reaches out to carve the raw material of the universe into understandable chunks.
But there is one particular area where knowing fuzzily seems quite inadequate. An area where precise knowledge is surely generated: knowledge that isn’t just analytic, like mathematics, but synthetic, making bold claims about the nature of reality that can be confirmed as objectively true.’
Great. But maybe another day. All I can think about is that sausage…
‘OK. Tomorrow we discuss the philosophy of science.’