Conceptual Schemes

At the end of the eighteenth century (c 1770) the German philosopher Immanuel Kant attempted to combine rationalism with empiricism in what is known as Kant's Synthesis. He was dissatisfied with the attempt by both rationalism and empiricism to make our knowledge of the world conform to how objects in the world really are, with one arguing that we have knowledge of them through sense experience, the other through reason.

The problem is that in both cases we don't have direct experience of them. Kant's solution was to ditch that attempt and to see if he could connect objects to our minds by looking at the problem the other way round - by explaining how objects conform to our understanding. In other words, it's not the objects that cause the way we experience them; it is our minds that "create" the object as we experience it. The concepts he called "the categories of experience" impose a particular conceptual scheme on the raw data of sense experience. Without that conceptual scheme, we would not have self-conscious experience at all. There would be no "I" to know it was experiencing the world. The categories include space, time, causation. They do not derive from experience; they are a priori in that they come before experience, but they are not mental processes that we control, as maths and inferential reasoning are. We can only experience objects in space and time, but we don't actually experience space and time themselves: they are part of the structure of our experience. Without experience, there could be no awareness of them.

So Kant is neither an empiricist nor a rationalist. He argues against both sides. Hume had argued that knowledge was either synthetic a posteriori (factual knowledge from experience) or analytic a priori (true by definition, not factual), but Kant argued that knowledge can be synthetic a priori - that is, knowledge that is known independently of experience but which is not analytic, not trivial. "Every event has a cause" is an example. It is factual but it depends on the a priori categories organising experience in such a way that we experience events chronologically and as causally connected.You couldn't even form the concept of an event without the categories: it depends on us perceiving things in space and time.

Today we would more naturally think of the categories as brain mechanisms (as Chomsky does in relation to the Language Acquisition Device) but Kant thought of them as a priori rules of the understanding which somehow unify our experience, imposing an order and coherence on it. Otherwise, our raw sense experience would be chaotic, confused, incoherent. We wouldn't even be able to say it was our experience because we wouldn't have a sense of an enduring self without the concepts of time, space and causation. This order comes from our minds, so it is in this sense that objects "conform to our knowledge". [Look at the quotation from Kant on page 25 of the textbook.]

Notice two important points. Firstly, Kant is accounting for our experience being as it is (phenomenon). He does not claim to explain how things in the world really are in themselves (noumenon). So you might well think that he still leaves us cut off from reality. Secondly, some philosophers (including Bertrand Russell) have argued that if the categories are a contingent fact about how we happen to filter raw sense data in our minds, then there is no logical necessity in our knowledge of the world. It could have been different. In which case, do we have the kind of certainty that is usually considered necessary for science, maths and inferential reasoning?

Kant certainly did consider that his conceptual scheme, his categories, were essential for human experience. Without them to filter raw sense data, we would not have experience because there would be no selves to experience. My sense of self depends on my ability to differentiate myself from other people, and that depends on my knowledge of objects in the expternal world, outside of my own experience, which in turn depends on the categories filtering sense experience and imposing a particular order or schema on it. The conceptual scheme makes the concept of self possible.

So even if they are not logically necessary, the categories are supposed to be essential, fundamental to what it is to be a human or a person. But we can think about conceptual schemes in a very different way. Perhaps we all need to experience the world through concepts that organise our perceptions in particular ways, but could the conceptual scheme vary from person to person, or at least from culture to culture? Perhaps the way people in one time and place organise experience (the way they see things) can be different from the way people in a different time and place do. If so, we might still consider that Kant's categories are essential, operating at a deeper layer in our minds, perhaps, but the particular way our culture conceptualises the world could be overlaid on top of them.

Think, for example, of different ways in which different cultures might categorise animals or plants. We see certain four legged animals as dogs, despite their different sizes, shapes, habits, etc. But do we have to? Could a different culture organise their concepts differently and therefore see things differently? Surely, the answer is yes. We interpret the world, we don't just see it, and our culture imposes an interpretation on us to a large extent, through language and concepts. So our world - our way of seeing how things are in the world - is not universal. It is specific to our culture, or to our culture's conceptual scheme.

This means that on top of the Kantian conceptual scheme that is considered to be essential to our human nature, to our human experience, there are culture-specific conceptual schemes that affect the way we perceive things. Both kinds affect our view of reality and of truth.

But the idea of variable conceptual schemes opens up the possibility of truth and knowledge being relative to particular cultures. Unless...we can decide that one conceptual scheme is better than others, that one presents the world to us as it is, and the others are false. In that case, we can dismiss relativism and say truth is not relative, it is absolute, fixed, and Culture X has it.

Problem is, which culture is Culture X, the one with The Truth? Is it, for example, the culture of the Sioux Indians, of Kalahari hunter-gatherers, of the Taliban or of fundamentalist Christians from the American mid-west? No, I don't think so either. I think my culture, my way of seeing the world is more right than theirs is. On the whole, I trust science and rationality. I am biased in favour of what I know and understand.

It was just this kind of problem that Foundationalism wanted to avoid. If we could trace all our knowledge and ideas back to absolutely secure basic beliefs, we could then build up truths from there with confidence, as Descartes and Locke both intended. But neither rationalism nor empiricism has convinced us that they can provide firm foundations. A priori reasons and sense experience lead to problematic theories. Kant's categories look promising, but they are contingent on human nature and don't give us logical necessity or certainty. In any case, the cultural concepts can be considered to be in addition to Kant's conceptual scheme.

The main alternative to Foundationalism is Coherence Theory. This says that all we can aim for is that our beliefs are coherent, that they fit together and help us to interpret the world in a particular way and to make predictions. They are useful rather than absolutely true. As long as conceptual schemes work, do the job, we cannot choose between them. Our scientific web of beliefs is no better and no worse than any other. Some schemes could be criticised, perhaps, for being contradictory or over-complicated, but most long-established cultures will have traditional beliefs that are coherent and with their own kind of elegance or beauty. (One twentieth century philosopher who has argued for this is W V Quine.)

Personally, I am not convinced. I do not believe that all conceptual schemes that are coherent are as good as each other, for I don't think they really can all do what, on the whole, our broadly scientific and rational conceptual scheme can do. Ours has more predictive and explanatory power. It has allowed us to bring the world to the brink of disaster by developing technology...but (with necessary adjustments) it may allow us to go beyond existing technology and save the world from disaster. It is very open to change; it revises itself as new evidence and techniques become available. It does not get quite so bogged down in traditional belief as some cultures do. It allows us to change not only the minor beliefs at the edge of the web of beliefs but even occasionally important beliefs towards the centre. Also, as the world shrinks due to globalisation and electronic media, it the scientifc culture of the West (USA and Europe) that is spreading. It is hard not to conclude that this is simply because it is more likely to bring progress through knowledge and truth. (Although, of course, many would disagree, including fundamentalists of all kinds - but they still use the technology that Western thinking invented.)

Of course, that is not a reason for disregarding other conceptual schemes or cultures. We can respect them without accepting that they have an equal claim to truth as ours. Also, of course, there may be some aspects of them from which we can learn, ways in which we can recognise that they are wiser or clearer.

So I want to avoid relativism about conceptual schemes. Some beliefs simply are more true than others. But you must decide for yourself, and when you do you will inevitably be using ideas and concepts (reasons, truth, knowledge, justification, fairness, freedom, etc) which are concepts from our culture, with which we organise our understanding or interpretation of the world. It is hard to escape the conceptual scheme - the web of beliefs, the language, the concepts - our culture has given us.

Criticisms of Empiricism

1. Empiricists talk about forming compound ideas by combining simple ideas, but what is a simple idea? If "golden mountain" is a compound idea, what makes "mountain" a simple idea - is it too not a combination of ideas? Is the concept "horse" simple or compound? It is not a clear distinction.

2. Do you agree that all ideas could come from sense experience - not just from experience but from sense experience? Take moods such as joy and melancholy - do we really acquire those through our senses? Also, we have concepts of things that we believe exist but no-one has ever experienced (eg atoms; and we have abstract ideas (such as justice and freedom). How do they connect with sense impressions?

On the other hand, perhaps we do experience justice and freedom in some way. Perhaps we can trace them back to basic human needs and experience. Perhaps we can relate them to evolutionary developments. What do you think? Also, the example of people who are born blind and therefore lack certain ideas may be relevant: they do not seem to form certain concepts or ideas. How convincing is this point?

3. Do you agree that Hume's empriricism about meaning is dodgy? Language doesn't have to relate to sense experience to be meaningful, but that's what the twentienth century Logical Positivists (such as A J Ayer) argued (as well as Hume). They were extreme empiricists. They argued that "metaphysical" words, such as "God" and "soul" were meaningless because they did not connect with things that had been drectly experienced and therfore couldn't be verified. But we do have perfectly meaningful converations about such things. They are trying to restrict what we can talk about, but if we couldn't speculate about such things our lives would be devalued. Do we experience beauty directly? Self? Justice?

Part of the problem is that Hume thought of ideas as images in the mind (copies of impressions), but this is a naive account of ideas, as Wittgenstein showed in "Philosophical Investigations". An account of language that is based on meanings being private images or ideas in the head is fundamentally incoherent: meanings have to relate to what is shared. (Remember the beetles in a matchbox example.)

4. Empiricism, like Descartes' version of rationalism, cuts us off from the physical world, from reality - because we can only know it indirectly, via sense impressions. This means that we can't be certain about anything except our own private sensations, which makes our beliefs subjective, whereas we normally like to think that our knowledge claims (ie good quality beliefs) are objectively true.

Also, it leads us into solipsism - ie scepticism about the existence of other minds - and towards the Matrix/Brain-in-a-Vat scenario: how do we know this isn't all a dream/gigantic illusion/creation of a crazy scientist, etc? So, empiricism leads us towards the trap of scepticism and solipsism - the same trap that Descartes could only escape by relying on God not to be a deceiver, and which Berkeley also required God to save him from. But what if you don't believe in God? Can empiricism avoid scepticism about the external world and soliosism about other minds without invoking God? Hume didn't think so, which is why he gave up the search for rationally justified beliefs and focused on explaining why we believe what we believe, accounting for it by reference to human dispositions.

5. Another form of scepticism concerns our knowledge of the past and of the future. Can empiricism give us good reasons for trusting our memories, the main source of knowledge of the past? Can it justify our beliefs about the future? All our inductive reasoning about the future depends on things continuing as they have done previously - but can empiricism justify our faith in things continuing as they have done up to know. Perhaps the laws of nature will suddenly change. Can experience rule that out?

6. We can question the empiricist account of the relation between concepts and experience. Concepts depend on our forming general categories, but how can we know how to categorize our sense impressions. Sense impressions just arrive - they are pre-conceptual, immediate, "the Given" - but without some kind of conceptualization or categorization they would be simply a "blooming buzzing confusion" (as William James said). This is why Kant and others argued that it couldn't be true that all ideas derived from experience because we need some ideas or concepts to make any sense of sense impressions in the first place. So it might well be that what empiricists such as Locke took to be immediate and incorrigible sense experience, it actually experience which is already filtered through concepts that make sense of it, organise it, conceptualise it. In which case, experience isn't the incorrigible "Given" that empiricists take it to be: it is already the result of intellectual (cognitive) processes.

7. Of course, rationalists criticise empiricism for prioritising sense experience over a priori knowledge. Without inferential reasoning, for example, we would not be able to generalisze from particular experiential cases. If it is reason that makes categorisation or conceptualisation possible, then it is reason that makes self-conscious experience possible. So reason comes first and rationalists are right. But is it?

In the next post I will write about Kant and Conceptual Schemes.

Empiricism

We agreed (I think) that rationalists are right to argue that we cannot achieve certainty about knowledge unless we can generalize from particular cases (experience, observations, experiments) and form general principles or laws of nature. I think we could also agree that in order to do that we need to use the kinds of intellectual tools that rationalists say we have a priori knowledge of: maths, inferential reasoning, etc. But do we agree that we have knowledge of such tools innately or by intuition? Probably not.

Empiricists certainly argue that we can have knowledge only as a result of experience, and that includes knowledge of maths and inferential reasoning. They trace the foundations of all knowledge and all ideas back to the firm foundations (as they see it) of sense experience - that is, the basic, raw perceptions that reach our senses from the outside world. Hume called them "impressions". More recent philosophers have referred to "sense data". Whatever we call them, the point is that we cannot have any knowledge of the world except via our senses.

John Locke argued (against Descartes) that we are not born with any innate ideas; we arrive with "tabula rasa" minds - clean, blank, empty. The knowledge and ideas we acquire all come as a result of sense experience. Hume agreed with this, although he also thought that as humans we are born with a certain nature, dispositions, for example, to associate one idea with another, or to form habits of expectation, so that when we see item A we tend to expect item B. This was how he explained our inductive reasoning, causal explanations and our knowledge of the external world. "Hume's Fork" is his argument that all ideas and meanings can be traced back to impressions (sense experience). Ideas are faint copies of impressions. The impressions (sights, smells, sounds, etc) have impressed themselves, printed themselves, on our minds, and our memory of them is what we call "ideas", but of course we can combine ideas of basic impressions to create more complicated concepts and imaginary objects (eg golden mountain). As for analytic truths - Hume was dismissive: they cannot give us genuine knowledge of the world because they only recycle ideas we already have, being merely true by definition. All genuine factual knowledge (synthetic truths) derives from impressions, sense experience.

One major difference between Lock and Hume is that Locke believed our beliefs about the world were rationally justified, whereas Hume was more sceptical and thought they were only justified in so far as our natures (imaginations, habits, emotions) allow us to interpret the world in particular ways. He didn't think we could know what the external world was really like because we only ever see impressions, not the objects that (we assume) caused the impressions. Locke and Descartes were both Indirect Realists, believing that we could rationally infer the existence of the objects that cause our perceptions of them. The ordinary "common-sense" view is that we just see objects as they are. This is known as "Direct Realism", and Hume thought both the Direct and Indirect versions of Realism were equally shaky. So he was a kind of sceptic or anti-realist. He had been persuaded by Bishop Berkeley's arguments against Realism but he didn't follow him all the way to Idealism ("to be is to be perceived" - ie only perceptions actually exist) - partly because Hume couldn't believe in Berkeley's God, who ensures existence is continuous by perceiving everything! So Locke, Berkeley and Hume are all empiricists, because they all believe that we acquire knowledge from sense experience (a posteriori) not before it (a priori), but they represent different kinds of empiricism.

Next we will look at some criticisms of empiricism.

Rationalism

Rationalism claims that our knowledge of the world rests on firm foundations derived from ideas that we know a priori, before experience. We have ideas or knowledge that are not merely kinds of "knowing how" (such as knowing how to suckle) but can be used in propositional knowledge - factual knowledge or "knowing that...". So, not mere instincts. Innate knowledge or innate ideas are ideas or facts we are born with, even if they require experience to realise their potential, to make us aware of them. Claims are made for the idea of God, of maths and inferential reasoning, for morality and beauty. A famous example of a rationalist's claim about innate ideas is Descartes' Trademark Argument for the existence of God: we find in ourselves the idea of a perfect being, only a perfect being could have placed that idea in our minds from birth, therefore the perfect being must exist.

Intuitions are ideas or facts that are simply self-evident - "clear and distinct ideas", Descartes called them. We grasp them by pure thought or introspection. Descartes thought that we grasp the real nature of objects in this way: not their secondary, perceptual features but their underlying measurable features (extension - mass, etc.) G E Moore, a twentienth century moral philosopher, thought the rules of morality were grasped intuitively, not by experience - they are simply self-evident.

Another kind of a priori knowledge is analytic knowledge. That is knowledge of the identity or sameness of two different ways of putting things. Analytic propositions are true by definition. So, for example, "Torture is cruel" might be considered analytic (rather than synthetic) because it is true by definition: it is built into the concept of torture that it is cruel. A lot of philosophy is about analysing concepts in this way.

In general, rationalists argue that a priori truths are necessarily true. They are not contingent on how the world happens to be: they simply have to be true, the world could not be otherwise. Empiricists claim the opposite: the only necessarily true propositions are analytic ones and they don't give us any new information about the world; they simply repeat the same thing in different terms.

Rationalists claim that it is by pure rational thought that we are able to generalize from particular instances (experiences, observations, experiments) to reach necessarily true laws of nature. We do this by maths and by ampliative inductive inference.

Criticisms of rationalism. Research these for yourself, but make sure you consider the innate, intuition and analytic points separately. Are there really any ideas or facts that are innate in every human being? God? Morality? Maths? Reasoning? Do all hmans really find the rules of morality self-evident, or grasp the underlying nature of physical objects? Is Hume right when he argues that analytic truths are not facts about the world; they are only definitions - analysing the necessary conditions for using a particular term or concept?

On the other hand, is it not true that we do need to generalise from particular instances? Repeating experiences or observations does not gives us knowledge of general principles or laws of nature> For that we need maths, logic, rules of inference (such as non-contradiction). So you might agree with rationalists to some extent - at least, that experience alone cannot give us useful, practical or scientific knowledge.

Reason and Experience: Introduction

Where does our knowledge come from? How do we acquire it? Epistemologists argue about what knowledge is – ie what its necessary conditions are. Is it, for example justified true belief? What properties must a mere belief have before we can count it as knowledge?

Epistemologists also argue about whether knowledge is a unitary concept: is knowing that Obama is the US president the same kind of thing as knowing how to swim or knowing your neighbour (knowledge by acquaintance)? But the arguments we are concerned with are about the sources of knowledge.

No doubt a great deal of what you know has come from other people (friends, parents, teachers) telling you things – ie testimony. Some people think they know things as a result of revelation – it has been revealed to them by a vision, or voices or a miracle. We might argue that our own memories can be a source of knowledge for us. But the main arguments traditionally have revolved around the claims that rationalists have made for reason as the primary source of knowledge, and the claims made by empiricists for experience as the only true source of knowledge.

On the whole, empiricists argue that none of our ideas come from reason alone, whereas rationalists argue that some do. Rationalists concede that much of our knowledge depends on sensory experience, but insist that the true foundations rest on what we know by reason alone.

Rationalism argues that we can know things by reason alone: thinking can give us truths without help from our senses. The truths of maths and logic are often used as examples. A solitary thinker sitting in front of his fire, with a candle burning by his side – Descartes’ meditator – can discover certain truths in his own mind. His senses may deceive him, he may be the victim of the tricks of an evil spirit, but certain clear and distinct ideas or intuitions can give him certain knowledge. In this way Descartes’ meditator achieved certain knowledge of his own existence and of the existence of God (the cogito argument and the ontological argument). He also discovered in himself innate knowledge of God(the trademark argument). He was then able to use such certainties as foundations on which to build a secure body of knowledge – the existence of the external world, science, etc.

So the twin claims of rationalism are that we can have knowledge without help from the senses (a priori knowledge) and we have a certain stock of innate knowledge or innate ideas (eg God, morality, rules of reasoning). These are separate claims and it is important not to confuse them. We could reject one without necessarily having to reject the other.

Two separate claims can be differentiated in the a priori claim: firstly, that substantive knowledge (facts) can be achieved by pure thought or intuition, and, secondly, that analytic truths can be known without sensory input – they are true by definition. Again, we could reject one without having to reject the other.

Both traditions, rationalism and empiricism, have generally (until the twentieth century) been forms of foundationalism. This is the idea that we justify our beliefs or claims to knowledge by connecting them to chains of reasoning going all the way back to basic foundational beliefs, the ones you can’t go beyond. For empiricists the basic foundation of all knowledge is sense experience – the raw sense data that comes into our mind from the outside world via our five senses. This is the Given; it is immediate, incorrigible, present. Rationalists argue that we can’t trust our senses: they can deceive us, so they cannot provide secure foundations for our judgements; absolutely certain knowledge can rest only on what we discover in our minds by pure thought.

One of the problems for empiricists is to explain how propositional knowledge (knowledge that p) can be built on perceptual experience, which is knowledge by acquaintance. But then rationalists have to explain how it is that we can discover propositional knowledge already in our minds. How did it get there? The problem for both traditions is to explain how something that is subjective and private (sense experience or thoughts) can be the foundation of something that is essentially public and shareable – knowledge.

At the end of the eighteenth century Immanuel Kant tried to reconcile rationalism and empiricism through his idea of an innate conceptual scheme, and in the twentieth century philosophers have questioned foundationalism and the possibility of founding knowledge on private experiences. Various versions of coherentism argue that beliefs are only ever justified by other beliefs: what you belief is justified and rational when it fits into your existing web of beliefs. There are no foundations. This opens the way to relativism about truth and knowledge.

The starting point for all modern philosophy is Descartes’ cogito ergo sum argument in Meditations. It was published in the 1640s at the dawn of the new age of experimental science, which combined observation and experiment with mathematical reasoning – in other words, it combined the empirical with the rational. Particular empirical observations needed to be generalised into theorems, principles and laws of nature. Descartes and others considered that you need the empirical data but you also need to interpret the data in the light of the general, universalisable truths that rational methodologies have discovered. Observation gives us data but reason gives us understanding. So the philosophical debates about reason and experience have real significance for science, for how we interpret the world to ourselves. Empiricists and rationalists have to achieve generalised truths or laws of nature in different ways.

Another key moment was when John Locke, towards the end of the senventeenth century, rejected innate ideas and argued that we are born without any ideas or knowledge- the tabula rasa argument. So, what are the key arguments for and against innate ideas?