Compare and contrast Hume and Kant’s accounts of causation.

Compare and contrast Hume and Kant’s accounts of causation.

      Both David Hume and Immanuel Kant were very influential philosophers who answered many topic areas and questions within philosophy. More pointedly, Kant often aimed to answer or amend Hume's answers to certain questions; causation being no exception. Hume's original standing on causation in 'A Treatise of Human Nature' and 'An Enquiry Concerning Human Understanding.' are central to his entire philosophies which makes the area a highly discussed philosophy. Kant, following in Hume's footsteps, wasn't satisfied with Hume's account of causation and so his own answer became highly discussed and central to his own philosophy too. As a result, the differences between Hume's account on causation and Kant's are highly important areas for distinction. In this essay, I shall cover Hume's account first and then move onto Kant to point out the comparisons and differences between them.

      David Hume, as an empiricist, believed that all knowledge comes from our experiences. That is, all information is received through our senses – our knowledge of the world can only have been seen, heard, felt, etc. With this in mind, Hume states that much of our worldly knowledge is often habitual with no real grounding. In his example of causation he states that when event A directly precedes event B often enough we become 'psychologically certain' that A is the cause of B. An example of this is when holding an object over a fire, it becomes hotter. In our experiences, because every time this has happened to our knowledge the object has become hotter, we are psychologically certain the holding something over a fire in the future will cause that object to get hotter. He goes on to say, however, that this kind of thinking is incorrect and it is only habitual to assume that future events will always be the same as past events.

      To fully understand why Hume makes this claim it is important to distinguish between what he calls 'ideas' and 'impressions'. Impressions are reflections in the mind – what we know that comes from our senses. Impressions are things we have experienced or remembered. Ideas, on the other hand, are purely thoughts that come from our minds alone; often by combining or applying out knowledge from out impressions. For example we can have an idea about a human with butterfly wings even though we have never seen one – because we can combine the two thoughts of butterfly's wings and people. That Hume claims all 'ideas' are merely copies of 'impressions' is called the copy principle. This means that all ideas can be traced back to being formed from impressions or the impressions have at some point been used to form the idea. This also explains Hume's reasoning for being an empiricist – due to all our knowledge that isn't taken from experiences / senses is actually derived from our impressions and therefore our senses.

      This means that, in terms of cause and effect, the ideas we have about something are only there from our actual knowledge of the object. So if, for example, we see something entirely new having never seen it do anything – any ideas of 'effects' it might 'cause' becomes arbitrary as we have no real idea about the object. An argument against this may be to say that we've seen enough things in the world to see a new object and make educated ideas about the properties of that object. If it looks rubbery, for example we might assume it will bounce when dropped on a hard surface. This would be because our experience has told us so, with the bouncing off the hard surface being event B and event A being the rubbery object being dropped from a distance. In our experience, B has generally always followed A – this causes the conjunction of the two events where we induce that B always will follow A in the future. This, however is where Hume's 'problem of induction' comes in.

      An inductive argument in causation is a logical / reasoned argument or conclusion one comes to after observing specific or particular instances and using that to make broader general claims. For example every time I stub my toe, I feel pain; therefore all stubbed toes cause pain. Or all life forms we are aware of are liquid water based; therefore all lifeforms are liquid water based. In causation it is the prediction that B will follow A if observed consistently enough. In 'An Enquiry Concerning Human Understanding' Section IV, Part I, Hume states that where 'relations of ideas' like maths and geometry retain their certainty of evidence (despite not existing in nature,) whereas a 'matter of fact' causes issues: “Matters of fact, which are the second objects of human reason, are not ascertained in the same manner; nor is our evidence of their truth, however great, of a like nature with the foregoing.” This suggestion is that 'matters of fact' that are ideas in our world about the laws of nature, etc. are never really factual as the assumption of things happening in the future never holds a contradiction when imagining the opposite effect. It is possible for me to imagine stubbing my toe and feeling pleasure instead. Or it is possible to imagine a rubber ball hitting the ground like a dead weight with no bounce. This shows that the adverse to our assumptions are not contradictions – as a result need an alternative way to prove their falsities if we are to truly accept them.

      Hume states that all 'matters of fact' are actually a cause and effect relation in practice. This is because if you were to try to give reasoning for one thing that is not immediately obvious, you give reference to another fact as reasoning. Hume uses the example that if you ask someone whether a friend is in the country he would give the reason based on having received a letter or knowledge of a promise, etc.. This, however, always requires experience as it seems implausible to, without any previous knowledge, correctly assume an effect purely from the cause. This is because with every cause, there could be a great number of effects that could happen and only through experiences can we make the correct assumptions. As a result, matters of fact are based upon cause and effect and cause and effect is based upon experience. This is because our knowledge uses different facts and reasons to come to certain conclusions, and those facts and reasons come from our observations and experiences. So it seems, from this that we use an inductive argument as when we observe some things we make assumptions about unobserved things. Then, because these things are unobserved assumptions in which alternative possibilities are imaginable – there is no contradiction and therefore is a false argument. Induction, then, seems not to bring about necessary truths.

      The principle of the uniformity of nature is an inductive suggestion of causation that the future will be like the past. This is a claim that all future events will be like all past events, the main argument for this being that all past events have been like future events in the past, and will continue to do so. It can be clarified by saying that all past 'A's have lead to 'B's, so the next 'A' will lead to 'B' and all future 'A's will lead to 'B's. On Hume's account, however, there are two major issues with this view as an explanation of causation. Firstly, as directly stated above, this seems to be an inductive argument, stating the generalisation that properties will remain the same because they always have done. It is possible to imagine, however, that something will change in the future like has never been observed and therefore make the principle of the uniformity of nature false. An example of this would be that it could be a property of the ocean's water (that is usually blue in large quantities) will one day become green. However because we have always observed it to be blue, the assumption becomes that the ocean is blue and will always be blue. The second problem with the principle of the uniformity of nature is that it creates a logical loop. As it doesn't have a set reason or explanation of how it reaches point A with conclusion B. The principle claims that A) in the past, all past events will be like future events. And B) So in the future, all past events will be like future events. This creates a logical loop as it is using it's own argument as it's own rule to assume future events. The only way you can assume that all As will lead to Bs in the future because all As have lead to Bs in the past is by assuming that all As have lead to Bs in the past and therefore will continue to do so. This, as an account of causation is false on Hume's reasoning and cannot be considered a true causal rule.

      Seeing as Hume rules out the possibility of predicting future events due to lack of a reasonable rule to do so, we are left with a causal problem of cause and effect. Cause and effect relies on the rule of temporal correlation that As will lead to Bs and so on, and as humans we believe in causation as if we can observe it. However, it is not a contradiction to take on the account that things do not cause other things. This is because it is possible to imagine something moving or acting without a cause – therefore it is logically possible for an effect to happen without a cause. For As to follow Bs, then, it is not unreasonable to suggest that it is merely correlation rather than causation. That Bs will happen correlationally when As happen without any direct link from A to B. Simply put, there are is no real way of proof that causation happens as fact and it is possible that all effects are arbitrary to causes.

      This puts us in a position of doubt, when sciences use universal generalisations for all their laws, as they attempt to look into the future, which there is no proof that this is correct. So we are left with the simple question of, under Hume's account of causation – why do we as humans follow causation, believe in causation and put our faith in causation. Hume claims that this is because we are creatures of habit, and because all past As have lead to future Bs, we habitually assume they always will. This is Hume's main account that causation is something that is only perceived and has no link to the real world outside of our assumptions that causation exists.

      Immanuel Kant agrees with David Hume on the account that causation happens in the mind from As to Bs, whether or not there is a causal link. He disagrees, however, that causation is not a thing that exists because the reason for his answer to Hume is because he believed that Hume's account did not leave us with any real answers – more that we do not know anything, or have any kind of foresight. With this in mind Kant aimed to save science and knowledge from Hume's ideas. Both Kant and Hume agree on two major ideas about causation; firstly, that we never really have a sensation, feeling or observation of causation – we simply see events A and events B. The actual thing that connects A and B is hidden to us, and yet we seem to assume that there is a definite link between the two situations. For example if I drop and object and that object hits the floor, there's no observation of the link between letting go of the object and the object hitting the floor. We see the object falling from being let go until it hits the floor, but that is merely part of the chain of events, not an observed link; we do assume, though, that these two situations are linked in some way. This is because Hume believes we are habitual creatures who can observe one event and then another, but never the link – which is why it is reasonable to suggest that it is not the case that there is a link. Kant believes the same criteria happens because whatever happens must be something that happens outside of our sensory experiences in the 'noumenal' world, which I will come back to.

      The second major agreement between Hume and Kant is that, whether or not causation exists, in our mind we link both events A and B as if causation does exist. This is to suggest that we do assume cause and effect that we observe in an activity if noticed more than once. So if we press a button and a buzzer goes off, then we press the button and the buzzer goes off a second time, we might conclude that the button creates a buzzer sound – when the buzzer sound is actually an unlinked sound that happened in correlation as the button was pressed. Both Hume and Kant believe that people can be wrong about what they believe causation to be, and that this is a habit that people have. Kant, however goes on to give reason why we might have this habit in the first place.

      Kant creates two major distinctions in causation. The first distinction he makes is between the noumenal world and the phenomenal world. Kant, like Hume, was an empiricist and therefore believed that all of our knowledge and information comes from out senses. Our sensory information is key to everything we understand about life; all of the things in this that we have senses of, that we can hear, see, touch, etc. are part of the phenomenal world. The phenomenal world is everything we interact with and understand – this is the world we live out life inside. The noumenal world, however, is an extra-ordinary world or plane of existence that we cannot possibly know about, and because we learn from our senses alone, it becomes meaningless to try to figure out what the noumenal world is. Any point that is made by anyone, in order to be meaningful, must provide reason or experience within it – if it does not, then it is meaningless, according to Kant. The noumenal world cannot be understood, because of this, although it is important to make the distinction between this and the phenomenal world. The main difference being that that noumenal only exists outside of our experiences and phenomenal exists for us to experience. The implications of this distinction allow for a lot of leeway as we have no real knowledge of what can happen in this other world – necessarily anything metaphysical could be a result of the noumenal world. Within the context of causation it could be the case that cause and effect have an actual link in the phenomenal world – this would explain the human trait that we observe things as causation, even if we get them wrong, it gives reason for why we innately understand causation.

      The second distinction Kant makes is a distinction between analytic and synthetic truths. This is very much the core of Kant's argument as his ideas on what type of statements are possible is what could possibly change Hume's externalism views to make science a meaningful thing. An analytic truth is something that is true by definition, so that the subject contains the information. An example of this is: “All cows are female”, where the fact that you've said it is a cow means it is female by necessity. Analytic truths need no external evidence or experience that they are true because the truth of the statement is contained within the statement it's self. A synthetic truth is the opposite to this, so that the truth of the statement depends upon the world and our experiences of it. An example of a synthetic statement would be “all cows are large”, because 'large' is not part of the definition of cow, so we need some knowledge of the world to find the truth in this statement. This distinction is very similar to the two types of knowledge that is stated by Hume, with 'impressions' and 'ideas', where impressions are similar to synthetic truths and ideas are similar to analytic truths. This is to suggest that Hume could claim that all analytic truths are based on our synthetic truths about the world, in a similar link to the way ideas are based on impressions. Kant wants to deny this, however, claiming that there is more to the distinction, adding a priori and a posteriori truths. A priori truths are truths that can be known previous to our experience of them, these are logically necessary as they can be validated by experience, but they must be true. An example of this would be “all adults are not children.”, or again “all cows are female”, as both of these statements could be tested, but are already true by necessity. A posteriori truths are those that can only be know after our experience of them, i.e. “All cows are large” or “chairs exist”, as we would have to experience cows to know if they were large; also, existence is not contained within the meaning of a chair, we need to experience a chair before we know it exists.

      Kant very much differs from Hume in this idea, as to most empiricists, like Hume, a priori statements are synonymous with analytic statements and synthetic synonymous with a posteriori. This means that analytic a priori statements such as all cows being female are the core of what Hume suggests with 'ideas' – things that are true by necessity that you know before you experience them. In the sense of causation a synthetic a posteriori possibility like “all Bs will follow all As” seems like it would be an inductive argument and therefore make no conclusion. A sentence like “all stones will fall downwards” is synthetic a posteriori, because falling downwards is not contained within the concept of a stone, and in order to verify if the stone will fall down, we must have experienced the stone falling down. To argue that this will always happen in the future would be an inductive argument. So as long as the empiricist view holds a posteriori and synthetic synonymous with each other and analytic and a priori the same, there can be no a priori truths about the future.
     Kant disagrees with this sentiment, suggesting that there can actually be such a thing as a synthetic a priori truth – that would be a truth about the world that we know by necessity previous to experience, but that we know because of experience its self. Kant claims that both time and space are such truths. He claims that because we have vision and are able to understand things such as distance or area, then we must know that space exists. While space is not specifically contained within the definition of vision, it must exist in order for there to be vision because seeing things means there must be space as an a priori truth. Time is the other thing that Kant argues must exist that we know through experience, and that is due to any experience in itself – in order for there to be any experiences there must be a passing of time. This is because in order for there to be an experience that has happened there must be a previous to the experience, a post of the experience and a time when the experience was happening – through experiencing we know time must exist, despite it not being contained within the definition of 'experience'.

      The implication of being able to use synthetic a priori truths saves science from Hume. This is because it suggests that there are certain truths that can be necessary because of experience, creating the possibility of things we didn't think there were any evidence for such as a link of causation. Kant argues that many things that we think are analytic a priori are actually synthetic a priori, meaning that there is evidence and reason for such occurrences and allows science to claim certain truths. Kant would argue that the fact that we have an idea about causation that naturally occurs to humans, in the same way that space and time naturally occur to humans, is because it must actually be a thing that exists. The whole idea that it is in our mind is the explanation for it's existence. This counters Hume's argument that ideas are just based on impressions because things like time and space are knowable ideas that we know because of experience, but we do not experience them directly.

      Though this argument for causation seems to come to a conclusion and synthetic a priori truths allow for the possibility of causation, it does not actually make a different argument to Hume's. This is because both Hume and Kant are both suggesting that causality is created by human knowledge of it. When people get used to causality it becomes a habit, and when we have this habit we create what we know as causation. Hume would suggest from this that it still does not necessarily give any proof that the causal link between cause and effect actually exists – it merely points out that it exists because we make it exist. The possibility of a synthetic a priori truth seems to validate certain ideas that can exist – but it does not make them sound. This is because Kant seems to put a large amount of trust into modas ponans type arguments. This is where the premises of the argument lead to a valid conclusion, but that doesn't make the conclusion sound. If Kant places trust in this type of argument it only means he has an idea that works, but that idea is not necessarily applicable to the real world. An example of this kind of argument for Kant is that he seems to think that 'if experience exists, then time exists', and 'experience exists' and 'therefore time exists.' While the two premises do lead to the conclusion, it doesn't necessarily mean that the premises are correct: in this case it could be false that experience exists (such as in Cartesian skepticism) or more likely that time is not a priori of experience.

      The major issue between Kant and Hume's accounts of causation is that Kant, in a way, falls directly into Hume's explanation, almost setting an example of the kind of habitual thinking he is talking about. Kant's ideas tend to be taking specific ideas such as the instances of causation or our certain experiences and giving meaning to them by generalising them; claiming that time and space exist which seems to be a kind of inductive argument. To suggest that time is something that exists because we have experiences is an inductive argument because we do not know that our experiences will not change in the future and our senses exclude space. So this argument to suggest that a synthetic a priori statement exists because of an inductive argument is exactly what Hume is warning again. Hume suggests that inductive arguments are what humans fall prey to because we get used to them, and Kant seems to be defending this way of thinking using reason which is flawed.

      Overall, it seems that David Hume set the basis for the ideas and reasons behind why we think certain things are true, when they are not necessarily true. He also shows why causation is merely created in people's minds when it is a product of faulty thinking and force of habit. Kant, looking to protect the idea that people are more than this, and that there is worth behind our beliefs and there is a point to reason and science, ends up agreeing to Hume on the basis that causation is created in people's minds – but that is because it is a presupposition of the nature of experience. Kant, using very valid arguments creates a reasonable explanation to how we can save knowledge and reason from Hume's views, but indirectly ends up supporting the same habitual thinking that was the reason for Hume's ideas. This is to suggest that, while Kant's views sound strong, there is little in the way of an actual answer to Hume, and Hume is yet to be disproved.

Refrences:

• Hume, David. (1740) A Treatise of Human Nature. Clarendon Press, Oxford, U.K., 2007, edited by David Fate Norton and Mary J. Norton.
• Hume, David. (1748) An Enquiry Concerning Human Understanding. Clarendon Press, Oxford, U.K., 2000, edited by Tom L. Beauchamp.
• Kant, Immanuel. (1787) Critique of Pure Reason. Cambridge: Cambridge University Press, 1997, edited by Paul Guyer and Allen W. Wood.
• Kant, Immanuel.(1783) Prolegomena to Any Future Metaphysic.Cambridge: Cambridge University Press, 2004, edited by Gary Hatfield.