The Justification Problem of Induction and the Failed Attempts to solve it. There has been much discussion on the problems of induction. In order to avoid diluting my essay into a summary of these problems, I will choose instead to concentrate on the problem of induction that is often associated with Hume, and consider some of the popular responses to this. The problem of induction concerns the justification of inductive reasoning. No other person has been more associated with this problem than Hume. For this reason, I will spend some time discussing and assessing the problem which Hume lays out in ‘A treatise of Human Nature’. After the problem has been laid out, I will move on to looking at popular attempts to solve the justification problem of induction. I will attempt to explain the problems that these solutions come across and show why they do not come up with an acceptable answer to the problem of induction. What I will not be doing in this essay is discussing in any great detail Goodman’s New Riddle of Induction. I believe that the ‘old problem of induction’ is still a problem that lacks a satisfactory answer, and therefore I believe it is a necessary to find a solution to this problem before adding more to it, as Goodman does with his New Riddle of Induction. I start my essay by briefly discussing an important difference, which may easily lead many astray if not fully understood. This is the difference between context of discovery and context of justification.
If you need assistance with writing your essay, our professional essay writing service is here to help!
The differences concerning discovery and justification is one which philosophers rarely merit mentioning, yet it can often lead confusion and cause philosophers to talk ‘past’ each other on entirely different aspects of the problem of induction. It is therefore important to have a clear understanding of the distinction between the context of discovery and the context of justification. As Paul Hoyningen-Huene (1987) explains in his paper ‘Context of Discovery and Context of Justification’, there are four or five different components of the distinction. I will only concern myself with the common component here and that is of two different historical processes, the process of discovery and the process of justification. The process of justification is concerned with the logic behind inductive reasoning. In other words, the process of justification may be seen as trying to answer the questions such as ‘Is it possible to derive, by some transformational rule, scientific hypotheses from sensory experience’ (Kelly, 2012) and if it is possible to formulate such a rule, what does the rule transmit from the premises to the conclusion? Does it transmit truth or some probability? In the process of justification, importance lies with the purely logical questions regarding inductive reasoning. No thought is given to the psychological account by which the hypothesis are derived but rather how the belief in these hypotheses can be justified inductively.
The process of discovery on the other hand is a process by which the consequence of a inductive inference is substituted with phychological cause and effect. The premises of an inductive argument are seen as the sensory experiences of one self. In the process of discovery no logical questions are asked. They are phychological question and are very much subjective. Questions such as ‘Why did the pen fall to the floor?’ can be answered with ‘I saw the pen being dropped’. Reichenbach who is often attributed with defining the distinction between the context of discovery and the context of justification gives the following as an example of the distinction.
‘The way, for instance, in which a mathematician publishes a new demonstration or a physicist his logical reasoning in the foundation of a new theory, would almost correspond to our concept of rational reconstruction; the well-known difference between the thinker’s way of ¬nding this theorem and his way of presenting it before a public may illustrate the difference in question (1938, p. 6).’
‘there is a sharp distinction between the psychological and social circumstances in which a discovery is made (context of discovery or external history of science) and the logical basis for confirming belief in the actual discovery (context of justification or internal history of science). The aim of science is to test, confirm and justify theories using rational criteria while ignoring any interfering social and psychological factors; thus, only the context of justification (internal history) is of interest to scientists.’
Now to show that this distinction has been clearly understood I will use an example to help show the distinction. In answering the question ‘What would happen if I jumped from a height?’ a rational person would say something like ‘You would fall to the ground’. This inductive inference can be set up as follows:
Premise – On the many cases where I have seen someone jump from a height, they have fallen to the ground.
Conclusion – Therefore, when someone jumps from a height, they fall to the ground.
(For the sake of this argument, we can assume that the person’s fall to the ground is unobstructed).
This is an example of simple enumerative induction. It can not be classed as deduction, as the conclusion goes beyond what is said in the premise. The premise does not guarantee the truth of the conclusion in the way in which they do in an deductive argument. With appeal to the context of discovery what we can say has happened in this case is that on the many other times where we have experienced someone jump from a height, we have experienced them hit the ground shortly after (cause and effect). There is not a problem here. However, a problem does appear when we appeal to the context of justification. We have no rational basis what-so-ever for accepting the conclusion. The premise does not guarantee the conclusion. We arrived at the conclusion using induction, and that is fine. We may choose to arrive at a conclusion using whatever method we so choose. What matters is how we can justify on a rational basis that the theory is true. This is the Problem with Induction that Hume hits on in his paper, the justification problem of induction.
Before delving into the problem and Hume’s sceptical solution, I should spend some time discussing his version of the problem of induction. The problem of induction is far from being a new problem in Philosophy. Discussion on the problems of inductive reasoning can be found as far back as Empiricus. The problems however did not gain much momentum until Hume wrote about it in A Treatise of Human Nature (1739). Hume however did not use the word ‘induction’ in his work. Nor did he give much attention to the so called problem of induction. Infact, Hume accepted his Sceptical Solution to the problem and left it at that. Hume was an empiricist, and a strong empiricist at that. Like all empiricists Hume claims that all knowledge comes from experience. And this is his reason for landing on the problem of induction. In an ‘Enquiry concerning Human Understanding’ (1748) Hume starts by the divinding the ‘objects’ of human enquiry into Relations of Ideas and Matters of Fact. Relations of Ideas are those propositions that are true independent of anything else. ‘Propositions of this kind are discoverable by the mere operation of thought, without dependence on what is anywhere existent in the universe’ In other words, propositions are relations of ideas iff the denial of that proposition is impossible. Deductive arguments are relations of ideas. No more will be said on relations of ideas. What I am concerned with in relation to the problems of induction are Matters of Fact. Matters of fact are those things we learn by use of our senses. In other words, a proposition is a matter of fact if it’s denial is possible, that is, both it and its denial are possible without contradiction. An example of a matter of fact is ‘If you jump from a height, you will fall to the ground’. It is not contradictory to believe the denial of this. Hume’s own example is ‘That the sun will not rise tomorrow is no less intelligible a proposition, and implies no more contradiction than the affirmation, that it will rise.’
The problem soon comes to light that if propositions of matters of fact cannot be shown to be true, how are we to find out whether they are true or not? Being an empiricist, Hume thinks that our senses and what we have in our present memory is able to provide evidence in support of the truth. This still does not help with matters concerning beliefs of the future. All our knowledge comes from past experiences and we have no knowledge of any future events. Hume says that beliefs we have of future events which are beyond the scope of our senses and knowledge present in our memory ‘seem to be founded on the realtion of Cause and Effect’. He cites several examples of causal inferences including believing that your friend is in France because you have received a letter from him. The question Hume now asks himself is ‘how we arrive at the knowledge of cause and effect’. Hume says that it is impossible for us to observe these causal connections. We cannot observe the causal connection between a flame and heat. Had we never had experience of a flame before we would have no rational grounds on which to relate it with heat. According to Hume, it is not possible to justify any inference beyond those objects of which we had experience. In order to justify induction one has to provide a deductively valid or an inductively strong argument to the effect that our inductively strong arguments will continue to lead us to true conclusions (most of the time) in the future. And this is what leads us to Hume’s Skeptical solution (if one can call it a solution at all). While induction has no rational basis (context of justification), we cannot help but make inductive inferences (context of discovery) just as we cannot help attributing causal relations – seeing the world in causal terms. Inductive reasoning lacks rational foundation and justification but that will not and cannot stop us using inductive reasoning. Hume’s sceptical solution gives us little hope of justifying induction so I will now consider responses to Hume’s problem and discuss whether or not the problem of justification of induction can be solved. I will start by discussing some of the more common (weaker) responses to the problem and move towards what I believe to be the strongest attempt at solving the problem of induction, and that is BonJour’s A Priori solution.
The first response I will consider is what some call the NaÃ¯ve Inductive Response to the Problem of induction (Anderson, 2000), and it is one of the weaker attempts to solve the problem by way of induction. This is a response that many people who know very little concerning the problem of induction would reply with. In the vast majority of cases where you would ask such a person the question “What justifies your belief that when you drop a pen, it will fall to the floor” the person would respond by saying something along the lines of “Because whenever I have done it in the past, the pen has always fallen to the floor”. In the past, this person has had experiences of a pen being dropped and it hitting the floor. They have seen this happening, they have heard the pen hit the floor and it’s very likely they have also dropped the pen themselves and observed it hitting the floor. After a number of occasions of observing this happening they come to the conclusion that when someone drops a pen, it hits to floor. When they next observe a pen being dropped they will be quick to assume that it will indeed hit the floor, and when it does hit the floor they will conclude that their assumption was correct. There is another similar example which a scientific realist would use to justify induction (which as will be seen suffers from the same problem), and that is the success of science itself (Maria Baghramian, relativism about science, //books.google.co.uk/books?id=XlTfb3-OavAC&pg=PA92&dq=success+of+science+response+problem+of+induction&hl=en&sa=X&ei=Zt6KUNLbLoPS0QX1wID4Cw&ved=0CEAQ6AEwAw#v=onepage&q=success%20of%20science&f=false). They argue that induction is justified by the very success of science. The inductive methods used in science have time and time been used to predict unobserved instances successfully. Scientific experiments are conducted with the greatest of care so surely the very fact that science has been so successful gives us grounds to rationally accept the inductive methods? Unfortunately it does not. Like the previous example, not only does this example fail to solve the problem of induction, but it helps in showing what the very problem is. Both examples are trying to solve the problem of induction by means of induction. Therefore, the question remains, how can we justify the inductive solution to the problem of induction? As can be seen, all attempts to solve the problem of induction by inductive means results in a circular argument. ‘Induction is justified because it has worked in the past’ is Hume would say “is begging the question” (quote). Even though all attempts to solve the problem of induction by inductive means will lead to this circularity, this has not stopped many philosophers from attempting to do so (Max Black being an example Black ,1963 //books.google.co.uk/books/about/Induction.html?id=dEMNAQAAMAAJ&redir_esc=y). Infact, Hume’s ‘Uniformity of Nature’ suffers from this problem. If our belief that nature is uniform is based on our past experiences then it too suffers from being circular. As Hume remarks “It is impossible, therefore, that any arguments from experience can prove this resemblance of the past to the future; since all these arguments are founded on the supposition of that resemblance.” (from Dicker, p85, Georges Dicker, 1998. Hume’s Epistemology and Metaphysics: An Introduction. Edition. Routledge.) Because of this, other Philosophers have tried to solve the problem by using different means. Peter Strawson is one such philosopher and I will now discuss his solution, and show where the problems arise.
Strawson’s solution to the Problem of Induction is to say that there is no problem of induction to solve. Strawson puts forward his solution in (Introduction to logical Theory, 1952). Strawson argues that Hume misunderstands what it means to be ‘reasonable’ (or rational). In other words, the problem is due to a misunderstanding over ordinary language and what the word ‘reasonable’ entails. It is ‘reasonable’ for a person to base their beliefs on evidence, be it inductive of deductive. For a deductive argument to be reasonable, it must be true that the conclusion follows from the premises, this is not the case with an inductive argument. It is reasonable for a person to believe that once they drop a pen from their hand, it will drop to the floor, and it is unreasonable to think otherwise. Just because an inductive argument may not be able to be proven to be true, Strawson says that this is no reason to not accept it as being reasonable. Asking if induction is justifiable is like asking if it is ‘reasonable to be reasonable’ which is akin to asking ‘is the law legal’, it simply makes no sense to ask this. Hume was right in saying that there is no problem of justification which concerns deduction. Given true premises, deduction will always prove the truth of the conclusion. Deduction is justified in performing deduction.
I do not believe Hume would find this to be a satisfactory solution to the problem of induction. I also don’t believe that the exact problem that Hume had in mind. (From //ls.poly.edu/~jbain/philsci/philscilectures/08.Induction.pdf)
As Salmon notes, Strawson fails to take into account the difference between ‘validation’ and ‘vindication’. Validation and vindication are both ways of justifying a principle. As Salmon notes “While validation cannot be carried out without appeal to principals more fundamental than those whose justification is at issue, vindication is not subject to any such limitation (from //books.google.co.uk/books?id=ngXNqiH_cDQC&pg=PA363&lpg=PA363&dq=strawson+validation+and+vindication&source=bl&ots=Z2P77diSZh&sig=8dNhoBewlUhFRBhHD7YCQmSbecc&hl=en&sa=X&ei=GPqKUOj8JYew0QWDv4DoBQ&ved=0CCIQ6AEwAQ#v=onepage&q=strawson%20validation%20and%20vindication&f=false )”
Our academic experts are ready and waiting to assist with any writing project you may have. From simple essay plans, through to full dissertations, you can guarantee we have a service perfectly matched to your needs.
Vindication is to show how a certain method achieves its goals. Strawson’s solution to the problem does not touch on vindication. In deduction, principles are vindicated when it is shown that it is impossible for the premises to be true and the conclusion to be false (//books.google.co.uk/books?id=Uq7xf0rcCQIC&pg=PA61&lpg=PA61&dq=salmon+vindication+and+validation&source=bl&ots=P5yod3RZ6H&sig=qJLOWULmDaSmJKfNbsvQPMAw184&hl=en&sa=X&ei=tQKLUNL0G8Wr0QX074BQ&ved=0CB0Q6AEwAA#v=onepage&q=salmon%20vindication%20and%20validation&f=false)
When Strawson says “Asking if induction can be justified makes no sense” what he is actually saying is that “Asking if induction can be validated makes no sense”. The other method of justification, vindication, as can be seen, is left untouched. As there is two different way of justifying, there are two different definitions of reasonable to go with them. Hume’s real question was asking “Can induction be vindicated”. When Hume asks (as Strawson puts it) “Is it reasonable to be reasonable?”, what he is actually asking is “Does it serve our goal of predicting correctly as often as possible (reasonable corresponding to vindication) to use accepted rules of inductive inference (reasonable corresponding to validation)” (from //books.google.co.uk/books?id=Uq7xf0rcCQIC&pg=PA61&lpg=PA61&dq=salmon+vindication+and+validation&source=bl&ots=P5yod3RZ6H&sig=qJLOWULmDaSmJKfNbsvQPMAw184&hl=en&sa=X&ei=tQKLUNL0G8Wr0QX074BQ&ved=0CB0Q6AEwAA#v=onepage&q=salmon%20vindication%20and%20validation&f=false)). And this is not trivially true.
As Strawson’s attempt to solve the problem of induction fails to take into account vindication, the question “Can induction be justified (vindicated)” remains. So now it would seem reasonable to find out if induction can be vindicated or answer the question “What vindicates induction”. Hans Reichenbach attempts to do just this with his pragmatic justification of induction.
Unlike others, Reichenbach does not try to solve the problem by using an inductive or a deductive approach, rather, he takes a pragmatic approach to try to solve it. If Reichenbach was asked “What vindicates induction” his response would be something like “It is at least just as good as any of the alternative methods”. (From ls.poly.edu). In other words, if we do choose to use the inductive method in reasoning, there will always be at least some chance of it being a success, and this is what vindicates induction. (proginosko.org). His reasoning behind this vindication can be easily shown by way of example. If we use induction as a way of projecting past experiences onto future unobserved events, then induction will be true if there is indeed a uniformity of nature. Induction will be shown to be false if there is not a uniformity of nature. Now, if we take any other method of prediciting what will happen in the future (card reading, wishful thinking, crystal ball prediction etc.), if there is a uniformity of nature, it will either work or it won’t work. If there isn’t a uniformity of nature, it won’t work (If it did work and successfully continued to predict future events, then this would show that there was some uniformity of nature). If it does work, it will show that there is a uniformity of nature, which will provide a basis for induction to work. The inductive method is therefore successful in projecting the regularity of past events to future events (//ls.poly.edu/~jbain/philsci/philscilectures/08.Induction.pdf) wherever any other method would also be successful. Therefore, if we use induction, there is nothing to lose and everything to gain. The conclusion being, it is possible that there is no chance of success in predicting future events based on past experiences (showing that there is no uniformity of nature), but if it is possible (there is a uniformity of nature) then it is rational for us to use the inductive method over other methods. Unfortunately there are a number of problems with this approach to solving the problem of induction.
The pragmatic approach is liable to coming under fire from arguments against other pragmatic arguments, one of the most famous ones being Pascal’s Wager. Pascal’s wager may give us reason to believe in God, but it does not tell us whether or not God exists, just as the Pragmatic Justification of Induction may give us reason to use induction over other methods, but it does not tell us whether or not induction is reliable. Reichenbach also agrees with Hume in that there is no epistemic justification for induction. For Hume however, for the problem of induction to go away, he would want an epistemic justification for it. The epistemic justification that Hume is looking for is “a reason for believing that induction is reliable” (//www.proginosko.com/docs/induction.html), not the pragmatic justification that “it is our best bet to use induction”.
So far, all the different attempts (considered here) to solve the problem of induction have failed to do so. The attempts have all taken different approaches. The first tried to solve the problem with an inductive argument, leading to circularity. The second attempted to show that Hume was confused over the meaning of reason, but only managed to show that asking if induction can be validated makes no sense, and failed to touch on the vindication of induction. The third attempt provided us with a pragmatic solution to the problem of induction, but it did not provide us with a reason for believing that induction is reliable. Of course, there are many other solutions which I have not considered here, including Popper’s deductivism, Will’s inductive response (Frederick L. Wills, ‘Will the Future Be Like the Past?’, in Flew (1953), pp. 32-50.) and Goodman’s inductive intuition (Goodman (Fact, Fiction and Forecast)) but I believe even these fail to provide us with a satisfactory solution to the problem of induction. The last attempt that I will consider in this essay is also one of the more recent attempts, and it is BonJour’s (1998) a piori justification of induction.
BonJour brings forward his solution in “In Defense of Pure Reason” (Laurence BonJour, In Defense of Pure Reason (Cambridge: Cambridge University Press, 1998). BonJours view is that all attempts to solve the problem of induction which are based on a posteriori factors will fail, and therefore we should attempt to look for an a priori justification. Hume and many of the philosophers who have tried to solve the problem of induction have accepted “that there is no a priori means for proceeding from the premises of an inductive argument to its conclusion.” //stephanhartmann.org/HHL10_Lange.pdf “Matters of fact, which are the second object of human reason, are not ascertained in the same manner [knowable a priori]; nor is our evidence of their truth, however great, of a like nature with the foregoing.” Hume
BonJour attempts to show that when we have a standard inductive premise such as “All observed copper objects have been observed to melt at 1083 degrees centigradeâ€¦one can derive a priori a standard inductive conclusion” – //sammelpunkt.philo.at:8080/1657/1/meixfull.pdf
I will not go into detail here on how BonJour attempts to do this, but I will consider the conclusion that he finally comes to. His conclusion is:
“‘[an a priori] solution to the problem of induction depends on the tenability of a nonHumean, metaphysically robust conception of objective regularity (or objective necessary connection). Bonjour , pp.214-215
What this says is that if there is an a priori solution to the problem of induction, there needs to be real and constant causal relation between the objects in the universe. Finding this out is no easy task, and BonJour himself gives no clues on how one might try to find this out.//www.proginosko.com/docs/induction.html#FN11
The importance of what BonJour has done here is to show that one should not necessarily take Hume’s view “that there is no a priori means for proceeding from the premises of an inductive argument to its conclusion” when attempting to come up with a solution to the problem of induction. An a priori justification of induction would make the problem go away, but I believe BonJour shows what the very underlying problem with the problem of induction is.
In this essay I have explained and assessed Hume’s Problem of Induction. I have also looked at four of the prominent attempts at finding a solution to the problem. I have also explained why none of these attempts successfully solve the problem. I believe if we are to find a solution to the problem then BonJours attempt is the way to go. BonJour outlines the underlying problem with justifying induction, and that is whether or not a real and constant causal relation between the objects in the universe. If there is, then an a priori justification of induction can be found. It is likely that this crossed Hume’s mind when he formulated his sceptical solution, and as he believed such an a priori justification was not possible, his sceptical solution is all he had. To be able to justify induction, we need to be able to say that there is a uniformity of nature. How we should go about finding this out, is the problem.