Eric Bleys
Philosophy 221
10/7/2016
Professor Marco Dees Bard College
The problem of induction is the question of whether or not it is rational to reason from limited, particular premises, to generalizations which assume the similarity of unknown factors to known factors. Another way of stating this problem would be to ask if it is rational to believe that un-experienced cases of something will be the same as the cases which we have experienced. For example, if we observe five thousand dogs and all of them possess the same fundamental quality x, can we then conclude that all dogs possess quality x? What if we reason that all objects follow Newton’s laws of gravitation because the objects that our group of scientists have observed all follow this law? Or we can provide other examples which do not make universal claims. For example, all the people in neighborhood J which I have known have been wealthy, therefore the next person I meet from neighborhood J probably will be wealthy. We can object to these conclusions by simply pointing out that there may be a dog which does not have the quality of x and there may be an object which does not follow Newton’s law of gravitation. Disproving a universal is far easier. As soon as I observe one particular dog without the quality of x then I can conclude that not all dogs possess the quality of x, which is the negation of the universal all dogs possess the quality of x. This differentiates deductive from inductive reasoning. For in deductive reasoning the conclusion follows of necessity from the premises. With induction the core problem that is at hand is not whether or not it is as rigorous as deduction but whether or not it is even capable of giving us justified beliefs.
Why would one consider inductive thinking to be irrational? One example would be the fact that events do not always follow the patterns which we ascribe to them. There are instances of snow in the spring. Our weather forecasts are not always accurate. Furthermore, the appeal to the solution of probability still does not explain why the future must necessarily conform to the patterns of the past. Just because it rained five out of the past seven days does not mean that the week after will be structured according to the same pattern. In fact the superior way to reason about the amount of rainy days which may occur next week is to observe the direction of the storms which do exist in the present time by considering their location and the direction of the wind. There is nothing about the amount of rainy days last week that of necessity implies anything about the number of rainy days which will occur next week. The probability of of something occurring under one condition does not of necessity determine the probability of an event under future conditions.
Sometimes the future is like the past and sometimes the future is different from the past. And therefore we cannot know if the future will resemble the past. This is an argument against the rationality of induction. However, if we argue that there is a specific element of a particular type of past event that is by its nature consistent then it would seem we could predict the future based upon this event. The rising of the sun is consistent whereas the falling of a tree happens haphazardly and without the same level of consistent patterns. However, if there is no essential form which the pattern corresponds to then why would we deem it to be a law? And why would we expect the future to correspond to this pattern? Just because every time I have ridden my bicycle I have never been in a serious accident does not mean that such an accident will never happen. Perhaps It is unlikely because it proves my capability to ride a bicycle.
In the practice of science particular observations and experiments are used as support for generalized theories of how things work. This leads us to an interesting question which is essential in the philosophy of science. Do the logical problems with induction pose a threat to the validity of science? If science is so dependent upon induction then how can science be true? How do we reconcile the irrationality (or seeming irrationality) of induction with the tangible reality that it works (or at least seems to work) in our day to day lives?
One way of thinking about the problem and of trying to provide a solution to this dilemma is to discuss the idea of induction as probability. What if we are to think of the problem as a question regarding the ability of the past to predict the future? Then we can say that under the assumption that the future will be the same as the past that our inductive generalizations about the world will probably predict the future. If we posit this as our solution to the problem of induction then we must understand that what follows from this is that our scientific theories are incapable of definite determinations about the future. Our belief that after I drop a pen (because of gravity) the pen will fall, will be nothing more than a probabilistic guess about how the future will occur. Does this contradict our beliefs about the purpose of science?
Another way that we could respond to this question is by pointing a way to a metaphysical solution. One could posit the existence of abstract forms which are learned about indirectly through the empirical world which, because they themselves are unchanging by their definition, provide a foundation for the consistency of the world as experienced by the success of our predictive theories about reality as derived from inductive reasoning which does not theoretically have such explanatory power. This solution would have fundamentally different implications about the nature of science. For one, it would have to imply scientific realism in regards to the abstractions known as laws. Although a law within itself need not have any empirical embodiment it must actually exist. For if their were not abstract realities which actually exist demanding conformity on the future then this way of resolving the issue would not be possible. However, this does not necessitate the literal equivalence of our laws with the forms but only that our laws would have to resemble the eternal order in some way.
So one way of addressing the problem of induction would lead to the conclusion that science cannot make definite predictions about the future; the other is that it may be possible for science to make definite predictions if our laws correspond to the abstract entities which control the order of the empirical world. The answer to this question would depend upon the ability of the empirical world to teach us something about this abstract world. We can see that Hume also understood that the rationality of inductive thinking depended upon the premise that nature is uniform. The stanford encyclopedia of philosophy cites Hume as writing, ““Now as concerns the argument, its conclusion is that in induction (causal inference) experience does not produce the idea of an effect from an impression of its cause by means of the understanding or reason, but by the imagination, by “a certain association and relation of perceptions.” The center of the argument is a dilemma: If inductive conclusions were produced by the understanding, inductive reasoning would be based upon the premise that nature is uniform;
that instances of which we have had no experience, must resemble those of which we have had experience, and that the course of nature continues always uniformly the same. (THN, 89)”” (Stanford encyclopedia Problem of induction 2. Hume on induction).
Because Hume understood the necessity of the uniformity of nature thesis for justifying induction he would have understood that there would need to be a certain kind of stability behind the world in order for this to work. Abstract entities are by definition stable. They are stable because abstract realities are not empirical and the empirical world by definition changes. However, as an empiricist Hume would not have posited the existence of such an abstract world because it is unknowable through observation alone.