Chapter 1.2: Induction and background theories

[Music] in the previous lecture we talked about the difference between a deductive and an inductive argument in this lecture we will focus on inductive arguments and especially on one crucial feature of them that in order to see whether an inductive argument is any good you always need to use background theories this means that we can't use induction to derive theories from neutral data we are always already relying on some theoretical beliefs and that means that scientific theories don't just depend on the observations they also depend on some presuppositions or even prejudices but let's start at the beginning remember that a deductive argument is a valid argument in which the truth of the premises absolutely guarantees the truth of the conclusion it is simply impossible for the premises to be true and for the conclusion to be false an inductive argument on the other hand is a valid argument in which the truth of the premises gives good reason to believe the truth of the conclusion but doesn't guarantee it here's the example of an inductive argument we saw in the previous lecture none of the medieval texts we have studied argues against the existence of God so nobody in the Middle Ages argued against the existence of God these are slightly different maybe more representative kind of example I put 25 frogs in the freezer for a week and all of them died so all frogs die when they are put in the freezer for a week in that argument we take a limited number of observations and draw a general conclusion from them that is an example of induction not deduction after all there might be freezer resistant frogs and it just happened to be the case that I didn't do my test on any of them in general no matter how many of patience we have done it is always possible that the next observation is going to be different so drawing general conclusions from data is always going to require induction there's always some risk involved now in the last lecture we learned that we can recognize deductive arguments just by looking at their logical form without needing to know anything about the content in this lecture we'll see that that's not the case for inductive arguments instead we always need background theories about the content in order to assess whether a supposed inductive argument is any good so let's look at the example of the frogs again I put 25 frogs in the freezer for a week and all of them died so all frogs die when they are put in the freezer for a week what is the logical form of that argument it's something like this I did a 25 times and every time B happened so every time I will do a B will happen is that a valid argument scheme well let's try it out for example let's say that a is putting my hand on a hot stove and B is it hurts so the argument becomes I put my hand on a hot stove 25 times and every time it hurt so every time I am going to put my hand on a hot stove it will hurt and that sounds pretty good in fact it would seem that perhaps I didn't meet 25 trials to find out that it's going to keep hurting most people might have stopped after 2 or 3 attempts but now let's fill in something else for a and B that a B ask someone a question in Dutch and let B be the answer in Dutch so the argument becomes I asked 25 people a question in Dutch and every time they answered in Dutch so whenever I will ask someone a question in Dutch they will answer in Dutch and that of course is a terrible argument I did my experiment in the Netherlands and in the Netherlands almost every once Dutch so that's why I got the results that I did but it would be wrong to conclude that everyone speaks Dutch if I go to Beijing and do the same experiment I will get very different results so why is the argument about hot stoves acceptable but the argument about speaking Dutch not well it has nothing to do with the logical form of the arguments because their logical form is the same instead what matters is the content our background knowledge about the human body makes us believe that is if something hurts once it will also hurt the next time on the other hand our background knowledge about human languages tells us that speakers of a single language are often concentrated in one or a few geographical areas and that means that you can't generalize linguistic observations made in one area to the whole world you have to make observations all over the world before we believe your inductive argument in order to assess this kind of inductive argument then there are several things we need to know about the content that the argument is about specifically we need to know at least two things first we need to know how probable it is that the things we are interested in behave uniformly it is quite probable that my nerves will react to the same stimulus in the same way every time it is highly unprovable that there is a language which everyone in the world speaks this is why we need less evidence in the case of the stove than in the case of the Dutch speakers so we can see that our background theories about human bodies and linguistic communities play an important role in deciding how much evidence we need in these two cases second we need to know whether the data that we have are representative that is whether they represent the subject matter as a whole and not just a smaller special part of it suppose that I want to know what the life expectancy was of people in 16th century France and suppose that all my data are about rich aristocrats then we know that those data are probably not representative socioeconomic status can make a big difference to life expectancy to have representative data we would also need information about middle-class townspeople servants farmers and so on on the other hand suppose that all my data we're about people whose names started with an A maybe because I only have a small part of an alphabetically ordered registry that would not make the data a non representative since we don't expect the first letter of someone's name to have any influence on how long they live obviously we are using background theories about life expectancy to make these judgments we already know that wealth is relevant and your name is not the general point then is this inductive arguments in which we draw general conclusions from our observations can only be judged based on certain background theories we already need to know a lot about frogs or the human body or linguistic communities or life expectancy or whatever before we can decide whether our data are good enough to draw a certain general conclusion about any of these subjects and that is a very important point for understanding science for it means that scientists cannot just take some data and then draw conclusions from those data in a completely neutral way that is acceptable to everyone when we draw conclusions we rely on background theories and other people with different background theories might draw different conclusions so just by looking at the logic of induction we can conclude that a scientist will always be a bit biased will always be looking at the world from a certain perspective there is no such thing as a completely neutral scientist who doesn't have any background opinions about the world and who looks only at the data for the data alone will never allow us to draw any general conclusions and without general conclusions we wouldn't really have science

Channel: Leiden University - Faculty of Humanities

Views: 66,179

Rating: 4.9423804 out of 5

Keywords: Universiteit Leiden, Leiden University (College/University), Humanities (field of study), Geesteswetenschappen, Bachelor, Education, gijsbers, victor, wetenschapsfilosofie, induction, philosophy

Id: DRx-3jvC918

Length: 9min 20sec (560 seconds)

Published: Wed Sep 27 2017