Kennisclips wetenschapsfilosofie
1.1:
We need to understand the difference between good and bad arguments. Logic is the study of
argumentation.
Argument consists of two parts
1. Premises: what we suppose
2. Conclusion: what we conclude from the premises
I) Geen enkele vaders is kaal
II) Peter is een vader
III) CONCLUSIE: PETER IS NIET KAAL
Valid argument: conclusion follow from the premises
Invalid argument: an argument that is not valid, it does not follow the premise
Two kinds of arguments
- Deductive arguments: an argument in which the truth of the premises absolutely guarantees
the truth of the conclusion.
o Geen enkele vader is kaal. Peter is een vader. Conclusie peter is niet klaar
o Is you use deductive arguments, you cannot make new mistakes. The only way for
the conclusion of a deductive argument to be false, is if one of your assumptions is
false. If assumptions are true, conclusions are guaranteed to be true.
o Whether a deductive argument is valid or not can be determined just by looking at
the form, ignoring its content.
▪ No A is B
▪ C is A
▪ SO C is not B
- Inductive argument: an argument where the truth of the premises gives good reason to
believe the conclusion, but does not absolutely guarantee the truth of the conclusion. Truth
of the premises makes the conclusion likely but it does not guarantee it. Generally the case in
science.
o None of the medieval texts we have studied argues against the existence of God.
o So nobody in the Middle Ages argued against the existence of God
,1.2:
Inductive argument
- To see of ay inductive argument is any good, we always need to use background theories.
This means that we cannot use induction to derive theories from neutral data, we are always
already relying on some theoretical beliefs.
- SO, scientific theories don’t just depend on observations, they also depend on some
presuppositions or prejudices
- Limited number of observations and draw a general conclusion from them
o I put 25 frogs in the freezer for a week, and all of them died.
o So all frogs die when they are put in the freezer for a week
- So drawing general conclusion from data is always going to require induction, there is always
a risk involved
We always need background theories about the content to assess whether a supposed inductive
argument is any good.
- I did A 25 times and B happened
- So every time I will do A, B will happen
Valid argument scheme? NO
- I ask 25 people a question in Dutch and every time they answered in Dutch.
- So every time I will ask someone a question in Dutch, they will answer in Dutch.
So why is the argument of frogs acceptable and the argument about speaking Dutch not? It has
nothing to do with the logical form of the argument, because that is the same. Instead, what matters
in the content. Our background knowledge on frogs and cold makes us believes that frogs cannot
survive that long in a freezer. On the other hand, our background knowledge on human languages
tells us that speakers of a single language are often concentrated in geographical areas, so we cannot
generalize observations of one area to the whole world.
In order to assess this kind of inductive argument there are several things we need to know about the
content that the argument is about. We need to know two things.
- First, we need to know how probable it is that the things we are interested in, behave
uniformly.
Background theories 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.
Representative data is that represent the subject matter as a whole and not just a special part
of it.
o I want to know the life expectancy of people in 16th century France and suppose that
all my data are about rich aristocrats, then we know that the data is not
representative.
o Suppose that all my data is about people whose name started with an A, that would
not make the data non representative since we do not expect that first letter of a
name to have influence on how long they life.
o SO we are using background theories about life expectancy. Wealth is relevant, name
is not
General point: inductive arguments in which we draw general conclusions from our observations can
only be judged based on certain background theories.
- Other people, with different backgrounds theories might draw different conclusions. So by
looking at the logic of induction we can conclude that scientists will always be a bit biased
and are looking from a certain perspective.
, 1.3:
Some ways in which human reasoning goes wrong:
Confirmation bias
- The tendency to collect and take seriously evidence that supports our beliefs, and to
disregard or ignore evidence that conflicts with our beliefs.
- Good scholars try to overcome confirmation bias, keep an open mind. Science tries to combat
confirmation bias by fostering a lot of different opinions.
Confusion of correlation with causation
- Correlation is the tendency of two things to occur together
o Lung cancer and smoking
o Learning and good grades
- In both examples we have a correlation but also causation, smoking is not correlated with
lung cancer, it causes lung cancer.
- Causation is the relation between causes and effects
- Correlation however does not imply causation. A historian cannot conclude that
unemployment causes fascism because perhaps there is something else that causes both
unemployment and fascism, or it is an accident that there is a correlation.
- Especially true in disciplines like history where you cannot do experiment. We cannot take
100 countries and turn up the unemployment in 50 of them, to see what happens to fascism.
Probabilities
- The fact that a hypothesis is unlikely does not mean that it is false. We need to think through
whether the alternatives might not be even more unlikely
1.4:
Inductive reasoning always happens against a background of prior theories. When we draw a general
conclusion from our data, we use background theories to see whether we have enough data, and
whether these data are representative. This means that a scientist will always look at data from the
perspective of the theories you already believe. So science is not purely driven by observation,
scientists are biased by a theoretical perspective.
BUT Karl Popper thought that this theory was not true. According to Popper, real scientists are always
trying to show that their own theories are false, falsify their own theories. If not, pseudo scientist.
- Was aware that you can only do induction if you accept some background theories but he
also wanted to say that scientists do not except any theories at all since they are always
critical. How can we combine this?
Popper said we need to stop claiming that scientists use induction. They do not. They only use
deduction and thus do not need background theories.
- 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
Normally we state that scientists engage in this kind of reasoning all the time. We have a limited
amount of observation and we draw a general conclusion from the observation.
ACCORDING TO POPPER, this is wrong. Why? Scientists do not draw general conclusions from a
limited set of observations. They should not be interested in claiming that any theories are true.
Scientists never claim that a theory is true because they know that there is the chance of a new
observation that shows them to be wrong.
Instead, scientists are interested in critically testing theories and showing that they are false. The only
conclusion they ever draw are conclusions that a certain theory is wrong
