300 likes | 345 Views
CAUSAL REASONING. 1. Many moral arguments are based on causal relations. E.g.: Playing violence video games increases the tendency to become violent. Therefore, these video games should be banned.
E N D
Many moral arguments are based on causal relations. • E.g.: • Playing violence video games increases the tendency to become violent. Therefore, these video games should be banned. • Allowing same-sex marriage encourages homosexual behavior. Therefore, same-sex marriage should not be legalized. 2
Questions • What is a cause? • How many types of causes are there? • How can we know which causes which? 3
causal law properties: A B causes events: a b • Causal relations involve both events and their properties (or attributes). • An event a causes another event b when they have properties that are implicated in a causal law (or a law of nature). 4
E.g.: • Suppose Katrina was a typhoon. Katrina caused damage, e.g., the falling of trees, because the property TYPHOON and the property DAMAGE are implicated in a causal law. causal law TYPHOON DAMAGE caused Katrina falling of trees 5
However, other properties of Katrina (and other properties of falling of trees) may not be implicated in a causal law. • Suppose Katrina is reported on the 1st page of SCMP. • But there is no causal law connecting the property of being reported on the 1st page of SCMP and the property of damage. 6
Hence, Katrina in terms of being a typhoon but not in terms of being reported in 1st page of SCMP caused falling of trees. 7
Straight vs. non-straight laws • Causal laws can be divided into straight laws and non-straight laws. • If “A causes B” is a straight law, no background conditions need to be fulfilled in order that A causes B. • In other words, A is a sufficient cause of B. 8
Straight laws can only be found in physics. • E.g.: Force =mass x acceleration. 9
If “A causes B” is a non-straight law, some background conditions, say, C1 and C2, need to be fulfilled in order that A causes B. • C1 and C2 are structural causes and A is called a triggering cause. • Both structural causes and triggering causes are necessary causes. • Most causal laws, and all laws in social sciences, are non-straight laws. 10
E.g.: In order for my dialing of your phone number to cause the ringing of your phone, both my phone and your phone must be charged and your phone is not turned on the “silent” mode. • Triggering cause: dialing your phone number • Structural causes: • My phone is charged • Your phone is charged • Your hone is not turned on the “silent” mode. 11
E.g.: In order for freedom of speech to bring constructive discussion, offensive language must not be allowed. • Triggering cause of constructive discussion: freedom of speech. • Structural cause: banning of offensive language. 12
If a non-straight law involves many structural causes, it is difficult to know whether all structural causes are present. • Consider the example of making a phone call again. • In this case, we can never be sure that the presence of A will bring about the presence of B. • I.e.: A causes B only in a probabilistic manner. 13
We can use Mill’s methods, which were developed by John Stuart Mill, to find out what are an event’s causes. • There are 5 methods: • The method of agreement • The method of difference • The jointed method • The method of concomitant variations • The method of residues
The method of agreement • E.g.: • E: You have diarrhea. • Situation: Having breakfast. • A: egg • B: bread • C: milk (cause) • However, A may also be a sufficient (although not necessary) cause of E.
The method of difference • E.g.: • E: having GPA 3.5 or above • Situation: individual student • A: taking logic class • B: doing part-time job • C: sleeping and getting up at regular hours (cause)
If the previous table represents an experiment rather than a survey, situation 3 represents a control experiment. • We try to set up a control experiment as similar as the original experiment but without C in order to see whether the effect will still occur.
The joint method • E.g.: • E: exhibiting violent behavior • Situation: individual child • A: have been neglected • B: having low IQ • C: have been abused (cause) • D: being a member of a single family
In this case, neither the method of agreement nor the method of difference can tell us which candidate is the cause. • The joint method is a combination of these two methods.
The method of concomitant variation • E.g.: • E: happiness • Situation: different individual • C: close relationship with family members
The method of residues • C1 and C2 cause E. • C1 causes part of E. • C2 causes the rest of E. • E.g.: • Increase in salary and increase in price of consumer goods both cause 4% rise in inflation rate. • Increase in salary causes 1% rise in inflation rate. • increase in price or consumer goods causes 3% in inflation rate.
This method is often used to discover new causes. • E.g.: • Suppose Leo is very happy today. • You know that he has bought a new mobile phone. • But you also know that he will not be so happy just because of the new phone. • So you can conclude that there must be another cause of his happiness. • This leads you to search for the new cause.
Limitations of Mill’s methods • When we use Mill’s methods, we have to make sure that the background conditions are the same in all the situations in which we carry out observation.
In fact, Mill’s method can only inform us the correlation between 2 events. But correlations do not imply causation because: 1. The correlation may be accidental. • E.g.: The sea level in Venice and the cost of bread in Britain have both been on the rise in the past 200 years.
2. The 2 events may have a hidden common cause. • E.g.: Children who wear bigger shoes tend to have better reading skills. • This is because older children read better and have bigger feet. • E.g.: In the US, the sale of ice-cream goes up with the rate of rape. • This is because, in the winter, people eat less ice-cream and outdoor rape will not occur.
If A correlates with B and causes C, B may be thought to be the cause of C. • E.g.: A medication may be thought to be a cause of symptom reduction, which is actually a psychological effect caused by the treatment procedures. • This is called the “Placebo effect.”
Mill’s methods also do not tell us, in a causal relation, which is the cause and which is the event. • E.g.: Drug use correlates with having psychiatric problems. • Does drug use causes psychiatric problems? • Or, does preexisting psychiatric problems cause people to turn to drugs?
E.g.: Playing violence video correlates with having violence tendency. • Does playing violence video games cause one to develop violence tendency? • Or, having such tendency makes one enjoy playing violence video games?