Chap 8,9,10 (Ph)

Chapter 8

Exercise I (pages 218-9)

1. Inductive, since the word “probably” suggests that the argument is not intended as a valid proof of its conclusion. Also, any use of one observed fact to conclude another matter of fact will always fail the standard of validity, so it doesn’t makes sense to hold it to such a standard.

2. Deductive, since it is deductively valid; that is, the conclusion can never be false if the premise is true.

3. Inductive, because this is invalid and has the form of a statistical generalization from a sample of past presidents. Again, any prediction based on past observations will always fail the standard of validity.

4. Inductive, because this is a statistical generalization from a single case.

5. Inductive, because this is an inference to the best explanation. Other explanations are logically possible, making this invalid.

Exercise II (page 224-5)

6. Ref Class – Americans

Sample – my friends

This is a biased sample since it is made up only of my friends (not all Americans are my friends), who are more likely to like me or have a higher opinion of me and thus think I’d make a good president. The sample is probably also too small, because most people do not have enough friends for a national poll. There is bias in investigation since I am conducting the survey of my own friends and they are more likely to tell me flattering things than the truth. And there is risk of bias in interpretation since I may interpret the replies of my friends in a way that is more favorable to me (because of my ego).

7. Ref Class – people that voted in the election

Sample – people at the exit polls

The sample size is indeterminate, but usually exit polls contain a sufficient enough number of people to justify a generalization (that’s what they are designed for!). There is no sample bias as far as we know since there doesn’t seem to be any connection between the people at the exit polls and people who didn’t go to the exit polls. In other words, there’s no reason to think that whether or not you go to the exit polls would be relevant to who you voted for. There’s no bias in investigation since it is obvious whether or not someone talked to the candidate (and what they said). And I’d say there’s probably not bias in interpretation since if someone took the time to go and speak with the candidate, then they probably are telling the truth about how they voted (I don’t imagine someone would talk to a candidate they didn’t vote for).

8. Ref class – All of Mary’s children (including the baby)

Sample – Mary’s older children

One problem with this might be that the sample size is too small (even though the sample is the overwhelming majority of the reference class!). Bias in the sample might happen if the baby has some sort of condition that the other children do not (that might affect intelligence or cognitive capacity), but otherwise it is ok. There is probably going to be big problems with bias in investigation and interpretation however. The investigation might just be the mother’s observations of her children (which is not probably sufficient as a test of genius), plus the interpretation of those observations will probably be skewed in favor of her own children (aka wishful thinking on the part of the mother).

9. Ref class – the people

Sample – everyone we talked to

We can’t really evaluate sample size or sample bias because there is no information given about the people talked to (note: the exam problems won’t have something like this happening!). There is huge bias in investigation here however, because of the phrasing bias in how the question was put to people. Any interpretation bias worries are a derivative of this bias in investigation.

10. Ref class – potential murderers

Sample – the hundreds of convicted murderers in states without the death penalty that we talked to

The sample size is probably ok given how hundreds of personal interviews is an incredible amount of research for a study of this kind! Sample bias is present however given that the sample concerns people who have already gone through the experiences of committing the murder, getting caught, tried, convicted, and punished – whereas not all potential murderers have had such experiences. These experiences probably would affect whether or not a person would be deterred from a motive to act a certain way (to kill). Bias in investigation comes from asking the convicts directly – they may have an interest in not honestly answering given that their parole might be at stake if they answered that they would not be deterred even by a death penalty. Interpretation bias comes from doubts about whether any person would be able to accurately report on how they would have acted differently under different circumstances.

Exercise IV (pages 228-9)

4. Ref class - Socialists with blue eyes.

Subset – Maureen

The percentage here is 3%, which is close to 0%, so this makes the argument stronger. And although the socialist aspect of the reference class may be relevant, the blue eyes part is not. A more relevant reference class would only include socialists.

5. Ref class – things John says

Subset – John’s claim that his father is named John

The percentage of 98% is doing well since this is close to 100%. Here the reference class is far too large to be relevant, since most things people say are true. The argument is still strong, however, if we assume that John is in a position to know his father’s name and has no reason to lie in this situation. Then it is relevant to look at John’s track-record at making honest claims to figure out if his current statement is true.

6. Ref class – things John says

Subset – John’s claim that the Giants are going to win

As in (5) the reference class is too expansive to be the most relevant. However, this case is different from (5) because it involves a certain subset of beliefs (namely who will win a baseball game) that more people tend to have false beliefs about than what they say in general. This is probably true of John as well. This makes the argument weak. A better reference class would be perhaps John’s predictions about sporting events since it wouldn’t mix the claims that John is probably in a position to know with other claims that he likely isn’t in a position to know.

8. Ref class – people

Subset – my physics professor

This argument overlooks a more relevant reference class, that of physics professors (many of whom probably understand quantum mechanics), so it is weak. The percentage is given informally with the word “most”. “Most” is a very ambiguous proportion, so depending on whether the true proportion is closer to 100% or to 50% will determine how well the argument is doing with this standard. However, the point is mostly moot since, as with most statistical applications, a problem with the reference class will trump any strengths with regard to the cited percentage.

9. Ref class – birds

Subset – this penguin

“Almost all” implies a ratio pretty close to 100%, so the argument is doing well in that respect. However, the reference class is not the most relevant since the category of being a “bird” doesn’t not only track features of animals that would be relevant for abilities at flight. The size of the animal relative to its wingspan is itself a more relevant category for determining whether it can fly.

Chapter 9: Causal Reasoning

Exercise I (page 234)

1. T

2. F

3. F

4. T

14. T

15. F

16. T

17. F

18. T

Exercise III (pages 238-9)

1.

a. B and D are eliminated by the SCT in Case 2

b. B and D are eliminated by the NCT in Case 3

2.

a. No candidate is eliminated by the SCT.

b. A, B, and D are eliminated by NCT in Cases 2, 3, and 1+3, respectively

3.

a. A, B, C, and D are all eliminated by the SCT in Case 1.

b. A, B, and D are eliminated by NCT in Cases 2, 3, and 3, respectively

4.

a. A and B are eliminated by SCT in Case 3.

b. A, B, and C, are eliminated by NCT in Cases 2, 2, and 1, respectively

5.

a. A, B, C, and D are all eliminated by the SCT in cases 1, 2, 1+2, and 1, respectively.

b. B and C are eliminated NCT in Case 3.

6.

a. A, C, and D are eliminated by SCT in Cases 3, 2+3, and 2, respectively

b. C is eliminated by NCT in Case 1.

7.

a. A, B, and D are eliminated by SCT in Case 1 + 3, 1-3, and 1 + 2, respectively.

b. None of the candidates is eliminated by NCT.

8.

a. A, B, C, and D are all eliminated by the SCT in Case 1+3, 1, 1, and 1, respectively.

b. A and B are eliminated by NCT in Case 2.

Exercise IV (pages 239-40)

The only candidate that is not eliminated as a sufficient condition of failure is the plug being out. (That cannot be eliminated because the plug is never out.) The only candidates that are not eliminated as necessary conditions of failure are the old software and the plug being in. (The latter cannot be eliminated because the plug is always in.) It does not follow that any of these candidates is necessary or sufficient, since the plug positions are not tested rigorously, and the claim that the old software is necessary for failure might be refuted by further cases, such as when the computer fails after taking the plug out.

One condition that might be both necessary and sufficient is: Old Software and (either New Monitor or New CPU), so the problem might be some incompatibility between the Old Software and the New Equipment. I would advise updating the software to solve the problem, but there is no guarantee that this will work.

(Note: Many more exercises can be constructed on this model simply by changing the results in the right-hand column of the table in the exercise.)

Chapter 10: Inference to the Best Explanation and from Analogy

Exercise II (page 264)

Note from Tim: remember that evaluations here are highly dependent on background assumptions, so your answers might be different from mine. If that happens, double check to make sure that you’re using the criteria properly!

1. The explanation has a story to tell since there is only one thing to explain. It is deep because if a witch made everyone invisible, then it would make sense that you couldn’t see anyone. It is of limited power, but still could probably be available to explain other things (like losing my keys and things like that where I expect to see something and then I don’t). The explanation is falsifiable since if I could see people, I couldn’t explain that by how they were invisible! It isn’t modest because the key bit is the invisibility part, not this extra stuff about witches that are wicked. Depending on background assumptions there could be either simplicity or conservative issues here. For me, I believe that witches with such powers don’t exist, so this hypothesis would mean adding something new to my beliefs about the world (not simple) and contradicting my previous beliefs (not conservative).

IMPORTANT NOTE : this is a nice full explanation and the sort of thing that I want to see on the exam. In what follows below I will only mention the standards that have issues, but treat my answer above for #1 as what you want to be doing.

2. This explanation has issues with the first criterion since even though it explains the absence of the students, it doesn’t explain the absence of the instructor. There isn’t a real problem with simplicity since I already believe that students sometimes intentionally skip class. However, it does violate the conservative standard since I have beliefs that while one, two, or a few students may decide to skip on a single day, having all of them skip at the same time is highly improbable.

3. This explanation is doing a fantastic job on all fronts! This depends on the background knowledge that Columbus Day is a holiday where school is canceled.

4. The biggest issue here is the lack of depth. If all I knew about a house was that it was painted red, I wouldn’t thereby expect that the house would fall down. One might also complain that there is an issue with the conservativeness of this explanation since I have beliefs about color that deny such powers to the property.

5. This explanation postulates a novel kind of earthquake, which would violate many of our beliefs about plate tectonics and seismology, so it lacks conservativeness. Furthermore, this explanation is not powerful because it only applies to one specific case. Notice that there isn’t a depth issue here!

6. Simplicity is the primary concern here. Though the new subatomic particles may not force us to give up any of our previous beliefs about particle physics, it is implausible to assume that a previously unobserved type of particle has suddenly appeared in a form specifically destructive to the boards in a single abode. Surely some other explanation is simpler.

7. This explanation is immodest. While it might be plausible that the segment of river you fished had a lack (or more likely, paucity) of fish, we assume too much in claiming that the entire river is completely devoid of fish. Probably there would be an issue over conservativeness too since you probably fished there thinking that there were fish (you wouldn’t fish somewhere you didn’t believe there were fish).

8. The explanation is not simple since it postulates the existence of river gods. And while some may hold this belief, it is not commonly accepted. Furthermore, the explanation is not conservative if, as seems likely, it conflicts with other beliefs you hold. This explanation is also not falsifiable, if “the will of the river gods” is highly inscrutable and could have been invoked to explain just about any outcome (i.e. it isn’t drawing a clear line about why one thing happened instead of something else).

9. If being unlucky can explain not catching fish, then being lucky can explain catching fish, so explanations by luck are unfalsifiable and never really explain why something happens rather than its opposite (just like the worries in #8).

10. It is not likely that you can tell which airline is flying overhead or where it is headed, so it is immodest to assume it’s United to LA. All that is needed for the explanation is that it is an airplane.

11. This explanation is probably not going to be simple (as most people don’t believe that aliens positively exist – or at least exist on Earth). Also, there are huge modesty issues here. Again, that it is a flying craft is all that is really needed – even saying it is a UFO is more modest. That it is a spaceship, moreover, one of alien design is adding unnecessary details to the explanation that don’t do any heavy lifting in the explanation.

12. This is not a very deep explanation, because it doesn’t not explain why your eye affliction produces only this one distortion. If I knew I had something wrong with my eyes, that would immediately make me think I’d see things moving more quickly than they really are. I’d also add that this is not simple since I’m not aware of an eye affliction with this sort of effect. Finally, in as much as this hypothesis is able to explain the light moving quickly at all (which is not much – see depth), it is also capable of explaining things if the light had been moving slowly. This means there are also worries with the explanation being falsifiable.

Exercise V (page 272)

These questions are intended to display the variety of contexts in which arguments from analogy are used. Complete answers to these questions would be complex and often controversial. Here are a few brief suggestions:

1. Given the background information that Cézanne often painted series of landscapes in similar styles, and that artists are often consistent within a period, this argument is pretty strong, even though it is based on only one other painting. The analogy is based on artist, time, and subject matter. In detail, having the same artist is a very relevant and important similarity for determining the beauty of a painting since the vision and skill of the artist is largely responsible for the result. That the paintings were from similar times is also important since artists can grow and lose their skill or abilities over time. The similarity of scene is relevant and important since a painter can be very good at one type of painting but bad at another; one great with landscapes could be terrible with portraits and vice versa. Other relevant factors in determining the beauty of a painting exist, but few could compete with these for their general impact. One possible disanalogy could come from the time spent painting the painting. If Cezanne generally takes his time with a painting but this painting he made very quickly, well then I could relate it to other hasty paintings I’ve seen that are sloppy and not very evocative. Judgments of beauty can be hard to determine objectively, but another possible disanalogy might come from if you’ve seen other Cezanne paintings that you didn’t like!

2. Siamese cats have a bad reputation in some circles, and that background belief might seem to make this argument strong. Also, there aren’t a lot of details provided about this cat that would give material for analysis. However, all things considered, I believe this argument is not very good since while the breed of a cat is relevant to the behavior of the animal, it isn’t the most important. Other factors matter more, (and we can make disanalogies out of them too!). Of these more important variables I’d pick out things like, whether the cat is mistreated or not, whether the cat is well fed, if it comes from a home with children (that it plays with), whether it is an inside or outside cat, etc. If some of these factors are present in this cat (like say this cat is well fed/loved) then I can make disanalogies out of the many very loving and wonderful cats I’ve met in my life (that were well fed/loved).

3. The background information in this case—study habits, intelligence, content of the course—seems relevant. Intelligence and a strong work ethic speak to the abilities one can bring to academic ambition and the similarity in course material speaks to the sort of challenge that will test those abilities. These are also fairly important factors in whether or not someone will be able to succeed in academics, but there may be some other traits to be sensitive to as well. The quality of the instruction is perhaps just as important. Also the attitude of the student, in my background assumptions, counts for a lot. A student who is intelligent and hardworking but is arrogant or has trouble responding to criticism or who has poor attitude about collaboration might not be able to succeed as easily depending on the subject matter (like if it is a philosophy course!). Disanalogies can again be built out of these factors. Here’s one: I had a student who was a straight A student, but could not get comfortable in my Phil 101 course with the idea that she needed to discern whether SHE thought that these philosophers were right or wrong. She was used to just accepting what she was told as truth and was very good at those sorts of tasks (smart and hardworking), but she ended up dropping the class halfway through because her grades had slipped to a C from not following the instructions for the philosophy papers the course required. Another possible disanalogy: if the person speaking is addicted to meth, then (from analogy) I’d expect them to fail the course even if they are smart and hardworking and the course has similar material.

4. Since there is a significant difference between cheating in the context of marriage and cheating in the political realm, the argument that the politician will display similar moral turpitude in both cases is fairly weak. This analogy is not completely irrelevant, however, since both reflect a disrespect for moral constraints (this is the ground of the relevance). However, the importance can be challenged since there are other factors that effect a person’s vulnerability to temptation – for instance, the consequences if one is caught. Cheating in a marriage has big consequences but not as big as if one was caught cheating in politics (that might involve jail time!). Disanalogies are probably going to come from people who have been highly principled and effective idealistic leaders while still engaging in affairs (since we don’t know much else about this person except that they cheated in their marriage). We could also just bring up many examples of how humans can be highly principled generally but still have moral “blind spots”.

FUN ALTERNATIVE ANSWER (this shows how different takes on a problem can still be using the standards appropriately): The main cited similarities in this problem are 1) the same person is in both cases 2) both scenarios are arenas of importance with high stakes and consequences and 3) there are similar temptations that would motivate cheating. Similarity #1 is relevant on the grounds that we’re talking of the same person in both contexts and this means that the personality we’re evaluating is the same. A single person will exhibit similar patterns of behavior in different circumstances since they have the same background beliefs and values wherever they go. This similarity is not so important however given my background assumptions that people are not usually very integrated or unified in their behavior and commonly exhibit moral “blind spots” for specific things, or act with double standards in different situations. Similarity #2 is relevant on the grounds that scenarios with higher stakes might deter cheating more than contexts with low stakes. If the person is already willing to cheat on something as serious as a marriage, then they probably are not going to be deterred by the high stakes of public office. I’m not sure how important this factor is however. I do think that the more serious the situation, the greater deterrence effect is possible, but a lot of times people that cheat are convinced that they will not be caught. So, this might not be the most important factor. Similarity #3 is very relevant and important. It’s relevance is about as direct as could be since cheating is something we almost never do for the sake of cheating itself and usually requires a powerful motive! The need for a motive for cheating is a major variable in whether or not someone is going to cheat. Disanalogies might include how this person acts in the business world since it is also a public sphere of life like political service. A marriage is a more private matter, so this could make a difference. If they are clean in the business world, that would blunt the force of this analogy.

5. Economies are complex, and it is not clear which factors affect the impact of a high minimum wage. This makes it hard to say what impact a high minimum wage will have in different countries. Thus, to make this argument stronger, more detail is necessary. Still, in the absence of any relevant disanalogy, the argument does have significant force. Minimum wage could be relevant to unemployment on the grounds that if a business has to pay its employees more, then it can’t hire as many. But this might not be the most important factor effecting unemployment. How much money is being given to other employees in the company can just as easily effect this! Further detail in evaluating this analogy starts to rely on more specialized knowledge (aka background assumptions!) about how economies function. Disanalogies would derive from comparing other details about the second country’s economy with economies that have decreased unemployment.

6. To some, this argument will seem quite strong since it seems as though the properties on which the analogy is based are all relevant (presumably nerve endings on the head, a functioning pain receptor in the brain, etc.). Common sense says that such similarities are actually quite important too! What is more central to feeling pain than the neurological architecture of my body? However, to others (specifically those who are approaching this at a deeper level than common sense), this will force them to wonder how similar their bodies are to other bodies, and beyond that, how similar their minds are to other minds (notice this is minds not brains). For the exam, don’t feel like you need to get all crazy philosophical! It’s ok to stick to the world of common sense as long as you don’t hold such intuitions dogmatically.

7. What are the morally relevant features here? Making claims you know are false. Acting in the best interests of the person you are lying to. Those are the big ones. Are those features the most important in determining the moral status of an action? Well, this might be challenged. I can imagine a debate occurring around the potential consequences of lying or telling the truth. Maybe the content of those consequences matter. There a much smaller risk involved in telling me the truth about my finances than about my health (as I’m a little more attached to my life than my money!). Disanalogies might come from cases where the gravity of the consequences for the patient are greater – like the analogy that paternalism is based off of (i.e. the relationship between a parent and a child).

8. Another moral analogy. What are the cited similarities here? Defects that caused harm and risk of harm. The relative seriousness of that potential harm is also the same. Harm and risk of harm is a classic moral concern so there is definitely a relevance connection. And certainly harm issues are of central moral importance. We also think that causing harm makes someone morally liable and that they ought (morally) to try to make amends to the harmed parties. So the similarities are relevant and important. As far as I can tell with the details we currently have in this example, there is not much room for relevant disanalogies. The fact that they are different companies and that the defects are in the gas tank instead of the suspension doesn’t seem morally relevant.