Introduction
Critical thinking sounds fancy but it's something most of us do everyday. In fact, most of us are quite good at it...so long as we are properly motivated. Unfortunately, we absolutely suck when the motivation is wrong--and I'm not talking about money.
What do I mean by all this talk of correct and incorrect motivation? What I mean is that, when someone is trying to argue against our own cherished position and beliefs, most of us are very good at pointing out the problems with our opponent's arguments. We are primarily motivated to want to be 'right' and to protect our most cherished beliefs from criticism and so, in such cases, our critical thinking skills are generally quite good.
We are absolutely horrible critical thinkers when an argument or evidence favors our existing cherished beliefs. Why? Because, as I said before, we are primarily motivated to be right and to protect our beliefs, thus, we will usually uncritically accept any argument or evidence that supports our position. Any argument or evidence that serves our purpose is, ipso facto, good. Consider: When was the last time you were critical of an argument or evidence that supported your own position on an issue?
What's the moral of the story here? Most of us already have a intuitive grasp of critical thinking, what messes it up is our motivation to be right and to protect pre-existing conclusions. So, if we want to be good critical thinkers, we need to manipulate our motives; or at the very least be aware of their capacity for distortion.
What is Critical Thinking?
To understand what good critical thinking is, it helps to contrast it with poor critical thinking. The way most people reason is that they look at the conclusion of an argument or the conclusion some evidence implies and assess whether that conclusion agrees with their pre-existing beliefs and positions. If the argument/evidence agrees with or supports their pre-existing position then the argument/evidence is considered good. If the converse is true, then the argument/evidence is considered defective. To summarize the problem: Most people focus on whether they agree or disagree with a conclusion rather than on the quality of the argument/evidence. This approach is not good critical thinking.
In critical thinking we don't care two hoots whether we agree or disagree with the conclusion: all we are interested in is whether the argument or the evidence is good support the conclusion. Critical thinking is mainly about two things: (a) standards of evidence (i.e., what constitutes good evidence?) and the logical relationship and relevance of premises of arguments to their respective conclusions (i.e., does the conclusion follow from the premises?). That's all. The end. Good night. (For convenience, I'll refer to both of these aspects as "quality of evidence/arguments").
The Secret that Professors Don't Want You to Know:
Good critical thinking is all about focusing on the quality of the arguments/evidence relative to conclusions but unfortunately our brains are hardwired to look at the conclusions relative to our pre-existing beliefs. Because we should really focus on the quality of arguments/evidence, we need a trick to overcome the tendency to focus on the conclusion.
The Ol' Switcheroo Version 1: (a) If an argument or evidence supports your position, ask yourself if you'd find the argument/evidence compelling if the same quality of evidence/justification supported the opposite conclusion. (b) If an argument/evidence is against your current position, ask yourself if you'd find the quality of evidence/justification compelling if it supported your position.
For example (a): Suppose you think vaccines cause autism and to support your conclusion you cite the fact that your nephew has autism and he was vaccinated; therefore, vaccines cause autism. To apply critical thinking special secret #1 we construct a similar argument but for the opposite conclusion: E.g., I have a nephew and he was vaccinated and isn't autistic; therefore, vaccines don't cause autism.
If your original position was that vaccines cause autism, would this second argument cause you to change your position? Nope, I doubt it would, and for good reason: a single case doesn't tell us anything about causal relations. Notice that applying secret thinking sauce #1 allows us to focus on the quality on the evidence rather than on whether we like the conclusion. So, if the second argument fails as good support for the conclusion, so does the first, even though it supports your position. Boom! goes the dynamite.
Lets try another example (b): Suppose you are an anthropogenic climate change denier. Someone argues against your cherished beliefs by saying 97% of climate scientists agree that human activity is responsible for climate change. Your natural reaction is to discount this as an insignificant argument because it contradicts your pre-existing position. Now apply critical thinking secret sauce #1 and ask yourself: If 97% of climate change scientists denied that human activity has any effect on the climate, would you consider this as good support for your position?
Lets try a moral example: In the "homosexuality is bad vs homosexuality isn't bad" debate both sides often make appeals to what is or isn't natural behavior as justification for their position. Lets apply critical thinking secret sauce to both sides to show why both justifications are weak:
"Homosexuality is morally wrong because it's unnatural." The justification here is that moral wrongness is a function of whether something is unnatural. Now, applying the ol' switcheroo, we ask the person who takes this position: Supposing homosexuality were natural, would you then agree that homosexuality is morally permissible? They will likely answer, "no" thereby indicating that naturalness is a poor justification for moral permissibility.
But it isn't just evangelical moralists that are using poor justifications for their claim. Lets apply the same test to those who argue that homosexuality is morally permissible because it is natural for a certain percentage of the population to be gay (usually some sort of genetic argument is given). Lets try applying the ol' switcheroo:
Suppose scientists discover that there is no "gay gene" and that homosexual behavior is purely a matter of some combination of socialization and personal choice. If this were the case, would proponents of the argument then say "welp, I guess homosexuality is morally wrong after all"? Probably not. And the reason is that whether a behavior is natural or not tells us nothing about that behaviors moral status.
Whatever one's opinion on the moral status of homosexuality, the ol' switcheroo shows us that both positions cannot be supported through appeals to "naturalness". That is, the quality of that particular justification is weak regardless of which conclusion we are sympathetic to.
The Ol' Switcheroo Version 2: Sometimes issues are such that the simple switcheroo won't work too well in helping to focus our minds on the quality of arguments/evidence; so, we need a variation of the switcheroo to deal with those situations. Here it is: (a*) If an argument/evidence supports your pre-existing position, ask yourself if a similar argument or evidence would be convincing to you in an different issue to which you are opposed. (b*) If an argument/evidence is against your cherished beliefs, ask yourself if a similar argument would be convincing in an issue where you are a proponent.
Basically, in this version we're trying to generalize the principle that is being used to justify a conclusion then apply it to other cases to see if the principle is being applied consistently or (as is often the case) the principle is being used when it supports a conclusion we like but is being denied when it supports a conclusion we dislike.
Example (a*): Suppose you think homeopathy works and that you are generally skeptical of conventional medicine. To support homeopathy you cite a particular scientific study shows that 70% of subjects no longer had condition X after homeopathic treatment. The study has a sample size of 10 and there's no control group. Ask yourself, would such a study convince you that a new conventional medication was effective for condition X?
Of course not. A sample size of ten is way too small to conclude anything of consequence and the lack of a control group makes a study, especially of this size, essentially worthless. If the evidence in the second case wouldn't be good support the conclusion, then the same applies to the first case. Critical thinking secret v.2 allows you to see why the evidence you've provided isn't good.
Example (b*) Suppose again your pre-existing position is that global climate change is not caused by human activity. Someone points out that 97% of climate scientists think the opposite: that global climate change is attributable to human activity. Now apply critical thinking secret sauce v.2: pick an issue where you have a pro position or even one where you don't have position: suppose it's that it's consistent with the 2nd amendment that people should be able to own guns free of restrictions. Ask yourself: if 97% of all constitutional experts agreed that unrestricted gun ownership is consistent with the 2nd amendment, would you consider this to be a good reason in favor of your position? If yes, then you have to also allow that it's a good reason in the first case too.
Showing posts with label rules of thumb. Show all posts
Showing posts with label rules of thumb. Show all posts
Thursday, January 16, 2014
Monday, April 8, 2013
Critical Thinking: Relevance Relevance-Related Logical Fallacies, and Sufficiency
Introduction
Last week we revisited premise acceptability and looked at rules of thumb for acceptance, rejection, or "questionable" status. This week we're going to continue our happy reinvestigation of foundational principles by reexamining premise relevance and sufficiency. Recall from previous discussions that relevance and sufficiency are measures of logical force; that is, the degree to which a conclusion follows from premises. In other words, premises have strong logical force if they force us to accept the conclusion (and the opposite is true of weak premises). Also, lets remind ourselves that when we evaluate logical force (relevance and sufficiency) we assume the premises to be true--regardless of if they are or not.
One more time with the definitions to be clear: Premise acceptability is our assessment of the truth value or reasonableness of a premise. Logical force (made up of relevance and sufficiency) is a measure of how strongly the premises (either individually or in conjunction) force us to accept a conclusion.
A deductive argument's premises are all relevant and sufficient. An inductive argument has premises that are relevant but not 100% sufficient.
Note: The book uses the term "inductive validity" to indicate an inductive argument that has strong logical force. For this course (and everyone else in the philosophy department) doesn't use this term because it causes confusion with actual validity.
In logic, the only arguments that are valid are deductive. So, in line with the UNLV philosophy department and conventional formal definitions, we will reserve the term "valid" for deductive arguments only.
If in your evaluation of an argument you come across an inductive argument that is has very strong logical force, say it has "strong logical force" but do not say it is valid.
Premise Relevance
When we first encountered premise relevance, we understood it as something like whether the premises and the conclusion are talking about the same thing. This is a good starting point, but we're going to modify this approach because sometimes a premise can be about the same topic as the conclusion but fail to be relevant.
Premise relevance increases or decreases the likelihood that a conclusion is true. For example, suppose I want to prove that (MC) a certain animal 'x' is a duck. I give the following premises
P1. The animal waddles.
P2. The animal quacks.
P3. The animal has a stomach.
P4. The animal has a reality TV show and is a primate.
Notice that P1 increases the likelihood that (MC) the animal is a duck. Same goes for P2. P3 is neutral. And P4 actually decreases the likelihood of the conclusion being true. It increases the probability of another conclusion: that the animal is a cast member of Duck Dynasty. Premises that swing the probability against the main conclusion can be used as premises in a counter-argument.
A note on extended arguments: The argument we just looked at is a simple argument because no premise requires further support. In the case of extended arguments some premises require further support. In these cases, the sub-premises will be relevant to the sub-conclusion they support but not directly relevant to the main conclusion. In your evaluation you should indicate this.
Summary and Key Points:
(1) When we evaluate for premise relevance we are looking at whether the information in a premise increases or decreases the likelihood of a conclusion being true.
(2) When we evaluate for relevance we assume the premises are all true--even if they aren't.
(3) When we evaluate for relevance we evaluate each premise individually. It's possible that some premises will be relevant while others might not be.
Informal Logical Fallacies Associated with Relevance (i.e. Failures of Relevance)
Ad hominem (against the person): When a claim is rejected or judged to be false based on an alleged character flaw of the person making the claim. A second form occurs whenever someone's statement or reasoning is attacked by way of a stereotype, such as a racial, sexual, or religious stereotype. A third form involves the use of circumstances of a person's life to reject his claims. Exception: denying someone's claim by calling them a liar and they have a reputation for being one.
Examples:Why should I believe what he says about our economy? He's not even a citizen!
Charities use this a lot in their advertising.
Give Bob a lighter sentence because he's an orphan that grew up in hardship. Have a heart!
Debatable cases: when human emotions are an important factor in the issue.
Appeal to Force: When the arguer essentially presents a threat of force instead of a reason for accepting a position.
Last week we revisited premise acceptability and looked at rules of thumb for acceptance, rejection, or "questionable" status. This week we're going to continue our happy reinvestigation of foundational principles by reexamining premise relevance and sufficiency. Recall from previous discussions that relevance and sufficiency are measures of logical force; that is, the degree to which a conclusion follows from premises. In other words, premises have strong logical force if they force us to accept the conclusion (and the opposite is true of weak premises). Also, lets remind ourselves that when we evaluate logical force (relevance and sufficiency) we assume the premises to be true--regardless of if they are or not.
One more time with the definitions to be clear: Premise acceptability is our assessment of the truth value or reasonableness of a premise. Logical force (made up of relevance and sufficiency) is a measure of how strongly the premises (either individually or in conjunction) force us to accept a conclusion.
A deductive argument's premises are all relevant and sufficient. An inductive argument has premises that are relevant but not 100% sufficient.
Note: The book uses the term "inductive validity" to indicate an inductive argument that has strong logical force. For this course (and everyone else in the philosophy department) doesn't use this term because it causes confusion with actual validity.
In logic, the only arguments that are valid are deductive. So, in line with the UNLV philosophy department and conventional formal definitions, we will reserve the term "valid" for deductive arguments only.
If in your evaluation of an argument you come across an inductive argument that is has very strong logical force, say it has "strong logical force" but do not say it is valid.
Premise Relevance
When we first encountered premise relevance, we understood it as something like whether the premises and the conclusion are talking about the same thing. This is a good starting point, but we're going to modify this approach because sometimes a premise can be about the same topic as the conclusion but fail to be relevant.
Premise relevance increases or decreases the likelihood that a conclusion is true. For example, suppose I want to prove that (MC) a certain animal 'x' is a duck. I give the following premises
P1. The animal waddles.
P2. The animal quacks.
P3. The animal has a stomach.
P4. The animal has a reality TV show and is a primate.
Notice that P1 increases the likelihood that (MC) the animal is a duck. Same goes for P2. P3 is neutral. And P4 actually decreases the likelihood of the conclusion being true. It increases the probability of another conclusion: that the animal is a cast member of Duck Dynasty. Premises that swing the probability against the main conclusion can be used as premises in a counter-argument.
A note on extended arguments: The argument we just looked at is a simple argument because no premise requires further support. In the case of extended arguments some premises require further support. In these cases, the sub-premises will be relevant to the sub-conclusion they support but not directly relevant to the main conclusion. In your evaluation you should indicate this.
Summary and Key Points:
(1) When we evaluate for premise relevance we are looking at whether the information in a premise increases or decreases the likelihood of a conclusion being true.
(2) When we evaluate for relevance we assume the premises are all true--even if they aren't.
(3) When we evaluate for relevance we evaluate each premise individually. It's possible that some premises will be relevant while others might not be.
Informal Logical Fallacies Associated with Relevance (i.e. Failures of Relevance)
Ad hominem (against the person): When a claim is rejected or judged to be false based on an alleged character flaw of the person making the claim. A second form occurs whenever someone's statement or reasoning is attacked by way of a stereotype, such as a racial, sexual, or religious stereotype. A third form involves the use of circumstances of a person's life to reject his claims. Exception: denying someone's claim by calling them a liar and they have a reputation for being one.
Examples:Why should I believe what he says about our economy? He's not even a citizen!
You can't accept her advice. She is so old she has no idea what goes on in today's world.
Why would you listen to him? He's too young to have any wisdom about life.
Type 3: Of course Senator X thinks my administration's tax proposals are bad for the country. After all, his political party lost the last election, and everyone knows that losers are jealous.
You don't want cars to get better gas milage because you are a self-centered rich bastard who isn't affected by gas prices. All you care about it how big your engine is.
Of course you think that people should take drugs. You work for a pharmaceutical company and you make more money if more people take drugs.
Notice in all cases, the issue is not being addressed by bringing up reasons for or against the position, rather it is the person or their circumstances that is being criticized.
Poisoning the Well: Pre-emptive ad hominem to discredit the opponent before they make their point.
Why would you listen to him? He's too young to have any wisdom about life.
Type 3: Of course Senator X thinks my administration's tax proposals are bad for the country. After all, his political party lost the last election, and everyone knows that losers are jealous.
You don't want cars to get better gas milage because you are a self-centered rich bastard who isn't affected by gas prices. All you care about it how big your engine is.
Of course you think that people should take drugs. You work for a pharmaceutical company and you make more money if more people take drugs.
Notice in all cases, the issue is not being addressed by bringing up reasons for or against the position, rather it is the person or their circumstances that is being criticized.
Poisoning the Well: Pre-emptive ad hominem to discredit the opponent before they make their point.
Eg. "Only an ignoramus would disagree with fluoridating water."
Tu Quoque: Another variety of ad hominem fallacy in which one person attempts to avoid the issue at hand by claiming the other person is a hypocrite.
Examples:You're always telling us to do our homework and study but you never did your homework when you were an undergrad.
Many Arab countries put house their prisoners in inhumane conditions. Who are they to lecture us about our prison practices?
Argumentum Ad Populum/Bandwagon Effect:
Ad Populum: Appeal to the people
This fallacy is committed when the arguer appeals to popular opinion to support their claim.
Tu Quoque: Another variety of ad hominem fallacy in which one person attempts to avoid the issue at hand by claiming the other person is a hypocrite.
Examples:You're always telling us to do our homework and study but you never did your homework when you were an undergrad.
Many Arab countries put house their prisoners in inhumane conditions. Who are they to lecture us about our prison practices?
Argumentum Ad Populum/Bandwagon Effect:
Ad Populum: Appeal to the people
This fallacy is committed when the arguer appeals to popular opinion to support their claim.
Eg. (Historically) Everybody agrees that group x shouldn't have equal rights, therefore they shouldn't have them.
Eg. Most people agree that vitamin C cures the common cold. Therefore, you should take it.
Eg. Most people agree that Hondas are better than Fords.
Bandwagon Effect (variation of argument from popularity)
Eg. Most people agree that vitamin C cures the common cold. Therefore, you should take it.
Eg. Most people agree that Hondas are better than Fords.
Bandwagon Effect (variation of argument from popularity)
Often used in advertising through images of beautiful/happy people using a product...”you can be like us too!”
Appeal to Emotions: When the arguer tries to elicit feelings of pity, outrage, compassion, pride, nationalism, etc...instead of providing reasons for or against a position.
Appeal to Emotions: When the arguer tries to elicit feelings of pity, outrage, compassion, pride, nationalism, etc...instead of providing reasons for or against a position.
Examples:
The new PowerTangerine computer gives you the power you need. If you buy one, people will envy your power. They will look up to you and wish they were just like you. You will know the true joy of power. TangerinePower.
The new UltraSkinny diet will make you feel great. No longer be troubled by your weight. Enjoy the admiring stares of the opposite sex. Revel in your new freedom from fat. You will know true happiness if you try our diet!
Charities use this a lot in their advertising.
Give Bob a lighter sentence because he's an orphan that grew up in hardship. Have a heart!
Debatable cases: when human emotions are an important factor in the issue.
Appeal to Force: When the arguer essentially presents a threat of force instead of a reason for accepting a position.
Examples:
If you don't get rid of your chemical weapons we will bomb you.
Missing the point (non-sequitur): Generic catch all for arguments where the conclusion doesn't follow from the premises and the premises seem to suggest a different conclusion should be drawn.
Examples:
Professor Brown is a really hard grader. Not only does he force you to attend class, participate in discussion, and do homework. He actually expects us to think about the material outside of class. So, you can believe that his class teaches students nothing about real life.
Ami likes cheese, ice cream, and yogurt. He should eat more vegetables.
Appeal to (Unqualified) Authority (Arugmentum Ad Verecundiam—Arg. from reverence or respect):
Celebrity endorsements: Jenny McCarthy on vaccines, Oprah on psychology and medicine.
Sufficiency
Sufficiency is the measure of whether there is enough evidence to guarantee the truth of the conclusion (or at least make it very likely). When an argument's premises are all relevant and 100% sufficient, the argument is deductive. When an argument's premises are relevant but do not 100% guarantee the conclusion, the argument is inductive (and invalid). Being able to make this distinction is one reason why evaluating sufficiency is important.
A few more comments: in an argument it's possible (but very rare) for a single premise to be sufficient to guarantee the conclusion. Eg:
P1. Bob and Joe are humans
C. Bob is a human.
Or an argument may require several premises working together to be sufficient for the conclusion:
P1. If the Canucks win the Stanley Cup, I will be happy.
P2. The Canucks won the Stanley Cup.
C. I am happy.
Notice in this argument that P1 and P2 on their own don't do much to guarantee the truth of the conclusion, but working together, the conclusion is guaranteed; i.e., together they are sufficient. When you make an analysis of sufficiency, you should point out whether it is a cluster of premises that are sufficient for the conclusion or whether it is a single premise. However, at this point in the course, you can probably bet that you won't be seeing too many single premise arguments and if you do, it means there's a hidden second premise...
Evaluations of sufficiency will take into account all the premises for the argument. In some arguments single premises will be sufficient for the conclusion whereas in other arguments, sufficiency will be achieved through several premises working together. In your analysis you don't need to say of each premise why it on its own isn't sufficient. But when premises are obviously meant to work together, you should evaluate their sufficiency collectively.
Some Heuristics
Because it's not always straight forward to determine whether a premise or set of premises are sufficient, here are a couple of general rules we can apply in our analysis:
1. How strongly worded or broad is the conclusion?
One thing we might look at is how strongly stated or broad the conclusion is. Suppose someone wants to conclude that "all men are evil." This is a very strong/broad claim. The arguer will have to provide enough evidence to show that every man on earth is "evil" (lets ignore the vagueness problem for now). That's going to require quite a bit of compelling empirical evidence.
Now suppose this person argues for their conclusion based on personal experience. "All my boyfriends were mean to me" or something like that. This is insufficient evidence to support the broad and strongly worded claim that all men are evil for several obvious reasons, one of which is that the arguer hasn't encountered all men (we can make that reasonable assumption, anyway).
To simplify this rule we can interpret it this way: universal claims in the conclusion should set off alarm bells in your head, as should strong (emotional) language. Also, the strength of evidence should be in proportion to the strength of the claim being made. We can call this the proportionality principle. Or to quote Carl Sagan paraphrasing Hume: extra-ordinary claims require extra-ordinary evidence
Heuristic: Words like "all","every", "most", "none", "never" should set off alarm bells when evaluating sufficiency.
2. Don't Jump to Conclusions
This rule is closely related to the previous one. It's basically a close relative of the proportionality rule (strongly worded/broad claims require a lot of evidence). If you have a small sample of evidence pointing to a particular conclusion, don't prematurely jump to that conclusion. Recognize that results from a small sample size can easily be over-turned by a large sample. In other words, when stating a conclusion based on initial evidence, temper the language of the conclusion.
For example, you grew up in a forest and want to argue that all birds can fly. Your evidence for your claim is that every bird you've ever seen can fly. Of course you haven't yet seen the humble chicken or majestic turkey or know of their existence. But as an intelligent forest-dweller you know that the world extends beyond the forest so you should opt for a more reasonable claim in proportion to your evidence: "every bird in the forest can fly." Or you might qualify your claim and make a weaker universal claim: "It's likely that every bird can fly."
Using qualifying language is the mark of a careful thinker. Even more so when it is used to bring a conclusion in line with the amount of current available evidence.
Heuristic: Words like "up to x%" "some" "a few" should set off sufficiency alarm bells.
3. Ensure The Arguer Isn't Engaging in Distortion or Omission
If an argument only presents positive evidence/reasons and doesn't at least acknowledge (reasonable) competing evidence/reasons, then we have reason to be skeptical of the sufficiency of the premises.
Suppose a boy wants to argue that "all girls are mean" because all the girls in his class tease him. His evidence is that each girl in his class teased him at some point. However, he's missing the following evidence: his teacher, mother, and sister are all girls and they are nice to him. Also, sometimes little Susie shares her lunch every week with the boy, even though she teased him only once. As we investigate, we find that sometimes girls are nice to the little boy. Not only that, but sometimes girls that the little boy never met in his life are always nice and never tease any of their classmates in other schools and classrooms. All this evidence is missing and undercuts the strength of the conclusion.
So, if we were evaluating the little boy's argument for sufficiency we'd say there is insufficient evidence for his conclusion because he hasn't considered contravening evidence.
Missing the point (non-sequitur): Generic catch all for arguments where the conclusion doesn't follow from the premises and the premises seem to suggest a different conclusion should be drawn.
Examples:
Professor Brown is a really hard grader. Not only does he force you to attend class, participate in discussion, and do homework. He actually expects us to think about the material outside of class. So, you can believe that his class teaches students nothing about real life.
Ami likes cheese, ice cream, and yogurt. He should eat more vegetables.
Appeal to (Unqualified) Authority (Arugmentum Ad Verecundiam—Arg. from reverence or respect):
Celebrity endorsements: Jenny McCarthy on vaccines, Oprah on psychology and medicine.
Sufficiency
Sufficiency is the measure of whether there is enough evidence to guarantee the truth of the conclusion (or at least make it very likely). When an argument's premises are all relevant and 100% sufficient, the argument is deductive. When an argument's premises are relevant but do not 100% guarantee the conclusion, the argument is inductive (and invalid). Being able to make this distinction is one reason why evaluating sufficiency is important.
A few more comments: in an argument it's possible (but very rare) for a single premise to be sufficient to guarantee the conclusion. Eg:
P1. Bob and Joe are humans
C. Bob is a human.
Or an argument may require several premises working together to be sufficient for the conclusion:
P1. If the Canucks win the Stanley Cup, I will be happy.
P2. The Canucks won the Stanley Cup.
C. I am happy.
Notice in this argument that P1 and P2 on their own don't do much to guarantee the truth of the conclusion, but working together, the conclusion is guaranteed; i.e., together they are sufficient. When you make an analysis of sufficiency, you should point out whether it is a cluster of premises that are sufficient for the conclusion or whether it is a single premise. However, at this point in the course, you can probably bet that you won't be seeing too many single premise arguments and if you do, it means there's a hidden second premise...
Evaluations of sufficiency will take into account all the premises for the argument. In some arguments single premises will be sufficient for the conclusion whereas in other arguments, sufficiency will be achieved through several premises working together. In your analysis you don't need to say of each premise why it on its own isn't sufficient. But when premises are obviously meant to work together, you should evaluate their sufficiency collectively.
Some Heuristics
Because it's not always straight forward to determine whether a premise or set of premises are sufficient, here are a couple of general rules we can apply in our analysis:
1. How strongly worded or broad is the conclusion?
One thing we might look at is how strongly stated or broad the conclusion is. Suppose someone wants to conclude that "all men are evil." This is a very strong/broad claim. The arguer will have to provide enough evidence to show that every man on earth is "evil" (lets ignore the vagueness problem for now). That's going to require quite a bit of compelling empirical evidence.
Now suppose this person argues for their conclusion based on personal experience. "All my boyfriends were mean to me" or something like that. This is insufficient evidence to support the broad and strongly worded claim that all men are evil for several obvious reasons, one of which is that the arguer hasn't encountered all men (we can make that reasonable assumption, anyway).
To simplify this rule we can interpret it this way: universal claims in the conclusion should set off alarm bells in your head, as should strong (emotional) language. Also, the strength of evidence should be in proportion to the strength of the claim being made. We can call this the proportionality principle. Or to quote Carl Sagan paraphrasing Hume: extra-ordinary claims require extra-ordinary evidence
Heuristic: Words like "all","every", "most", "none", "never" should set off alarm bells when evaluating sufficiency.
2. Don't Jump to Conclusions
This rule is closely related to the previous one. It's basically a close relative of the proportionality rule (strongly worded/broad claims require a lot of evidence). If you have a small sample of evidence pointing to a particular conclusion, don't prematurely jump to that conclusion. Recognize that results from a small sample size can easily be over-turned by a large sample. In other words, when stating a conclusion based on initial evidence, temper the language of the conclusion.
For example, you grew up in a forest and want to argue that all birds can fly. Your evidence for your claim is that every bird you've ever seen can fly. Of course you haven't yet seen the humble chicken or majestic turkey or know of their existence. But as an intelligent forest-dweller you know that the world extends beyond the forest so you should opt for a more reasonable claim in proportion to your evidence: "every bird in the forest can fly." Or you might qualify your claim and make a weaker universal claim: "It's likely that every bird can fly."
Using qualifying language is the mark of a careful thinker. Even more so when it is used to bring a conclusion in line with the amount of current available evidence.
Heuristic: Words like "up to x%" "some" "a few" should set off sufficiency alarm bells.
3. Ensure The Arguer Isn't Engaging in Distortion or Omission
If an argument only presents positive evidence/reasons and doesn't at least acknowledge (reasonable) competing evidence/reasons, then we have reason to be skeptical of the sufficiency of the premises.
Suppose a boy wants to argue that "all girls are mean" because all the girls in his class tease him. His evidence is that each girl in his class teased him at some point. However, he's missing the following evidence: his teacher, mother, and sister are all girls and they are nice to him. Also, sometimes little Susie shares her lunch every week with the boy, even though she teased him only once. As we investigate, we find that sometimes girls are nice to the little boy. Not only that, but sometimes girls that the little boy never met in his life are always nice and never tease any of their classmates in other schools and classrooms. All this evidence is missing and undercuts the strength of the conclusion.
So, if we were evaluating the little boy's argument for sufficiency we'd say there is insufficient evidence for his conclusion because he hasn't considered contravening evidence.
Subscribe to:
Posts (Atom)