Professional Documents
Culture Documents
PHI345 - Metaphysics - Writing Tips
PHI345 - Metaphysics - Writing Tips
Do NOT break Rule #1 in your paper! A deductively valid argument is an argument that is
logically airtight in the sense that the mere grammatical structure of the different claims in the
argument guarantees that its conclusion must be true if the premises are true, regardless of the
particular concepts they employ. Most of science does not rely upon deductively valid arguments
(except when doing mathematical calculations). Rather, science mostly relies upon inductive or
probabilistic forms of reasoning. Logic is a matter of possibility and necessity. Science is generally
more concerned with what is probable or likely given our evidence. I’m not denigrating science
when I say this. Probabilistic reasoning is generally far more useful. Moreover, it is wrong to think
that logic gives you certainty. Logic can only guarantee that your conclusion is true if all of your
assumptions are true. You have to rely upon other evidence to get anything out of logic other than
tautologies, and even then, your conclusion will only be as certain as your assumptions.
For Valid Arguments: Junk in → Junk Out
If you start with false assumptions, you can still get a false conclusion, even if the argument is
valid. A sound argument is an argument that is both valid and has all true premises (and thus a
true conclusion). Logic can’t save you from making false assumptions, and it can’t tell you how
to make your argument sound unless your conclusion is a tautology (and thus rather uninteresting).
This means that a valid argument isn’t automatically a good argument. In fact, the primary reason
why valid arguments are useful in philosophy is because they help us identify bad arguments!
Making a valid argument forces you to draw out all of your tacit background assumptions to close
the logical gaps between your premise and conclusion. Quite often, when we do this, we find that
we were relying upon principles that might not seem all that plausible upon closer inspection.
All of that is well and good. But how do you know whether an argument is valid or invalid?
The good news is that there is an entire course where you can learn to do this (PHI 210 Symbolic
Logic). The bad news is that there is an entire course where you learn to do this. We can define
validity in propositional logic in terms of truth tables. An argument {φ1, φ2, φ3, …, φn, Con} is
valid if and only if there is not a single row in the truth table where the premises φ1-φn are all true
and the conclusion Con is false.
Did you follow that? Probably not, because you likely haven’t taken symbolic logic. Here is one
of the most basic valid arguments (one name for it is ‘and elimination’ or ‘&EL’ for short):
PHI 345 WRITING TIPS
Here is an example:
Premise 1: John and Mary both went to the store.
Conclusion: Therefore, John went to the store. [1/&EL]
Premise Con
VALID: No counterexamples
A B A&B A
T T T T Not a counterexample because conclusion is true
T F F T Not a counterexample because conclusion is true
F T F F Not a counterexample because premise is false
F F F F Not a counterexample because premise is false
This truth table allows us to abstract away from the particular meanings of the sentences A and B
and just focus on the logical structure of the argument. A and B could be replaced with any full
sentences, such as ‘grass is green’, ‘2+2=4’, ‘The color purple smells funny’, ‘the moon is made
of green cheese’, etc. It doesn’t matter which, because the four rows in the truth table represent all
four possibilities: Top Row: A and B are both true sentences. 2nd Row: A is true and B is false. 3rd
Row: A is false and B is true. 4th Row: both claims are false. The claim ‘A & B’ is true only in the
first case (top row) and false in the rest. Thus this truth table tells us that regardless of what
sentences we replace A and B with, there is simply no way for the conclusion to be false if the
premise is true. You can do this with most arguments. Obviously, this inference is so trivial that
no one even really feels the need to make it explicitly.
Here is another type of valid argument called disjunctive syllogism:
Premise 1: Either A or B is true.
Premise 2: B is false.
Conclusion: Therefore, A is true. [1,2 / disjunctive syllogism]
Here is an example:
Premise 1: Either John or Mary must have taken the car to the store.
Premise 2: Mary did not take the car to the store (because she is in the living room).
Conclusion: Therefore, John must have taken the car to the store. [1,2 / DS]
An ‘or’ claim is only false when both of the claims it connects are false. (At least for inclusive
‘or’. Sometimes we use ‘or’ in a way that insinuates that it is one or the other, but not both. This
is a different logical concept.) So if you know one side is false and the or claim is true, then you
know that the other side must be true. (We can swap out A and B in premise 2 and the conclusion
and it won’t make a difference.)
Note that in these cases the premises are not backed by logic but by other evidence. Again,
logic just gets you from the premises to the conclusion. It doesn’t tell you whether the premises
are true. The following is also a valid argument:
Premise 1: Either 2+3=7 or 2+3=11.
Premise 2: It is not the case that 2+3=11.
Conclusion: Therefore, 2+3=7. [1,2 / DS]
Here the conclusion is obviously false. But the argument is still a form of disjunctive syllogism,
and it is still valid. The problem is that Premise 1 is false. It isn’t the case that either 2+3=7 or
2+3=11 because both of these claims are false. Thus the argument is valid but unsound.
Here is another valid inference rule, called modus ponens:
Premise 1: A.
Premise 2: If A then B.
Conclusion: Therefore, B. [1,2 / modus ponens]
Here is an example:
Premise 1: Ben is a human being.
Premise 2: If Ben is a human being, then Ben is a mammal.
Conclusion: Ben is a mammal. [1,2 / modus ponens]
In general, an ‘if … then’ statement—a conditional—is only false when the if-statement is true
and the then-statement is false. That is, when the antecedent condition holds and the consequent
statement is false. In other words, when the thing that the conditional says is supposed to follow
from the initial condition does not in fact follow because the first is true and the second is false.
The bottom two rows of the truth table here represent the case where the conditional statement is
vacuously true. While this seems counterintuitive, consider that the following statement is true:
If Santa Claus is real then he delivers presents to children on Christmas eve.
PHI 345 WRITING TIPS
Unfortunately, Santa Claus is not real. Thus the antecedent statement is false. As is the consequent.
This corresponds to the bottom row of the truth table. Conditionals can be vacuously true because
we need to reason hypothetically about would happen if things were different.
Another valid argument form is called modus tollens:
Premise 1: It is not the case that B.
Premise 2: If A then B.
Conclusion: It is not the case that A. [1,2 / modus tollens]
Another example:
Another example:
Premise 1: It is not the case that Snoopy is a cat.
Premise 2: If Snoopy is a cat, then Snoopy is a mammal.
Conclusion: It is not the case that Snoopy is mammal.
Here is a truth table illustrating why it’s invalid:
Do not make an argument of this form. This is called denying the antecedent, and it is a fallacy!
PHI 345 WRITING TIPS
Another example:
Premise 1: Snoopy is a mammal.
Premise 2: If Snoopy is a cat, then Snoopy is a mammal.
Conclusion: Snoopy is a cat.
Here is an example:
Premise 1: If Garfield is a cat then Garfield is a mammal.
Premise 2: If Garfield is a mammal then Garfield is warm blooded.
Conclusion: If Garfield is a cat then Garfield is warm blooded. [1,2 / HS]
Here is a truth table illustrating why it’s valid (since we have 3 basic claims here we need 8 rows
to represent all the possibilities):
In fact, the chain of conditionals can be as long as you like. The following is also valid:
Premise 1: If A then B.
Premise 2: If B then C.
Premise 3: If C then D.
Premise 4: If D then E.
Conclusion: If A then E.
I know that this might sound like a slippery slope fallacy, but the problem isn’t with the form of
the argument itself. Rather, slippery slope fallacies just make false assumptions. For example:
Premise 1: If you give a mouse a cookie, then it will survive another day.
Premise 2: If the mouse survives another day, it will breed.
Premise 3: If the mouse breeds, then all of its offspring will breed.
Premise 4: If all of the mouse’s offspring breed, then their offspring will breed until
there are millions of mice in your home!
Conclusion: If you give a mouse a cookie, there will be millions of mice in your home!
In terms of its logical form, this argument is valid. The reason why it is a bad argument is that the
premises are likely false. Even Premise 1 is doubtful: the mouse could be killed by a cat, snake, or
owl two minutes after you give it the cookie. But suppose Premise 1 just happens to be true and a
mouse happens to survive after being given a cookie. Premise 2 is also doubtful: the mouse could
die the next day before it could breed, or it could simply fail to find a mate. Premise 3 is even more
ridiculous: there’s a very good chance that many of the mouse’s offspring will die or fail to breed.
And Premise 4 is just obviously false. So the problem isn’t with the logical form of the argument,
but with its premises. Making the argument explicit like this allows me to point out what’s wrong
with it by attacking specific premises and coming up with specific counterexamples to them. In
general, a counterexample to a conditional claim has the following form:
‘If A then B’ is false when A is true and B is false.
Let’s take a the following conditional as an example:
If x is a part of y and y is a part of z, then x is a part of z.
This is known as the transitivity of parthood. It is a metaphysical principle that strikes many people
as being true. For instance, the cells in my toe are parts of my toe, and my toe is a part of me, so it
seems that the cells in my toe are also a part of me. But consider the following:
Justice Kennedy’s toe is part of Justice Kennedy and Justice Kennedy is part of the
Supreme Court, but Justice Kennedy’s toe is not part of the Supreme Court.
This sounds true. But if it is, it would be a counterexample to the transitivity of parthood. This is
interesting from a metaphysical perspective. How did I come up with this counterexample? Well,
I read the SEP article on mereology. But the philosopher who originally came up with it obviously
had to use their imagination. But you at least know what you’re looking for: a case where x is a
part of y and y is a part of z, but x isn’t a part of z. How do the defenders of the transitivity of
parthood respond to this example? Well, they have three options:
PHI 345 WRITING TIPS
Option 1: Deny that Kennedy’s toe is a part of Kennedy. [Not very plausible.]
Option 2: Deny that Kennedy is a part of SCOTUS.
Option 3: Insist that Kennedy’s toe is a part of SCOTUS.
You can combine these different arguments in complex ways. For example:
Premise 1: A
Premise 2: If A is true then so is B.
Premise 3: If B is true then so is C.
Premise 4: If C is true then so is D.
Conclusion: Therefore, D must be true.
[Basically, this just involves applying modus ponens and hypothetical syllogism over and over
again.] A final common kind of deductive argument is a dilemma (or trilemma, etc.):
Premise 1: One of the following must be true: A, B, C, …
Premise 2: If A is true then so is Z.
Premise 3: If B is true then so is Z.
Premise 4: If C is true then so is Z.
⋮
Conclusion: Therefore, Z must be true.
There are infinitely many types of deductively valid arguments, with far more complex structures
than this. But at your level, these are really the only sorts that I would recommend.
Let’s look at an example:
(Sorites 1) 1 grain of rice is not a heap. [Assumption]
(Sorites 2) If n grains of rice are not a heap, then n+1 grains are not a heap. [Assumption]
(Sorites 3) If 1 grain of rice is not a heap, then 2 grains of rice are not a heap.
[From (Sorites 2) / universal instantiation]
(Sorites 4) Therefore, 2 grains of rice are not a heap.
[From (Sorites 1) and (Sorites 3) / modus ponens]
(Sorites 5) If 2 grains of rice are not a heap, then 3 grains of rice are not a heap.
[From (Sorites 2) / universal instantiation]
⋮
(Sorites Con) Therefore, 1,000,000 grains of are not a heap.
This is called the Sorites Paradox. Pretty much everyone agrees that there must be some solution.
The bitter disputes arise over what the solution should be. There are five main options:
PHI 345 WRITING TIPS
Option 1: Deny (Sorites 1) and admit that a single grain of rice is a heap. This is not very intuitive,
to say the least. If we apply it to similar cases, it would mean that a single grain of sand
is a dune, a single H2O molecule is a tidal wave, etc.
Option 2: Deny (Sorites 2) and admit that there is some unknowable sharp cut-off point where
adding a single grain of rice to a pile of grains will suddenly make it a heap. The
problem with this option is that it seems bizarrely arbitrary.
Option 3: Accept the conclusion that there are no such things as heaps of rice. Basically, this is
like saying that normal objects can’t really exist because they would have to be vague
and reality isn’t vague. One way to cash this out is by saying that a ‘heap’ is just a term
we apply to some grains of rice for our convenience and doesn’t pick out a real kind of
object. But this reasoning would seem to also entail that there are no waves of water,
or dunes of sand, or bald men, or anything else with vague boundaries! In fact, this
might seem to entail that there are no people, cats, cars, buildings, tables, or chairs, as
all of these things seem to have vague boundaries as well.
Option 4: Claim that most language is vague, so there is no fact of the matter about what which
piles of rice the term ‘heap’ applies to. For many people, this seems to be the most
reasonable option. The problems mostly concern the semantics of vague language,
which are really too technical for us to get into in the course. Some people argue that
these semantics proposals fail, so we are forced to either adopt one of the other options
or admit that most language is meaningless.
Option 5: Reject classical logic by saying that there are some things that both are and are not
heaps of rice, or are such that it is indeterminate whether they are heaps of rice. Since
much of mathematics rests upon classical logic, this would cast much of physics and
engineering into doubt.)
A good paper could focus on cashing out and defending just one of these options. Or considering
one way of cashing them out and presenting further objections showing that it doesn’t really escape
the original problem.
Here is a way to reconstruct something like Bargle’s initial line of reasoning in the ‘Holes’ article:
P1: There is at least one hole in this piece of cheese. [NLA]
P2: There is at least one hole. [P1]
P3: Holes do not have mass. [NLA]
(i.e., For all x, if x is a hole, then x does not have mass.)
P4: The only physical objects without mass are photons and gluons. [NLA]
(i.e., For all x, if x is a physical thing and x does not have mass,
then x is either a photon or a gluon.)
P5: Holes are neither photons nor gluons. [NLA]
(i.e., For all x, if x is a hole, then it is neither a photon nor a gluon.)
P6: Holes are not physical objects. [P3-P5]
PHI 345 WRITING TIPS
Con: There is at least one thing that is not a physical object. [P2 and P6]
This one is a bit trickier because it uses quantified logic, but hopefully all of the intermediary steps
are obvious. If you want to deny the conclusion of the argument, you must deny one of the non-
logical assumptions (which I’ve labeled ‘NLA’). What you can’t do is deny P2 while saying that
P1 is true. How can there be holes in the piece of cheese if there are no holes?! Argle was basically
forced to admit this in the article, at which point he switched to denying P3 instead (by identifying
the holes with the thin layers within the cheese surrounds what we normally think of as the hole).
But another alternative is to deny P4. You could do this by saying that space is a physical thing
even though it doesn’t have any mass, and then identify holes with empty regions of space. In fact,
that is probably how we normally think of holes! There are good reasons to think that space is a
physical object all on its own: according to current physics it has a curvature that can change as
massive objects pass through it. The only other option is to deny P6. But that doesn’t seem like a
real option at all. Clearly, the holes in the piece of cheese aren’t subatomic particles!
One good way of writing a paper is to try to construct a valid argument like this for some
conclusion and then spend the rest of the paper examining each premise and how one could deny
or defend it. Not every argument you make in the paper needs to be put in this explicit premise-
conclusion format. Rather, the format is helpful because it forces you to make some main line of
reasoning explicit, and allows you to pick apart the different pieces of the reasoning. The point of
representing an argument in this sort of explicit format is that it breaks your work up so you can
focus on one issue (premise) at a time.
1.4. Thought Experiments
But a lot of the times it won’t be all that helpful to construct an explicit logically valid
argument in premise conclusion format like this. Another way to write a good paper is to try to
evaluate different claims with thought experiments. In general, a thought experiment is much like
a scientific experiment: you are trying to isolate and test a specific variable while controlling for
interfering variables. Don’t be afraid to come up with something outlandish to illustrate your point!
What’s important is that you try to think of how your opponent would try to explain the intuitions
about your example away. Try to eliminate confounding factors. For instance, in the case about
the supreme court above, the defender of the transitivity of parthood might claim that the SCOTUS
isn’t a material object and so it doesn’t have any parts, but rather members (Option 2). While
Kennedy is a member of SCOTUS, neither he nor his toe are a part of it, properly speaking. This
is a somewhat plausible reply. One could try to control for it by coming up with an example that
features only material entities. How about the following: a screw is part of a machine, the machine
is part of the factory, but the screw is not part of the factory. Does that work? Maybe not. So how
about the following: my gut bacteria are a part of my digestive system, my digestive system is a
part of me, but my gut bacteria are not a part of me. Does that work? Hmm… If you’re proposing
some general metaphysical principle, it is important to reflect on whether there are any
counterexamples!
PHI 345 WRITING TIPS
2. Citations
Whenever you attribute a specific claim or position to an author you must provide a citation
and preferably a page number if there is a specific place in which she makes the claim.
• Quotes require specific page numbers on which they occur.
• Quotes from internet sources should try to specify the section or paragraph number in
which the quote occurs.
• Quotes longer than two sentences should be block quotes. They should be separate from
the main text, single spaced, and indented half an inch from the margins on each side.
When describing what an author says in an article or book, use the present tense.
• GOOD: “Thomson argues that abortion is permissible in most cases.”
• BAD: “Thomson argued that abortion is permissible in most cases.”
• BAD: “Thomson has argued that abortion is permissible in most cases.”
• BAD: “Thomson was arguing that abortion is permissible in most cases.”
Your Works Cited should be in some standard format like APA or MLA (look them up). I don’t
care which so long as you consistently use the same format for each entry and it contains enough
information to look the source up.
Example:
Thomson summarizes the upshot of her thought experiment as follows: “a woman surely can
defend her life against the threat to it posed by the unborn child, even if doing so involves its
death” (2010, pg. 91).
Thomson, Judith Jarvis. 2010. “A Defense of Abortion”. In David Boonin and Graham Oddie
(eds.), What’s Wrong? Applied Ethicists and Their Critics, 2nd Ed. OUP: pp. 89-98.
3. Common Grammatical/Stylistic Errors
I can’t even begin to cover all the basic grammatical errors people make. The best rule for catching
such errors is to just read what you’ve written aloud. That said, here are some common errors:
• Make sure your pronouns are consistent.
o Either ‘in one’s power … then one is morally …’ or ‘in your power … then you
are morally …’. But don’t mix ‘one’ and ‘you’.
• Avoid the passive voice.
o Avoid talking about arguments, discussions, views, ideas, claims, concepts, etc.
when you can instead describe the author as arguing, discussing, claiming, etc.
▪ BAD: “It is a point made by Thomson that ….”
▪ GOOD: “Thomson makes the point that…”
PHI 345 WRITING TIPS
4. KEEP IT SPECIFIC!
Make what you are talking about as explicit and specific as possible.
• Avoid using ‘his’, ‘her’, ‘the’ when you’ve just discussed multiple cases or people unless
it is completely clear who and what you are talking about!
• Generally speaking, ‘makes a claim about’, ‘argues on’, ‘argues about’, etc. are vague
and uninformative. One can make a claim or argue about something by either agreeing or
disagreeing with it.
For instance, if you just described Thomson’s Violinist case and want to say that it is
permissible for the person to detach the violinist from herself you have to say this explicitly.
There are multiple people acting in the case. It is reasonable for the reader to infer that you
mean it is permissible for the person to detach the violinist and not for the people to abduct
the victim and attach the violinist to her. However, your writing is far too vague if it forces
the reader to charitably reconstruct what you’re saying in this way.
o BAD: “I believe it is permissible” or “this case is permissible”.
o GOOD: “I believe it is permissible for the person in Thomson’s case to detach
the violinist”.
o Better yet: use different names for the different people in the case so you can
easily refer to them.
Example: “The are at least three agents involved in Thomson’s violinist case: the Violinist;
the Victim to whom the violinist is attached; and the Abductor, who attached the violinist to
the Victim. It was clearly impermissible for the Abductor to abduct the victim and attach the
violinist to her without her consent. However, the relevant moral question is whether it
permissible for the Victim to detach the Violinist after the fact. …”
• BAD: “Nozick’s view is horrible. I grew up in a poor neighborhood, so I can tell you
from experience that there is nothing ‘just’ in how landlords acquire their holdings.”
o Here you are appealing purely to your experience. There is no argument to speak
of. You’re basically just stamping your feet and asserting that Nozick is wrong.
PHI 345 WRITING TIPS
▪ But note: the fact that this is a bad argument doesn’t mean you’re wrong
about landlords! Rather, the problem is that you haven’t explained why
landlords acquire their holdings in an unjust manner.
• BETTER: “Nozick’s view fails to apply to the actual world. For instance, one can hardly
argue that landlords are in the same position as Wilt Chamberlain, given that people need
a place to live and are therefore forced to acquiesce to whatever terms the landlords set.”