This brief writing guide is designed to help you develop your ideas into a strong, well-written paper. I’ve tried to make the advice as reader friendly and accessible as possible. I’ve also tried to make the sections as independent as I could, so you shouldn’t lose much by picking and choosing sections to read. In this writing guide, I provide helpful tips for writing papers in ethics, although the advice I apply here is readily transferable to other sub-disciplines of philosophy.
On this page, I offer short pieces of advice for writing an argumentative paper, including how to write a thesis well, how to structure your essay, and how to construct a strong argument. You’ll notice that these categories correspond to most grading rubrics for argumentative philosophy papers and exams.
If you want more detailed guidance, some sections have more in depth explanations. You can find the link to the longer versions of the guide at the end of the sections.
This is very much still a work in progress, so take my advice with a grain of salt. Always feel free to reach out if have some suggestions for improvement.
Goal of the Guide
Thesis
-
The first thing you’ll want to do in your paper is to write a strong thesis statement. Put simply, a thesis statement is two to four sentences that tell your reader what you will argue for and how you will argue for it. Since the thesis statement is supposed to make sure the goal of the paper is clear, you need to be very explicit about what your conclusion is and how you’ll reach it.
A strong thesis, then, needs to have three things:
A specific conclusion: tell your reader exactly what it is you’re arguing for. Be specific here. A thesis for an argumentative essay shouldn’t say things like “Such and such practice poses an ethical problem.”
A reason to accept that conclusion: give your reader the main reason you will discuss to accept that conclusion. When writing an ethics paper, you can often do this well by presenting a general moral principle that you’ll defend in more detail throughout the paper (see “Moral Principles” below).
Some initial details to support the argument: give your reader the main detail about the big picture reason you have for accepting the conclusion. In other words, provide some initial support for the big picture idea you’re sketching. For instance, you might give some brief reason to support the moral principle that forms the backbone of your argument.
This is just a big picture overview of what a strong thesis statement should look like. Click here for a more detailed development of these ideas.
-
This guide is meant to serve as a primer on how to write an argumentative essay well. You should generally have an idea about how you will structure the paper. By a paper’s “structure” I mean the order in which you present your ideas. You can structure a paper in a variety of different ways. Here, I outline just a couple.
There are at least two broad approaches you could take to structuring a paper or essay response. First, you might take an offensive approach – where I mean “offensive” in the sense of an attack, not in the sense of offending someone. You might take this approach if you encounter a thesis or argument that you think is seriously mistaken. If your goal is to attack a thesis or argument you disagree with, the following is a pretty solid format for a big picture flow of ideas:
your thesis
key terms/assumptions
thesis/argument you are attacking
your main argument against that thesis/argument
an objection to your main argument, and
your response to that objection.
(If you’re preparing to answer an argumentative question on an exam, don’t be afraid to ask your professor if (5) and (6) are required to score well.)
The second broad approach is defensive. On this approach, the goal in your essay is to defend a thesis or argument against some objection. You might take this general approach if you found a serious objection to a thesis/argument you endorse, or even if it seems that many people around you think that objection is a serious problem. In this case, you may want to structure your paper as follows:
your thesis
key terms/assumptions
a summary of the position/argument you want to defend
an objection to that position/argument, and
your response to that objection.
an objection to your main argument, and
your response to that objection.
(Same comment as above: If you’re preparing to answer an argumentative question on an exam, don’t be afraid to ask your professor if (6) and (7) are required to score well.)
These are just two ways to organize your papers, and there are certainly others. For the rest of this guide, I’ll briefly provide some tips for how to write each of these sections of the paper.
-
Your essay needs to be focused. That means everything you write – every section, paragraph, and sentence – must support the thesis (see also Irrelevant Details). But determining whether some detail supports the thesis is just the first step; next, you need to make sure that it’s clear how what you’ve said supports the thesis. Even if some section of the paper supports the thesis, the essay can still be disorganized if you don’t explain why what you’re writing supports the thesis. Here, I’ll outline two tips on how to keep your essay focused.
First, you’ll want to state explicitly how that sentence and the sentences that follow are relevant to supporting the thesis. It’s important to make these explicit statements in the right place. Usually, you’ll want to say explicitly how the following sentences are relevant to the thesis at the very beginning of the section/paragraph you’re writing. Be as explicit as you can. You can even be as direct as saying something like “In this section/paragraph, I’ll argue that this thesis is true by…”. For every section/paragraph/sentence you write, ask yourself: “If one of my peers read this paper, would it be obvious how this section/paragraph/sentence supports the thesis?” If the answer is no, then you should do your best to clarify how that section is related to and supports the thesis. You should also occasionally summarize what you’ve said in some section/paragraph at the end of that section/paragraph. This will help your readers stay focused on the big picture without getting lost in the details.
Second, you’ll want to say explicitly how this section/paragraph – the one you’re just now writing – is related to the previous section/paragraph. Explaining very briefly at the beginning or end of a section/paragraph how it relates to everything else you’ve just said is a clear way to link your train of thought – like a literal train. So, for instance, at the start of the paragraph following a section that clarifies your definitions/assumptions, you might say something like “in the previous section, I aimed to clarify my thesis that [insert brief version of thesis] by defining [insert key term(s)]. Having clarified my main claims, I’ll now argue…”. Don’t be afraid to tell your reader exactly what you’re doing.
Organization
-
Having a strong thesis is only the beginning of writing an excellent argumentative essay. Now you’ll need to fill in the details of the argument itself. Generally, in your thesis, you will have used terms or phrases that roughly correspond to one of the following three categories:
(a) terms that most people would be familiar with and there is no (relevant) technical way in which prominent/relevant authors have used the term,
(b) terms that most people would be familiar with but there is a technical way in which prominent/relevant authors use the term, and
(c) technical terms that most people are not familiar with.
You should always define key terms in categories (b) and (c) as they become relevant in the paper. More often than not, this will be at the very beginning of the paper (but perhaps not always). This can be done in a few different ways, for instance, by briefly defining a term early in the paper in terms that are easy to understand before defining it more carefully in a section very early on in the paper. For more detailed advice, click here.
-
In an argumentative essay, it is crucial to outline any crucial background assumptions that you will be holding/using throughout the paper. Explicitly stating relevant assumptions will both clarify the thesis and narrow the scope of your paper. Your thesis should tell you what background assumptions you need to make; not always explicitly, of course, but you can generally see what might be plausible and helpful to assume for the sake of the argument once you’ve written the thesis.
There isn’t any hard and fast rule about what makes an assumption implausible, but there are some ways of narrowing down what assumptions you need to include. When determining what assumptions to include in the paper, ask yourself the following questions about the assumption/claim you’re looking to include in the paper:
(1) Is the assumption/claim more implausible than the conclusion?
(2) Will most or all of my opponents automatically reject this assumption/claim?
(3) Will my argument work if this assumption/claim is false?
If your answer to all three questions is “no,” then you have a good candidate for an assumption that is both plausible and helpful for the paper. I should emphasize that this is merely a rule of thumb; it is not a hard and fast formula. I work through this rule of thumb in greater detail here.
-
Many arguments in ethics rely on general principles. Moral principles are, roughly, claims about how we ought, morally, to live. Moral principles usually have two components, what I’ll call the type of moral principle it is and the scope of the principle. There are a variety of types of moral claim you might be dealing with. The most common types of moral principles are principles of:
(1) Obligation (it is morally obligatory to act or refrain from acting in some way just in case one has strong moral reasons to act or refrain from acting in that way and those reasons are not easily outweighed by other moral reasons).
(2) Wrongness (it is morally wrong to act in some way just in case not acting in that way is morally obligatory), and
(3) Permissibility (it is morally permissible for one to act in some way just in case that action is not morally wrong).
So, a moral principle has to tell you if an action is morally obligatory, wrong, or permissible.
Moral principles can be narrower or wider in scope. You can think of a principle as narrower or wider in scope, roughly, depending on how many instances of that kind of action the principle classifies as morally obligatory/wrong/permissible. The moral principle “it is morally wrong to kill humans” is quite wide: it classifies a lot of cases of killing as wrong. A wider principle governing the ethics of killing might be “it is morally wrong to kill animals.”
Arguments often rely on moral principles, so learning how to develop them is important. Here, I sketch some suggestions for developing plausible moral principles.
-
As a reliable general rule of thumb, you should always use examples to illustrate key points in your paper. Using examples will help the reader understand what you have in mind, even if you think what you’ve said is already pretty clear. Examples in philosophy essays serve mainly to clarify and support key terms or claims.
You might use an example, first, to clarify a key term. Suppose you’re writing a paper on clinicians who refuse to provide some service to a patient and you’ve just defined the key term “conscientious objection.” To conscientiously object to some practice is to refuse to provide that service based on sincerely held religious and moral beliefs. To clarify the definition, you might follow it up with an example: “for instance, a clinician conscientiously objects to providing an abortion when, due to her Christian faith, she believes it is immoral.” Some examples clarifying key terms can be included in parentheses or brief asides: “(e.g., refusing to provide an abortion because of Christian beliefs).”
You could also use an example to help support an assumption or specific claim in an argument. You probably already do this on a regular basis. If you said something like “Joe is not living a very healthy lifestyle,” you might naturally follow it up with something like “just this past week, he’s eaten two pizzas every day!” Eating two pizzas every day for a week is an example that you’re using to support your main claim, which is that Joe isn’t living a healthy lifestyle.
You need to remember to do this same basic exercise for the main claims you’re making in an argument. If you write something like “abortion is very harmful for women” but then never mention any cases of a harmful abortion (after carefully defining what you mean by “harm”! See section on defining key terms), your argument won’t be persuasive. You need to provide support for your argumentative claims, and this often comes from using examples that give you reason to think the main claim – in this case, the claim that “abortion is harmful” – is true.
-
A strong essay needs to include sufficient detail (see rest of the detail section), but there is such a thing as too much detail. You need to be selective about what you include in your essay. A very important piece of advice is that you only include information that supports the thesis. If something you’ve written doesn’t support the thesis, cut it.
You can tell if something supports the thesis by asking two important questions: first, is your argument weaker if you remove a particular section/paragraph/sentence? Whether the strength of your argument is changed at all by cutting that detail will depend, of course, on what type of argument you’re presenting (see Argument - Types). But the following is true regardless of whatever kind of argument you’re presenting: if taking out that detail makes the argument weaker, then it’s probably important to keep; if it doesn’t affect the strength of the argument, you can probably cut it. It might support the thesis by providing some detail that supports the premise of your main argument. Or a sentence you write might help clarify an assumption that you’re making that’s relevant to the argument. But if the answer is no, then you can cut that section/paragraph/sentence. This is worth emphasizing: cut everything that does not affect the quality of your argument.
To determine whether something you’ve written supports the thesis, you can ask, second: is your main argument or thesis more or less clear if you remove a particular section/paragraph/sentence? If removing some detail makes something else you’ve written less clear, then it’s well worth keeping. But try removing that section/paragraph/sentence. Can you understand the argument or thesis just as well without it? If you can, then it’s probably good to cut it. In general: be strict about what you include in the essay, cutting everything that doesn’t help strengthen or clarify your main argument or thesis.
Detail
Arguments - General Advice
-
When engaging with alternative positions – whether you’re presenting an argument against someone else’s view or thinking of how someone might argue against a view you hold yourself (see Big Picture Structure) – it’s important that you interpret and represent their views charitably. Here, I’ll briefly describe how you can begin to do this well.
Let’s start with interpreting charitably. There are some theses that might be ambiguous. In other words, there might be multiple ways that you can understand what someone has said. Here’s an everyday example. Suppose your friend Joe tells you that he needs to do some shopping and that he shops at Aldi and Walmart. He follows it up with “I’m going to the store now.” In this case, he could mean he’s going to Aldi, Walmart, or both. We don’t know, and, without more detail, we could interpret what he said either way.
Ambiguity can crop up in philosophical arguments, too. Here’s an example. Suppose someone presents you with the following pro-life argument: “the fetus is alive, so it’s wrong to kill it.” Now, there are at least two ways of interpreting what this person mightmean. This person might mean that it’s wrong to kill everything that is alive; or they might mean that it’s wrong to kill living human persons. The first interpretation isn’t a charitable interpretation because that’s probably not what they mean. For one thing, most pro-lifers think it’s morally permissible to kill at least some things that are alive (e.g., animals, bugs, cells, plants, etc.). And this is certainly an implausible version of the pro-life view more generally. So, this person is probably making the second claim: that it’s wrong to kill living human persons.
How do we know that’s what they mean? Most tellingly, it’s just the most plausible version of the view; it’s the view that many people already find obviously true, namely, that it’s wrong to kill persons. Of course, it’s debatable whether fetuses are in fact human persons; that’s beside the point. The main point here is that, regardless of what view you’re arguing against, there may be more or less plausible interpretations of specific claims. Always summarize and discuss the most plausible of those versions.
Even if you present the most plausible version of the view you’re attacking, it’s still possible to fail to faithfully represent that view. For instance, I’ve seen students, while arguing against the permissibility of abortion, merely mention in passing bodily rights. An argument in favor of abortion might best be put in terms of bodily autonomy, so, in the case I’m imagining, the student hasn’t picked an implausible interpretation. But this person, by only very briefly mentioning the main point, also hasn’t given a helpful and detailed explanation of the view – for instance, by ignoring why someone might think bodily rights are of central moral importance. This is one way you can charitably interpret a view without representing it well. To avoid this mistake, it’s best to put yourself in the other person’s shoes, so to speak. If you think abortion is impermissible, how wouldyoudefend the opposite position? Or, to put the same rule of thumb slightly differently, you’d want others to present all the important details of your view before they start providing some objections. This golden rule of arguments will help you practice charity in both your interpretation and presentation of views you disagree with.
-
Good philosophical arguments are designed to convince you of their conclusion. In this section, I’ll provide some helpful advice on how to make your arguments more persuasive.
The strength of an argument – at least, in terms of how likely it is to convince someone of the conclusion – depends partly on who you’re presenting the argument to. Why? Because when you present an argument to someone, they have a range of commitments that might influence their position. If you’re presenting a pro-choice argument to a pro-life person, for example, that person is not only pro-life but also committed to the belief that consciousness is not what grounds the moral value of humans. So, you should hardly be surprised if you tell that person “well, fetuses aren’t conscious early in development, so they don’t have moral value.” Why? Because they reject the assumption that consciousness is the only thing that grounds value, so they won’t find the argument convincing. The conversation would look very different if you were talking to someone agnostic about the issue; that person may not have this prior commitment, and so may be persuaded by the argument. The main point is that whether your argument is convincing at least in part depends on who you’re talking to. Big picture takeaway: your argument won’t be at all convincing if you just assume (and without further argument) that your opponent’s main claims are false.
So, how do we make more convincing arguments? In the example I gave, the pro-lifer wasn’t convinced because she rejected a key assumption/premise of the argument. That gives us a clue as to what might make it more convincing: you’d have to find a way to make that premise/assumption itself more convincing. To do that, you’d present a further argument for the crucial assumption/premise – in this case, an argument for the claim that something without consciousness doesn’t have moral value (for constructing these arguments, see Arguments – Types).
But what happens if you present that further argument and your interlocutor still rejects one of the premises? The strongest argument in this case will appeal to claims that your interlocutor already accepts. If you ground an argument on premises your interlocutor already accepts, then she can’t merely deny your support (not, at least, without being inconsistent). Thomson’s argument in defense of the permissibility of abortion is an excellent example of this approach. Her argument is roughly as follows: “let’s assume that the fetus is a person with the same rights as you and I; even still, if it depends on my body for life-support, I can permissibly remove that support. Abortion is just like removing yourself from a person who is attached to you in order to survive.” There’s a lot more to the argument, but the details don’t matter here. What matters is the kind of argument Thomson is making. She’s saying “Ok, let’s grant your claim that the fetus is a person and move from there.” Of course, she doesn’t believe that claim herself. But she’s making her argument much more persuasive by granting her interlocutor’s major assumption/premise. Here’s the take away: if you can argue that some conclusion is true even if someone accepts your opponent’s major assumptions/premises, you’ve stumbled upon a very strong argument.
Arguments - Types
-
One of the most well-known forms of argument is called modus ponens. This type of argument has two features.
The first feature is what we can think of as a linking premise called a conditional claim. Conditional claims link two independent sentences with an “if…then…” connector. Here’s an example: if the moon is made of cheese, then the moon is edible. This conditional claim links the independent sentences “the moon is made of cheese” and “the moon is edible” with an “if…then…” connector.
The second feature confirms that the first part of the “if…then…” clause is true. This first part – in our present example, the claim that “the moon is made of cheese” – is referred to as the antecedent of the conditional. The antecedent is just everything you find between the “if” and the “then” of the conditional premise. The second premise in a modus ponens argument just restates this antecedent. So, if the first premise of our argument is “if the moon is made of cheese, the moon is edible,” then the second premise will be “the moon is made of cheese.”
And that’s it! We now have the two premises of our argument. We just need to state the conclusion. The conclusion is just the second half of our conditional claim, which is known as the consequent – in our present example, it’s “the moon is edible.” So, our full argument looks like this:
(1) If the moon is made of cheese, then the moon is edible.
(2) The moon is made of cheese.
(3) Therefore, the moon is edible.
That’s how you set up the argument. But, in this case, the conclusion is obviously false. The moon is not edible (or, at least, not like cheese!). If the conclusion is false, it must be because either (1) or (2) is false. In this case, (1) seems true. If the moon were made of cheese, then we’d certainly be able to eat it. We can eat cheese, after all. The problem is with the second premise. The moon is in fact not made out of cheese (as far as I’m aware). So, in this case, (1) is true but (2) is false, and that’s what gave us the false conclusion.
There is a second way for a modus ponens argument to fail: the linking premise might be false. Here’s an example where that is the case:
(4) If bears eat meat, then they never eat honey.
(5) Bears eat meat.
(6) Therefore, bears never eat honey.
In this case, the argument again fails. This time, however, it fails because the conditional claim is false. How do we know? Because we can easily think of cases in which bears eat meat and eat honey. So, to figure out if a conditional claim is true, ask yourself two questions: is the antecedent true? And is the consequent false? If the answer to both is “yes,” then your conditional statement is false. We see that in the present example. The claim “bears eat meat” is true, and the claim “bears never eat honey” is false.
To summarize, the general form of a modus ponens argument is as follows:
(1) If p, then q.
(2) p.
(3) Therefore, q.
where “p” and “q” here just represent sentences. The argument fails if either (1) or (2) is false.
-
A very common kind of philosophical argument in everyday conversations run something like this: “you claim x. But, if x is true, then so is something else, y. But y is obviously false! So, there’s no way you can believe x.” Here’s an example: “you just said that all coffee is good. But, if all coffee is good, then so is Folgers! But Folgers is not good coffee. So, you’re wrong that all coffee is good” (for those of you not familiar with coffee, Folgers isn’t really that good). Notice what’s going on is this train of thought: your friend makes some claim but perhaps doesn’t realize what that claim entails. You respond by pointing out to her that, if she’s right, then something crazy has to be true. But we shouldn’t accept that crazy claim, in which case we should reject the claim in question. Hence the name of this type of argument, reductio ad absurdum, which means “reduction to absurdity.”
This type of argument is very common in philosophy too. I’ll just sketch one quick example used in moral philosophy against a view called moral relativism. According to (one type of) moral relativism, whether some action is morally wrong depends on the culture in which that act is performed. So, for instance, a relativist might claim that eating meat is morally wrong in primarily vegetarian cultures. Why? Because some action – in this case, meat eating – is morally wrong if the culture claims it is wrong. Now, here’s a common reductio ad absurdum against this view: if you think that the moral wrongness of an action depends on the culture in which it is performed, then it seems like you’d have to also believe some crazy things. For instance, it seems like you’d also have to believe that, in cultures where marital rape isn’t viewed as morally wrong, it is not morally wrong in that cultural setting. But, of course, the wrongness of marital rape doesn’t depend on how the prevailing culture views it. Therefore, moral relativism is wrong.
We can extract a general form of argument from this:
(1) If p, then q.
(2) Not q (because q is absurd!).
(3) Therefore, not p.
This type of argument fails if either (1) or (2) is false. (1) is false when the alleged connection doesn’t hold. I’ve covered elsewhere how to determine if (1) – which is called a conditional claim – is true (see Simple Deduction – Modus Ponens). The argument can also fail if (2) is false. The consequence might not actually be absurd. Of course, part of whether the consequence of a particular view is absurd depends on who you’re presenting the argument to (see Convincing Arguments). It’s very possible that someone just doesn’t find a particular implication absurd. Going back to the coffee example, someone might just deny that Folgers is a bad coffee (but they’d be wrong, in my professional opinion). If they do, then they won’t find the argument convincing. Try pointing out a crazier implication, one that your interlocutor already rejects.
-
Not all arguments guarantee that the conclusion is true; some just make it more likely that it’s true than false. More specifically, if a certain view can explain a wide range of data, then we have reason to think that view is true. We reason this way all the time. If I see smoke off in the distance, I have some reason to think that there’s a forest fire. The fact that there’s smoke makes me more confident that my thesis is true, namely that there’s a forest fire. And we’ll become more confident in our thesis the more evidence it can explain. So, for instance, if we know that it’s about the time of year when forest fires tend to spread, that’s more evidence in favor of our thesis that there’s a forest fire. If we read a news article that confirms there are forest fires nearby, then that too increases our confidence there’s a forest fire. Notice, importantly, that none of this guarantees there’s a forest fire nearby; it may be the case that a nearby abandoned building has caused the smoke. But, on balance, the evidence favors the thesis that there’s a forest fire.
Here’s the general form of the argument:
(1) There are pieces of evidence x, y, and z.
(2) Theory 1 explains x, y, and z better than Theory 2.
(3) Therefore, Theory 1 is more likely to be true than Theory 2.
In the present example, the theory that there’s a forest fire better explains the available evidence than the claim that it’s a nearby building fire, so it’s more likely to be true.
The same sort of reasoning is used in philosophy all the time. To illustrate the idea, consider a famous argument against abortion by Don Marquis. Marquis starts his argument with a claim that seems self-evident: killing people is wrong. This widely held moral claim serves as a piece of data for our moral inquiry. How do we explain this fact? According to one theory that Marquis rejects, it’s wrong to kill people because they have the capacity for consciousness. That would certainly explain why, in most cases, it’s wrong to kill people. But, as Marquis points out, that theory fails to explain why it’s wrong to kill people who presently lack the capacity for consciousness – for instance, those in temporary comas.
Marquis poses a different hypothesis: it’s wrong to kill people because doing so takes away their valuable futures. Not only does this explain why it’s wrong to kill you and I (at least, assuming we have valuable futures!), but it also explains why it’s wrong to kill people who are temporarily comatose. Even though they’re not conscious, they too have valuable futures. So now we have two pieces of data (i.e., self-evident claims that we endorse prior to thinking about our theory) that his view explains: it’s wrong to kill people and the temporarily comatose. And, he argues, his view explains these facts better than the main alternative, the consciousness theory – the latter can’t explain why it’s wrong to kill those in temporary comas. Since his view best explains these facts, we can be more confident, he argues, that abortion is wrong because it takes away a fetus’ valuable future.
Whether or not the argument is successful, it provides us with general lessons for making this type of argument. First, your proposal must explain the relevant data. In Marquis’ case, he had to explain what we already know about the wrongness of killing. The relevant data will depend on what your thesis is about. Second, it will depend on whether people endorse that data. If Marquis was arguing against someone who thinks killing innocent people is permissible, his argument would not be at all convincing to that person. Make sure the data you’re using in the argument is also considered reasonable by your interlocutors (see Convincing Arguments). Finally, you’ll need to show how your thesis better explains the data than the best alternative. If someone can explain why killing people and the temporarily comatose is wrong without claiming that abortion is wrong, then Marquis’ argument becomes much less convincing; in that case, his explanation isn’t the only game in town.
-
At some point in your life, you’ve probably heard someone make an argument using a comparison as a central feature of the argument. Plato makes exactly this type of argument, and it goes something like this: “if you just live your life seeking pleasure, you’re no better off than someone who is constantly working all the time. You just can’t have peace that way! In the same way, a life seeking pleasure can’t be a good life because it lacks peace.” This is what we call an argument from analogy. These arguments are simple: “these two things share some features; so, it’s plausible to think that they share this further feature, too.” In Plato’s argument, he’s claiming that constant pleasure seeking and working share some important features - most importantly, the stressful toll of constant pursuit of a goal. And we have one other crucial piece of information: working constantly, Plato suggests, cannot possibly result in sustained peace. So, if seeking pleasure is like that, then we know it likely doesn’t bring peace either.
Arguments from analogy share the same general structure:
(1) A and B are relevantly similar because they share two (or three, four, and so on) crucial features, x and y.
(2) A also has a further feature, z.
(3) Therefore, B likely has feature z.
This big picture structure makes it clear how arguments from analogy can fail. First, someone might dispute the fact that two things are relevantly similar in the way that someone claims. Consider Plato’s argument from analogy against the constant pursuit of pleasure. Is it really the case that constant pleasure seeking is stressful in the same way that constant working is? Perhaps, but someone might dispute that claim. It may be the case that constantly seeking pleasure isn’t stressful. So, the first way the argument can fail is this: the two things we’re comparing (e.g., pleasure seeking and work) may not be relevantly similar because they don’t exhibit the same features. In that case, premise (1) would be false.
Premise (1) can be false for a second reason. Even if two things share some important features, they might fail to be relevantly similar if they are dissimilar in other important ways. For instance, even if we grant that Plato is right to point out that constant pleasure seeking is nearly just as stressful as constant work, pleasure seeking might be different from work in other ways. It might be the case that, unlike work, constant pleasure-seeking results in much more pleasure! Constant work (at least usually) doesn’t have that effect. If that’s right, then Plato’s wrong to think that pleasure seeking is relevantly similar to work, even if they share one important feature
Finally, an argument from analogy can fail if premise (2) is false. It might not be the case that something has a further feature. Applying this possible objection to Plato’s argument, someone might claim that constantly working is not extremely stressful. That seems doubtful to me, but I mention it as one way someone might object to an argument from analogy. If that’s right, then the conclusion doesn’t follow.