Here you can find hints for improving the quality of your argument analyses. These hints are based on feedback written for students' weekly problem sets.

1 Be maximally clear

Use few multi-syllable words; do not use needlessly long sentences; thoroughly proofread your maps before submitting them. Ensure that each claim box contains exactly one sentence, and that this sentence is capable of being true or false.

1.1 Be concise, but not at the expense of clarity

Avoid language that makes it difficult to understand a claim without referring to something outside the claim itself—for example, demonstratives (like “that” and “this”). In other words, write clearly and minimize your use of anaphora.

2 Exclude “logical language” from claims

Argument maps represent the logical relations denoted by words like “and,” “but,” “so,” “because," etc., visually. These words will rarely appear in a good map because, rather than describing an argument using English words, the best maps display the argument using color, line, and shape.

Using words like "and," "but," and "therefore," frustrates the main benefit a really good map provides: the ability to quickly see exactly how the various parts of a complex argument all hang together. For example, we represent “A, therefore B” and “A and B” with the small map-fragments close to this text.

2.1 Rarely include conjunctions in claim boxes

Consider this sentence: [The moon is made of green cheese (A) and the sun is 93 million miles away (B).] This sentence is a conjunction because it was formed by conjoining two independently complete sentences, A and B, with the word “and.”

Avoid using conjunctions in your maps. Instead, split the conjuncts up into separate white boxes which can be unified within a single reason (green bracket) or objection (red bracket). The reason for this rule is simple:

If you put a sentence like “B1 and B2” into a single white box, it will often be ambiguous whether you're supporting B1, B2, or both. The map near this text represents a single claim, A, supporting a conjunction of two claims, B1 and B2. If A only supported B1, and had no bearing on B2, we would have no way to represent this because B1 and B2 are smooshed together in a single claim! There are a few instances where this advice should not be applied (namely, when a proposition supports the conjunction as a whole but does not support its conjuncts taken individually), but it holds in the majority of cases.

3 Exclude background material

For example, there is no need to represent definitions or other stage-setting materials in your maps. All materials appearing in maps should be helpful for persuading someone who’s already familiar with the basics. When it’s helpful to include some background material, you may wish to use a note (hit “n” in MindMup). Your maps should strive to capture the most valuable nuggets from the passages you are assigned.

4 Use parallel language whenever possible

Using different words to mean the same thing makes analyzing and assessing arguments more difficult; so, use consistent language throughout your maps. Don't be afraid to repeat entire claims.

4.1 Eliminate danglers.

Danglers are claims that do not contribute anything to the strength of the argument. Here is a simple example:

If we deleted 2.6, this argument would not be weaker; so, 2.6 is a dangler and should be cut. Too, notice that 2.6 is not linked up via parallel language to either 2.4 or 2.5. For example, neither 2.4 nor 2.5 make any use of the concept of dreaming, which is central to 2.6. This is an excellent clue to cut 2.6.

5 Avoid “cheesy co-premises”

Cheesy co-premises are claims, often implicit, of the form “If my premises are true, then my conclusion is true.” While these arguments are all deductively valid, they’re usually a symptom of lazy thinking. So, be on the lookout! They will follow the same pattern as the map nearest this text.

Maps which reliably follow this pattern are (very repetitive) logic puzzles. This defeats the purpose of mapping, which is to think more lucidly about difficult arguments.

Philosophers sometimes begin journal articles by announcing a conclusion and listing premises which are supposed to jointly entail the conclusion. But once you get to the meat-and-potatoes of the articles—the support for the premises themselves—the presentation usually becomes less formal. Deductively valid reasoning is rare; instead, what you tend to find are considerations that in good cases favor the premises to various (often unclear) degrees. Assessing an argument involves figuring out (a) to what degree one believes those considerations and (b) to what degree they favor the premises. Cheesy co-premises do not help in either of these tasks, and their rigorous-looking veneer can make it more difficult to detect shoddy or absent reasoning.

So when you catch yourself relying on a cheesy co-premise, always search for a more thoughtful implicit premise by searching for the considerations that might support the cheesy co-premise (albeit, non-deductively).

Often the best way to revise a cheesy co-premise is to ask how you would support that statement. (In the map near this text, the cheesy co-premise is 2.2: “If the tickets for a ski trip will cost much less than the tickets for a beach trip, then a ski trip will be less expensive than a beach trip"). Then, once you have come up with some kind of support, just replace the cheesy co-premise with that.

To see this hint applied, consider the previous example. In that map, 2.2 is a classic cheesy co-premise. To revise it, begin by asking yourself why someone might find 2.2 credible…

Maybe they believe that the cost of travel makes up the main difference in expenses between ski trips and beach trips. This is much more thoughtful than 2.2, so in the map nearest this text I've revised the implicit co-premise to what 2.2 should have read all along.

6 Never conjure concepts out of thin air

Make sure that none of your claims "conjure up" their central concepts. That is, each of the important concepts in a claim should appear in at least one reason or objection immediately beneath that claim. (One obvious exception to this rule is the bottom layer—every argument has to end somewhere!)

The map about aging nearby this text is guilty of "conceptual conjuring” because the conclusion is centrally concerned with the idea of public resources, which does not appear in the reason beneath the conclusion. The conclusion is simply conjured out of thin air. It is impossible to conclude anything about what resources society should devote to X without any premises that mention public resources! But that is exactly what this map attempts to do.

Watch out for this common and easy-to-avoid error!

7 Two claims should be placed into a single reason (green bracket) if and only if they support some conclusion more strongly together than they do individually

Here is a simple example:

(C1) Only 1 in 100 million people carry genetic marker A,

(C2) The person who killed Dylan carried genetic marker A,

(C3) Whitney carries genetic marker A.

Together, C1--C3 jointly support the conclusion far more strongly than any subset of the three. That’s why I’ve placed them all in a single green reason in the map above. Here’s one of the wrong ways of mapping this passage:

A good test for whether two claims, C1 and C2, provide support to a conclusion independently or jointly is to ask yourself, "Would C1 continue to support the conclusion (as strongly) were C2 false?" (and vice versa). If the answer is no, then C1 and C2 should be unified in a single reason (or objection). For reasons that have more than two claims, apply this rule to all possible subsets.

Double-check your work for this mistake—it is one of the most common.

7.1 Two claims, C1 and C2, should be placed in separate reasons (or objections)

if and only if C1 would continue to support (or oppose) the conclusion just as strongly were C2 false, and vice versa.

8 Start wherever you understand the argument best

Many students tend to map papers as they read through them for the first time. This is rarely a good strategy. A much better alternative is to begin by skimming the paper. When you hit a passage that you feel you understand, stop reading and get that passage mapped out perfectly in MindMup. Then, use your understanding of the small passage to bootstrap yourself to a better understanding of the surrounding passages, and eventually, the rest of the argument. Getting as clear as possible about the bit of the argument you find easiest and working outwards from there will often be the most efficient way to analyze a difficult philosophical text.

9 Take what's given freely

Consider the following passage:

``Do you think your local butcher will reduce the amount of meat she orders from the slaughterhouse if you become a vegetarian? [1] Of course not! [2] The supply chain for meat just isn’t sensitive to the quantities that a single person consumes. So, [3] by becoming a vegetarian, you'll never save a single animal’s life! But [4] by becoming a vegetarian you’ll deprive yourself of the pleasure of eating meat, and [5] you should only do that if the benefits would outweigh the losses. So, [6] you should only become a vegetarian if doing so would save some animals’ lives.``

It seems simple enough, but before reading any further down this page, try to sketch out a map of this argument in MindMup using this link (if prompted, select "Open with MindMup 2.0"). You can learn how to use MindMup by watching the short walk-through video.

If you attempted the exercise, you now know that it's far from easy to map this short passage. Beginning mappers at Princeton regularly struggle for an hour (or more) to figure out the structure of this deceptively simple argument. And they still don't always get it right: comprehending arguments is really that hard. But if you make the most of little blessings—words like “so” and “and”—you can figure it out quickly.

Here’s how. Look up at the passage again and notice the word “so” that comes just before Sentence [3]. That "so" should jump out at you and slap you in the face! It should cause you to draw something like the map beneath this text.

You should get warm, fuzzy feels in spades when you compare this map-fragment to the text it represents.

At this point, we don’t know how this chunk connects up with the rest of the passage, but we can be confident we’ve got this much of it right. So, let’s continue looking for little blessings we can connect up with this fragment.

Here's the next bit of the passage:

``But [4] by becoming a vegetarian you’ll deprive yourself of the pleasure of eating meat, and [5] you should only do that if the benefits would outweigh the losses. So, [6] you should only become a vegetarian if doing so would save some animals’ lives...``

The “and” linking [4] and [5], and the “so” that comes right before [6], suggest something like the three-claim map near this text. You’ll also notice that you can combine this argument unit with the one above to support the claim "You shouldn’t become a vegetarian." Putting them together nearly gets us all the way to a complete map!

Now that you've seen how we can get most of this difficult passage mapped for free just by noticing words like “and,” “but,” and “so," finish this map off by thinking about what’s assumed in the inferences supporting [6] and [3]. Then, compare your thoughts with the map that ends this document.

10 Be nice

When interpreting a text, begin by assuming that the author is smart and has thought carefully about the topic. This assumption should lead you to search for a representation of the author's argument that does not commit obvious blunders or rely on unsubtle falsehoods. As Kahneman and Tversky wrote, "a refutation of a caricature can be no more than a caricature of refutation.”

Yet, unfortunately, authors do sometimes assume fishy-smelling claims. So, just because a claim seems false to you, you shouldn't conclude that it has no place in a map of someone else’s argument! Remember that people can sometimes be blind to assumptions they themselves find incredible.

11 Expect that you’re going to revise. A lot

Mapping is hard, but not because the rules on this page are hard—they're fairly straightforward to understand and apply. Mapping is hard because reading argumentative texts is hard. People tend not to notice this fact when they aren't forcing themselves to rigorously analyze the text using a method that can reveal their biases and blind spots.

So, when you hit a difficult map, remember that you’ll rely on the skills you develop here to succeed in many other classes. They are all-purpose tools that will serve you well not just in your Princeton classes, but also in the political, professional, and civic reasoning you employ for the rest of your life.

Use this list of hints as a checklist when revising your assignments

Get into the habit of checking every argument unit for clarity, charity, concision, logical vocabulary, correctly individuated reasons and claims, and so on. With practice, students usually find themselves checking for these things automatically and their work improves dramatically. It’s important to develop good habits early on, so practice using these hints. As you learn to map arguments, practice applying these hints to every assessable chunk of your homework assignments!

Bonus map

Here’s a complete map of the anti-vegetarianism argument. Can you see anything wrong with it?