Types of Assumptions in GMAT Critical Reasoning: Necessary and Sufficient

Let’s start with a quick definition of what an “Assumption” is in GMAT Critical Reasoning.

gmat critical reasoning thinker

If you’re from Midwestern America, like me, I’m sure you know that “to assume makes an ass out of U and ME.”

You could say that on some level the GMAT feels the same way. Assumptions suck, unless they’re sanctioned by the testwriters. 

After all, it is technically an Assumption that we both speak English–that you know how to read it and that I know how to write it. Now perhaps this article is just the chance product of 20,000 monkeys at 20,000 typewriters tapping away for 20,000 years. 

There’s no sense behind it; it is merely a product of chance. Perhaps you’re scanning the words and making stuff up as you go along that you just assume–yes–corresponds to the words on the screen. Don’t scoff at me–one does not know what one does not know.

These assumptions would be a bit outlandish, and probably not something that would come up on GMAT Critical Reasoning. That is, the GMAT would think it’s obvious that a person taking the GMAT speaks English well enough to, well, take the test. 

The GMAT would think that it’s obvious that a person who wrote for the test–which is given in English–would speak English and not actually be a football stadium full of monkey typists.

In other words, these wacky, hyper-skeptical Assumptions are “reasonable by GMAT standards” and therefore trivial: they can be ignored. It’s not helpful to be paranoid about absolutely everything–let me tell you from experience.

Rather, what the GMAT dislikes is Assumption that overtly makes an argument incomplete. Put another way, some of the link between the Facts and Conclusion of the argument go unstated. This creates a logical gap in the argument. 

If you think of the facts as wood pieces and the argument as the chair created when the wood is pieced together a certain way, the Assumption would be the crappy glue used to hold the wood together. 

A master woodworker, however, can cut joints in wood so well that the glue isn’t even totally necessary–it’s just used as a strengthener. Likewise, given the social, theoretical, and academic structures of the modern Western world, certain arguments will essentially fit together without requiring any non-trivial Assumption.

However, if the wood isn’t jointed precisely and the only thing keeping your ass from hitting the ground–hard–is a gigantic smear of that glue that you bought out of the sales bin at Harbor Freight, you might just realize what a liability glue is–and, if it’s not clear, by glue I mean Assumption.

You would be correct to recognize that you’re never going to turn an argument that involves Assumptions at all into a perfect deductive argument. You’re just trying to minimize the Assumptions to eliminate any other non-trivial or “reasonable” possibilities.

Let’s now look at the two main species of Assumptions in GMAT Critical Reasoning: Necessary and Sufficient.

Necessary vs. Sufficient Assumptions

  1. Necessary Assumptions

If an Assumption is Necessary, then it is something that absolutely MUST be true if the argument is to function properly. Conversely, lack of a Necessary Assumption would completely break the argument.

This is a little trickier to imagine, and is probably best done with an example. Let’s think about a classic Correlation vs. Causation scenario:

My Great Aunt Gladys traditionally plays bingo every day at noon. Yesterday, my Great Aunt Gladys wasn’t at bingo and someone knocked over a liquor store on the other side of town. Therefore, my Great Aunt Gladys knocked over said liquor store.

Now, of course, you might say that no one in their right mind would agree that a frail 86-year-old would be capable of armed robbery, but I counter with this: you don’t know my Great Aunt Gladys.

Let’s look at a Sufficient Assumption:

CCTV indicates that only one customer entered the shop and despite wearing a ski mask, the store attendant said from his hospital bed that the lady who shot him with the sawed-off moved rather slowly and couldn’t have been more than five feet tall. 

OK–it’s still not 100% because there could be other little old ladies with illegally modified shotguns out there (you don’t know Gladys’ friends), but the odds are looking non-trivially safe here. 

That is the beauty of a Necessary Assumption: there’s a really straightforward way to drive a wedge into it. (Alas, we never should have allowed the judge to remove her ankle bracelet.)

Questions Involving Necessary Assumptions

There are a few rough question types that indicate that you’re after a Necessary Assumption:

The argument assumes which of the following?

–The argument depends on the assumption that

–Which of the following is an assumption required by the argument?


Breaking Necessary Assumptions

The easiest way to break a Necessary Assumption is to use Reduction to Absurdity. Simply ask yourself this: if all the Facts are held true, how is it possible that the Conclusion is still false?

I’m just going to leave it there, because it’s discussed at length elsewhere.

The other, essentially similar, way to think about it is this: if a counterexample presents itself–

e.g. There is a known gang of elderly armed robbers stalking the other side of town

-then maybe the argument isn’t so good after all. 

Next, consider making the answer choice false. If you assume the negation of the answer choice itself and this harms the validity of the argument itself, you have found a Necessary Assumption.

  1. Sufficient Assumptions

These guys are the ones that seem weaker, but can actually be a bit trickier to deal with.

These are the scenarios that don’t necessarily track with obvious logical fallacies (e.g., Correlation vs. Causation, Don’t Kill the Bird While It Still Lays Eggs, Affirming the Consequent, etc.).

Actually, they don’t often sound that bad. In the real world, they’re probably reasonable leaps to make. However, remember that we’re not in meatspace here. In GMAT-Land, remember that we must eliminate all non-trivial options. 

Let’s take a look at an example: 

Dr. Tanner makes her grading criteria unclear, which makes it impossible for students to understand what will allow them to achieve good grades. It follows that in Dr. Tanner’s class, no one can be certain of getting a good grade.

Well… hold your horse, there, bucko. 

Start by asking yourself: “How is it possible that no one can be certain?” That’s a bit of an absolute. Try ye olde Reductio and see what happens. 

How is it possible that one can be certain of getting a good grade in Dr. Tanner’s class?

This is where the counterexamples start flying. The trick here is not to break the argument, but rather figure out what would make it work. This is an abstraction, of course, but the abstraction actually comes from being able to eliminate concrete counterexamples.

So just spitballing here: Dr. Tanner is a royal skeeze and she always gives good grades to the pretty sophomore she’s sleeping with. Dr. Tanner has previously said that “only a writer of talent and oeuvre comparable to Bill Shakespeare will get a good grade in her class.” And someone–perhaps even her little boytoy–proves to be that writer. 

Look–there are, theoretically, about a zillion (plus gamma) ways to have someone get a good grade in Dr. Tanner’s. Feel free to be ridiculous. This is a situation where a sense of the absurd comes in handy.

What our job is, of course, is to take this information and distil it into something that will make the argument work consistently. Essentially, we are taking an argument with a large Assumptive leap and defining the Assumption as fact–turning glue into wood. I’m sure that’s in the Bible somewhere.

This actually moves us from the realm of Inductive Logic to a purely Deductive scenario (again, I’ve discussed this at length elsewhere). 

Remember, with a functional Deductive logic, what is set up will get knocked down, just like an Algebra problem. You’re just metaphorically “rearranging to solve for x.

The problem is, though, that you gotta figure out the abstraction that will work here! What abstraction would get behind our example and flush out all the wacky counterexamples?

We need to find a way to clearly link the Facts and the Conclusion. 

Fact: Dr. Tanner’s grading criteria are unclear.

Conclusion: No one can be certain of getting a good grade in her class. 

How are those things linked? If you think about it, in reality they aren’t. At the most basic level, having unclear grading criteria does not make one uncertain of getting a good grade.

We are missing a key point here: only if we assume that unclear grading criteria directly implies uncertainty of getting a good grade, then we are golden. 

Fact: Dr. Tanner’s grading criteria are unclear.

Assumption: One can never be certain of getting a good grade when grading criteria are unclear.

Conclusion: No one can be certain of getting a good grade in her class. 

See what I did there? I routed out all the wacky side-theories about what people might possibly do to get a functional, abstracted link between the Fact and the Conclusion. 

Remember, the key with the Sufficient Assumption is that it won’t be factual. It isn’t about making up stories or edge-cases about why some shit might work if I look at it from a cockeyed angle. 

Rather, it is coming up with a generalized–that’s what I mean by “abstracted” if not clear–that clears up the holes in the argument and gets us functional like if a = b and b = c, then a = c. 

(If that doesn’t make sense, please refer to Russell and Whitehead’s Principia Mathematica. If I recall, it’s on about page 4300.)

Questions Involving Sufficient Assumptions

You can identify a question that is asking for a Sufficient Assumption by the following prompts:

–The conclusion above is properly drawn if which of the following is assumed?

–The argument above relies on which of the following assumptions?

Yeah, those sort of sound like the same questions as for Necessary Assumptions, don’t they? I wouldn’t worry about it. I think the distinction is more in “are we looking for a specific that can help” for a Necessary or “are we looking for an abstraction that can help” for a Sufficient.


Do you need to understand the distinction between Necessary and Sufficient Assumptions in order to succeed on the GMAT? After all this hoo-ha, I’m going to give that a solid “NO.”

NOW–that does not mean that your time here was wasted. Remember, it’s good to learn the details, integrate them, and forget them. 

How do we do that in this case? Look for the key takeaway.

What’s common between the two? Use your hand-cannon: Reduction to Absurdity. With a Necessary Assumption, you’ll blow a hole right through it. With a Sufficient Assumption, Reduction to Absurdity will highlight how to build your very own airtight, abstracted Assumption.

Those ancient Greeks really knew what they were talking about, eh? 

At the very least, this provides an interesting angle on how the questions are built. Understanding better what the testwriter is up to is the name of the game for GMAT Critical Reasoning.

As usual, forget the silly names and learn the skills. You’ll be fine. I promise.