In law, a warrant is a document that authorizes an officer to make a search, seizure or arrest; it guarantees that what the officer does occurs within the purview of the courts and so is within legal sanctions. In other words, it gives the OK to the officer's actions, and, when those actions are subsequently the subject of discussion in a court of law, they will be deemed to have been aboveboard and, thus, admissible.

In argumentation, the warrant serves a similar purpose. It guarantees, within a certain degree of probability, that the inference can connect the cited evidence and the claim being advanced. Warrants shore up our inferences and ensure that they will be found to be aboveboard. In case you've not yet seen the analogy I'm attempting to make, the inference is the arrest, and the arguer is the officer. The warrant, which may or may not be stated, is there to OK the arguer's inferences so that when the argument comes under scrutiny, the inference may be both acceptable and defensible.

Who, then, is the judge in this case? Who issues the warrant? That's right: the audience. When it comes down to it, the audience will decide whether the inference will pass. This is not to say audiences determine what is true or not truealthough that is so in a very pragmatic sense it is to say, however, that the arguer must back up his inferences with warrants his audience will accept.

Let's return, for a moment, to our diagram of the Toulmin schema. For simplicity's sake, we'll consider just four parts: claim, evidence (grounds), warrant and inference.


We can see here that the inference is the connection we are making between the evidence and the claim and that the warrant is what authorizes us to do that: in the model, the warrant is "holding up" the connection. As mentioned briefly in the link at the Toulmin model page, there are six (6) different types of warrants. Each warrant is categorized according to the kind of inference being made. Below, we will examine those a little more closely.

Types of Warrants


Very often we reach conclusions (i.e., make claims) based on some limited examples, whether those are statistical or anecdotal. When we furnish examples, we are assuming that what is often true of the parts may be taken to be true of the whole. In such cases, we are saying that the examples are "representative" of a larger class. Conversely, we may often reason that what is true of a class is also likely to be true of its members. Either way, we are reasoning in a way warranted by generalizations.

Example: There's a lack of education going on in today's schools [CLAIM]. Of the twenty-five college freshmen I spoke to, nine of them had never heard of Shakespeare, twelve didn't know how to factor a trinomial, and four had difficulty spelling their own names [GROUNDS].

To watch out for: fallacies of composition, division, secundum quid (hasty generalization), dicto simpliciter (sweeping generalization).

Grounds: LOGICAL


Merriam-Webster Online's first definition for analogy has an argumentative meaning: "inference that if two or more things agree with one another in some respects they will probably agree in others." Note the words "inference" and "probably." These two words are key to our understanding of the analogical warrant. When we compare known ways in which two things are alike to infer ways in which they are not known to be alike, we are reasoning by analogy. (Remember, informal argument seeks to go from what is known to what is not known.) Analogies may be literal or figurative. In the first case, two things are compared for what makes them similar in known respects to infer what makes them similar in other respects.

In the second case, that of the figurative analogy, not two things but two relationships are compared. For example: David will never win this battle with Goliath because he's basically brought a knife to a gun fight. (We know from the story that the analogy wouldn't hold up.)

Example: Iraq is turning into another Vietnam.* (Evidence will then be cited for the Vietnam and Iraq wars are similar. The inference
the leap to new information is then that these similarities justify treating them in similar ways.)

Let's take another example, this one from the recent Yahoo! article about NASA's October lunar mission:

Has a hyperactive five-year-old taken over as the director of NASA? It sure seems like it. [CLAIM] On Friday morning, an unmanned spacecraft launched in June will crash into the moon's surface. On purpose [GROUNDS].

Let's put this into the Toulmin form more fully. Claim: It sure seems like a hyperactive five-year-old has taken over as the director of NASA. Grounds: An unmanned spacecraft launched in June will crash into the moon's surface on purpose. Warrant (unstated): Analogy.

What is the already known information in this analogy? That hyperactive five-year-olds smash things on purpose. It is also known that the director of NASA has given the word to smash things on purpose. Therefore, so goes the claim, the director of NASA is like a hyperactive five-year-old. This is the new information. How we judge the actions of a hyperactive five-year-old, so goes this argument, is how we are to judge the actions of NASA's director: as undisciplined, irresponsible, destructiveand in need of correction. (While these things are left unstated, they are nonetheless "there.")

Question: Is the similarity between the five-year-old and the NASA director enough to justify viewing the latter as we would the former?

Note: Sometimes, the "like" or "as" of comparison is present, making the claim a simile. When the "like" or "as" is not present, however, we have a metaphor. Both are held up by analogical warrants.

To watch out for: false analogy: if an analogy shows there to be more relevant differences between two things than there are relevant similarities, the analogy is unsound. I would argue that the analogy offered in the Yahoo! article is a false analogy because, while both the leader of NASA and the hyperactive five-year-old smash things on purpose, the hyperactive five-year-old does not do so after months, even years, of planning and deliberation in order to gather scientific data.

Grounds: LOGICAL, but sometimes with a PATHETIC element (as when the comparison is meant to evoke an emotional response)


The weight of experience accumulated over many years shows some things to be highly reliable indicators of other things. A simple example is smoke, which is taken to be a sign of fire. Our accumulated experience tells us that where there is smoke, there is—or very recently has beena fire which caused it. Not all signs, however, are infallible. Some believe, in fact, there are no infallible signs, and that, as with all informal argumentation, there is always a degree of probabilityand so the chance that the inference based on sign may be wrong.

Example: Beth studies diligently for four hours everyday [GROUNDS]. She is probably an A student [CLAIM]. The warrant is that long, consistent and diligent study is a sign of a high grade-point average.

To watch out for: signs which may counter the claim suggested by the sign we cite; also, the fallacies of post hoc ergo propter hoc and cum hoc ergo propter hoc (i.e., correlation is not causation).

Grounds: LOGICAL

Cause and Effect

It sounds easy enough, but reasoning warranted by established causal relationships has several potential pitfalls. For one thing, what do we mean by "cause"? Do we mean "sufficient condition"? Do we mean a human element that gave rise to a certain effect? This is a complicated matter that I'll leave for the more high-flown explanations of philosophers to deal with.

Obviously, in trying to establishing cause and effect, we are trying to establish responsibility. Reasoning in this way can thus serve epideictic ends: where shall we assign either praise or blame? But reasoning this way is essentially a matter of forensics, which seeks to ascertain the truth about what happened in the past.

Example: Sarah dropped out of school because she spent more time partying than she did studying [CLAIM]. Once she started partying, her grades started to drop [GROUNDS].

To watch out for: as with sign, the fallacies of post hoc ergo propter hoc and cum hoc ergo propter hoc (i.e., correlation is not causation). In the example above, what if Sarah's partying and her dropping out of school merely occurred together? Further evidence might suggest that Sarah started partying more because she realized she didn't have enough money to continue her education, and for that reason dropped out of school. In that case, both her partying and her dropping out of school were caused by the same thing.

Grounds: LOGICAL


None of us can possibly know everything about everything. Because knowledge, like labor
and the two often go hand in handis highly specialized, some people are more reliable experts on some subjects than they are on others. When arguing a case, many times we may be arguing in favor of a view we believe to be true not because we have studied the literature vastly or have conducted experiments ourselves but because we are acquainted with the findings, conclusions or pronouncements of a certain authority. To boil this warrant down to its simplest form, the warrant of authority says, "In the absence of or as supplement to our own direct observation, credible sources can provide acceptable evidence."

Here is where we must be careful. Not all authority carries equal weight on every issue. Sometimes authorities in the same field may not be as able to speak on some matters as on others. For example, English teachers may be excellent sources of information about what is contained in Shakespeare and may even be very good people, but they may know very little about political science and how a government should be run. Citing an English teacher as a reason for one's political opinions about, say, welfare or the national budget is unwarranted, unless the professor is especially well versed in the topic. This is not to say that opinions from humanities professors do not sometimes legitimately inform political opinionsindeed they canbut it is to say that our knowledge cannot possibly cover every area. When you have the choice in citing authorities, go for the one who has the most expertise.

Example: Don't fight back when pushed by bullies [CLAIM]. After all, Gandhi said that "an eye for an eye and a tooth for a tooth leaves the whole world blind and toothless." [GROUNDS]

To watch out for: argumentum ad verecundiam (appeal to irrelevant authority).

Grounds: ETHICAL


Reasoning from commonplace is reasoning from what Peter Berger and Thomas Luckmann call a "common stock of knowledge." Common beliefs, assumptions, attitudes and values all factor into this warrant, sometimes in the form of proverbs and maxims. As David Zarefsky notes, here is where the enthymeme often comes in. An enthymeme is a syllogism with the major premise left unstated because it exists at the level of basic assumptionit doesn't have to be explicated for the audience, in other words.

Example: Bob should marry Jane [CLAIM] because they love each other [GROUNDS]. (The unstated assumption is that people should marry for love, not for money or convenience.)

To watch out for: Commonplace assumptions are sometimes at variance with one another. For example, if Bob were already married to Jennifer, then he might not be justified, in the eyes of certain audiences, in divorcing Jennifer to marry Jane.

Also to watch out for is the dicto simpliciter, also known as the sweeping generalization. Stereotypes, which often come in the form of sweeping generalizations, look like legitimate arguments based on a warrant of generalization. However, the sweeping generalization is most often acceptable only to audiences who are already predisposed to believe in them, and the warrant is really a commonplace
and a shaky one at that.


*This is obviously a claim of fact with an implicit claim of policy, i.e., we should get out of Iraq now. Here, the claim of fact functions as the evidence of the unstatedbut certainly understoodclaim of policy.

Sources Consulted

"Criteria for Analyzing Arguments." http://www-rohan.sdsu.edu/~digger/305/criteria.htm. 13 September 2009.

Reynolds, James. "Blocks to Critical Thinking." Invitation to Critical Thinking. New York: Holt, Rinehart, & Winston, 1984.

Zarefsky, David. Argumentation: The Study of Effective Reasoning. 2nd Edition. Chantilly, VA: The Teaching Company, 2005.