arguments |
ABSTRACT Stephen Toulmin's model of an argument structure has three elements: grounds, warrants, claims. A simple argument illustrates this model. A claim in an argument is a conslusion inferred from evidence. Claims have to be limited or they are vulnerable to rebuttals. Grounds are descriptions of experiences. Warrants are assumptions. Sometime warrants require backing. Sometimes they are implicit. Arguments invite rebuttals and can be refuted. |
Arguments are one of the most important discursive structures in our uses of language. It is the outcome of our ability to reason, that is, to draw inferences from experiences. Arguments are the expression of a reasoning process. Reasoning has three necessary cognitive components: experiences, inferences, and conclusions. The most widely quoted instace of reasoning in our culture is "where there is smoke, there is fire." From our experiences of observing things burn, we know that fires produce smoke. So, when we see smoke, we infer that there is a fire causing it.
The discourse pattern: "if smoke, then fire" is the expression of an inference (drawing a conclusion by linking one phenomenon to another by a causal relation). The reasoning process goes something like this: Given the fact that I see smoke, I assume that where there's smoke there's a fire, so I conclude there is a fire where I see the smoke. Formulaically: Given x, if y, then z.
Stephen toulmin's uses of argument
The philosopher Stephen Toulmin's theory of informal logic is the most widely used account of argument discourse structures. His model has replaced the older Aristotlean model of the syllogism largely because syllogisms are artificially contructed discursive patterns than rarely appear in our typical discourses whether common sensical or technical. Taking his clues from legal discourse Toulmin identifies the three constituent parts of an argument structure as "claims," "warrants," and "grounds." The following chart gives the main synonyms for Toulmin's terms:
| grounds | Given x | experience | data | evidence | facts |
| warrant | if y | inference | hypothesis | reason | premise |
| claim | then z | conclusion | proposition | belief | assertion |
In Toulmin's model there are two other discursive markers typically found in argument structures: backing of warrants and modality of conclusions. Sometimes, the warranting assumption in an argument is not widely accepted. So, it has to be "backed up" with additional arguments or authorizations. And, most claims are limited by modifiers (modals). Instead of saying: "Robert Redford always plays in excellent films," it helps to limit your claim to "most often plays" or "usually plays" in excellent films.
The Chicago Bulls will win the NBA championship given their superb defense because defense trumps offense in the playoffs.
| given their superb defense | because defense trumps offense in the playoffs | The Chicago Bulls will win the NBA championship |
| GROUNDS | WARRANT | CLAIM |
| Given x (Bull's defense) | If y (we assume that the better defensive team will defeat a better offensive team) | Then z (Bulls win NBA championship) |
Though this argument has all of the necessary components, it is not a very good one.
However, notice that the argument can begin or end with differing elements:
The Chicago Bulls will win the NBA championship given their superb defense because defense trumps offense in the playoffs.
Given their superb defense, because defense trumps offense in the playoffs, the Chicago Bulls will win the NBA championship.
Because defense trumps offense in the playoffs, the Chicago Bulls will win the NBA championship given their superb defense.
Even though it's impossible to say, that "argument structures begin with grounds, followed by warrants, and end with claims," it is easy to locate a declaration and examine the neighboring sentences for descriptions that warrant it.
A claim is an assertion, belief, proposition, or conclusion. In the preceeding argument, the claim is best described as a "belief" that is being proposed or asserted. [From a discursive point of view, claims are "declarations." They are public statements that insist something is the case, that some state of affairs exists.]
The claim that "The Chicago Bulls will win the NBA championship" is clear enough but it is unnecessarily overstated--"will win."
[Often claims are stated in unclear or ambiguous terms. For example, the claim that "The best defensive team will win the NBA championship is unclear (it doesn't identify a team) and also ambiguous (since the NBA season hasn't started no one knows which team will be the best defensive team).]
The assertion that the Bulls WILL WIN the NBA champtionship is overstated. There are various ways to modify this claim and make it less vulnerable to rebuttals. For example:
The Bulls should win ....
The Bulls may very well win ....
Providing there are no serious injuries to their starting five, the Bulls could win ...
Unless Gordon does not improve his defense, the Bulls are capable of winning ...
Unlimited claims are quite vulnerable to rebtualls.
All women are ...
All Frenchmen are in the habit of ....
Every time he drives, he .....
Whenever she feels like it, she ...
Some women ...., some frenchmen....., often times...., Usually ... stregthen your claim because it becomes easier to support it.
When a "claim" is made but is not supported with evidence (grounds), then it is simply a declaration of opinion. But, when claims are grounded, an argument structure is created.
Support for claims are invariantly direct or indirect descriptions of experience. As descriptions they can be "witnessed." Other persons should be able to experience the situation described in support of a claim.
Not all descriptions, however, are of equal weight in arguments. Descriptions have varying degrees of support. The strongest type of description is "factual." As Toulmin remarks, facts show "specific features that mark off this precise situation from others and so point toward this specific claim or conclusion rather than others" (Introduction to Reasoning, 33).
Not all descriptions of experience are factual. Facts are observable. For example, in the argument above "given their superb defense" is not observable. If I can back it up with statistics, I indicate that authoritative witenesses have observed and counted the number of steals, blocks, and turnovers the Bulls have marshalled over the course of several games if not an entire season.
Steals, blocks, forced turnovers are observable. Furthermore, all NBA games are recorded on video and the statistics can be recounted and verified.
We have to note at this juncture that statistical evidence can be misleading. For example, if the Bulls have more steals, blocks, and turnovers than their opponents but lose the game, then the statistics are misleading. This brings us to the assumptions underlying the argument.
The overriding assumption in the argument is that "defense trumps offense in the playoffs." This assumption is not only ambiguous but also not definitively proven. It's a common assumption, but not necessarily a scientific law. First, let's remove some of the ambiguities and vagaries in the warrant.
The metaphor "trumps" is understandable but open to a wide range of interpretations. Does "trump" mean that the better defensive team (as yet undefined) will beat a better offensive team (also undefined)? Does "better" mean "gives up fewer points than" as well as "scores more points than"? Ok, let's accept this gloss. Now can we establish that "steals," "blocks," and "turnovers" are related to scoring fewer or more points?
Here again we are thrown back to statistics but this time statistics over a period of time when the game was played in relatively similar fashion. Let's stipulate that the game has not changed drastically in the last five years. This is a difficult stipulation because one of the rule changes that was instituted in 2006-2006 season had to do with "hand-checking," a defensive maneuver that was outlawed.
The intial assumption "defense trumps offense in the playoffs" has been complicated. It can't hold up unless we "back it up."
We have already anticipated the backing in the previous section. To make the assumption (now rephrased to be less vague and ambiguous) that "in the playoffs, the "D" team that typically holds it opponents to fewer points than the "O" team it is playing against will normally defeat the "O" team even if that team scores more than the "D" team does." Quite a tangled sentence and even so, it requires further backing.
Not only do we have to back this warrant up with statistics about the average number of points the "D" team give up and the average number of points the "O" team scores, but we also have to calculate the differential in points with respect to the caliber of the teams they played. And then we also have to note which players the "O" & "D" teams and their opponents played in the games in question. And so on, and so on.
Fortunately, ESPN's John Hollinger does all of this for us, saving us a whole bunch of time.
In analyzing an argument structure, it is very important to recognize that many arguments, especially everyday ones, do not make their warranting assumptions explicit. In such instances, the writer or speaker believes that either the warrant can be taken for granted or that his or her audience holds the same warranting belief and, therefore, it is not necessary to make it explicit.
It is common to hear abbreviated arguments such as:
Hybrid engines get twice as many miles to the gallon as simple combustion engines. I'm going to buy one.
It is not necessary to make the warrant in this argument explicit. cars that use less fuel are economical. The fact that a hybrid engine gets more miles to the gallon than conventional engines is grounds for buying one. Most people do not believe that cars that use more fuel are more economic to run than cars that use less fuel. In this case, the warrant can be taken for granted.
In other instances, for example persons very familiar with each other who frequently discuss a given subjecy, the speaker/writer already knows that the person with whom he or she is in conversation holds the same warranting belief. For example, if I am discussing the Bulls with my son, I know that, like me, he believes that defense trumps offensive in the playoffs. So, if I say that the Bulls are the best defensive team in the Eastern conference and, if their main plaers are healthy, they are very likely to win the conference and get to the championship game. Normally, our conversation will focus on the evidence for the Bulls being a better defensive team than Detroit or Cleveland. This brings me to rebuttals and refutations.
If my son, Greg, challenged the evidence for my claim about the Bulls, he might point out that Ben Wallace is a year older, that Ben Gordon gets smoked, that they don't have the size of Cleveland or Detroit, and so on. His "rebuttal" (counter argument) would focus on my grounds.
On the other hand, if I were arguing with a fan of Phoenix, he might attack my warrant thata "defense trumps offense in the playoffs," pointing out that in the western conference win by outscoring their opponents and that Phoenix's success with this strategy has altered te way coahes think about the game of basketball.
I might counter with the fact that Phoenix not only did not make the championship game but that the tema that did in the western conference was the best defensive team, namely, the Spurs.
He might counter with the fact that Phoenix lost because Comminssioner Stern suspended two of Phoenix's starters and quote Rsheed Wallace (PIstons) who remarked that under Stern, the NBA is no different than the World Wrestling Association--all orchestrated for TV.
Such counter-arguments, challenging evidence or warrants, can go on for a long time in areas such as sports.but in technical disputes refutation can put an end to an argument. The difference between a rebuttal and a refution hinges on prediction for the most part. If a scientist predicts that the drug he invented will cure cancer, then he is predicting that persons who are suffering from that disease will regain their health if they take his drug as prescribed. But, if, instead, they die, his argument is refuted by this outcome.
jjs
← previous | next →
last revised:
June 13, 2007
Send comments to jjs.
copyright © jjs, 2007