Instructions for reconstructing arguments from texts

1.

Is there an argument at all?

Does the author try to support or justify a position by reasons? Look for indicator words such as “thus,” “hence,” “because,” “since,” etc.

Sometimes there are no indicator words but an argument, and sometimes there is one but no argument

2.

What is the position that the author tries to justify or support? Formulate it as the conclusion of the argument

Look for cues like a summary or an introductory remark such as “I will show that …,” “the point that I want to make clear is …,” etc.

Be prepared to change this formulation later if the reasons provided are not sufficient to justify what you thought the conclusion should be.

Justifiable positions can be descriptive / factual statements (“snow is white,” “knowledge is possible”) or normative statements (“we should pass the law,” we should strive for justice,” “prudence is an important virtue”) or evaluative statements (“this is a good argument”)

3.

Where does the argument start, where does it end? Keep in mind that authors often justify the reasons of a main argument by further sub-arguments, and so on.

Everything that belongs to the justification of one position is part of the argument for this position. Look again for the same cues or try to separate the argument from the topic the author discusses before and after the argument

Sometimes an argument can be reconstructed as plausible only if claims are included that the author makes in the context of what you first considered to be the argument. These claims need to be included. So be prepared to extend the boundaries of the argument.

Sometimes further analysis shows that there is more than one argument in the text you first identified as “the argument.” There is more than one argument if there is more than one conclusion. Even if different conclusions might differ only slightly, as long as the reasons that you can identify can support only either of them, they differ. In AGORA each conclusion needs to be justified on its own argument map.

4.

List all claims that the author makes within the boundaries of what you identified as the argument

Separate each claim—that is the statement that needs to be true for the argument to work—from phrases such as “X is convinced that ...” or “we believe that ...” whose truth is not relevant for the argument. Omit what is not relevant

Some of the statements you list will later turn out not to be part of the argument. They might be introductory remarks, explanations, comments, or background information. Include them nevertheless in your list

It is best to write each claim in a movable text box of some mindmapping or argument visualization software (you can use also tools such as Powerpoint).¹ You have to play with these claims (see below)

Of course, you might decide right from the beginning that certain statements do not belong to the argument proper. Be prepared, though, to go back to the text and look for further statements if you need them later to get to a plausible argument reconstruction

Check that each claim is formulated as a complete sentence so that you can assess its truth or acceptability without using information from the context. Neither claims nor the conclusion “should include pronouns such as they, my, it, that, and this. Instead the appropriate nouns should be used. Premises and conclusions should be in the form of statements—not questions, commands, or explanations” (Govier)

For each formulation you create, balance accuracy and charity. This means: try to remain as close as possible to the original formulations of the author whose argument you reconstruct, but provide an interpretation that is as plausible and reasonable as possible.

“Check that no premise or conclusion itself expresses an argument. For instance, if one premise says, ‘The party will do poorly in the election because the leader has made serious mistakes,’ you need to break down this premise further into 1) the leader of the party has made serious mistakes and 2) the party will do poorly in the election.” (Govier)

5.

Reconstruct the main argument by entering first the conclusion and then one or more reasons into your newly created AGORA argument map

Take the conclusion and play with the claims of your list. Which claim—or which combination of claims—would fit best to justify the conclusion?

It might be necessary to tweak the claims of the main argument for a better fit. But keep in mind that using entirely different claims might allow a better fit. Keep an open mind and look again at the text and your list of claims

Look at the text and ask yourself whether the structure of the main argument could be reconstructed as a modus tollens, a disjunctive syllogism, or a not-all syllogism. Only if this is not the case, select modus ponens in AGORA to complete the argument when the system asks you to do so. Have a look at http://agora-info.spp.gatech.edu/learn/validity

In AGORA you can change both the selected argument scheme and the language form of the argument scheme later. This will change the way the parts of the argument are connected. Always check carefully whether the enabler that the system creates automatically based on your input and selection is acceptable! The formulation of the enabler shows you if the relation between reason and claim is sufficiently strong. Start from this assessment to see whether you need to change anything. For this step, consult http://agora-info.spp.gatech.edu/learn/goal#reflect

Are there several independent arguments for the same conclusion or just one main argument with mutually dependent reasons?

Check http://agora-info.spp.gatech.edu/learn/arguments to answer this question. Everything that is connected by one enabler in AGORA forms one independent argument.

Sometimes one of the claims in your list might be very similar to a system-generated enabler. In this case do not enter this claim as a reason. It would only make the argument more complicated. If there is such a claim in your list, it is most likely that the author develops a logically valid argument.

6.

Reconstruct sub-arguments for the reasons of the main argument

Follow the instructions above for the main argument

After reconstructing the main argument the number of schemes available for sub-arguments is limited by the fact whether the reason you want to support is a negated (“it is not the case that”) or an affirmed statement. But whatever it is, you will always find at least two different schemes that you can use to reconstruct the sub-argument. If you identified the main argument as a modus tollens or a disjunctive syllogism, go from there and reconstruct the rest so that it fits to the main argument.

7.

… and then sub-arguments for the reasons of these sub-arguments, and so on ...