Because

  How to Analyze and Evaluate Ordinary Reasoning

  Section  11:  Instances

 G. Randolph Mayes

 Department of Philosophy

 Sacramento State University

When people reason they often provide concrete examples of the sorts of things they are talking about.  For example:

• Chuck is really not a very considerate person.  He hardly ever expresses any form of gratitude or appreciation.  Yesterday I bought all of his kids Slurpees and doughnuts and that guy didn't even say thank you.

Here the speaker is arguing that Chuck is an inconsiderate person based on the claim that Chuck rarely expresses any form of gratitude.  She then provides an example of Chuck's ingratitude: the fact that on one particular occasion he did not say thank you. 

There are basically two different ways of looking at this reasoning.  On the one hand, you could interpret the example given as a reason.  In other words, you could reconstruct the chain of reasoning as follows.

So far we have only been interested in reconstructing rationales, not evaluating them.  But in order to talk about the problem with this reconstruction, we need to express ourselves in an evaluative mode.  R1' just gives one example of Chuck's ingratitude.  So if R1' were seriously meant as a reason to believe R1 we would say that it is a very poor reason.  After all, Chuck could have just forgotten, or been in a hurry, or he might actually may have been a little bit pissed that you bought his kids a bunch of junk food.  Our point here would be that you have to be careful about generalizing from just one example, and it's a very good point.  This is a logical error that goes by different names like "hasty generalization" and "confirmation bias", which we'll talk about later. 

But generally speaking we want to avoid attributing logical errors to people.  If there is a way to interpret their reasoning as stronger rather than weaker, that's the way we should do it.  (This is what we call the Principle of Charity in interpretation, which we'll also talk about later.)  One way of guarding against the interpretation above, then, is to say that what we've called R1' isn't really intended as a reason at all.  It's just one of possibly many examples, or instances, of what the person is talking about.  In fact, this is really common.  Here is another example:

• The reason Chuck doesn't like people buying his kids junk food is that obesity runs in his family.  Chuck's mother and his sister Sarah are morbidly overweight.

This is an explanation that uses specific instances of the reason provided, namely Chuck's mother and sister.  It does not really make any sense to call these reasons, because they do not  represent causes of the tendency toward obesity in Chuck's family. 

In order to deal with examples like this we formally introduce the idea of an instance.

• Instance:  An instance is a concrete example of something claimed in a reason or conclusion.

When we represent a statement as an instance we connect it to the corresponding reason with a line segment rather than an arrow and label it with an 'I' as follows.

(You will note that in the explanation above, R1 does not occur explicitly in the original reasoning.  However, It occurs implicitly in the sense that the speaker just takes it for granted that we will understand that Chuck objects to his children eating junk food.  When we reconstruct reasoning it is important to make implicit reasons and conclusions explicit.  Otherwise the logical connections will appear to be weaker than they actually are.)

11.2 Instances and implicit reasons

The concept of an instance is a useful tool for identifying implicit reasons and also for providing strong unified interpretations of people's reasoning.  The basic point to notice here is that people often provide what appear to be several reasons for a conclusion which are actually more charitably interpreted as several instances of a more general reason that has not been explicitly stated.  For example:

• People tend to avoid Harold because he's belligerent, stupid, self-pitying, and dishonest.

Superficially, this example presents as four distinct reasons for a conclusion.  So one way of reconstructing it would be:

There is nothing technically incorrect about this representation, but it seems to be missing something, and that is that all these reasons are reasons of a certain kind.  Specifically, they are instances of offensive or unpleasant characteristics.  Hence, an alternative way of representing this reasoning is:

The reason this is a better interpretation is because R1 was implicit in the original reasoning, and now it is explicit.  We say that R1 was implicit, because the original reasoning could not have made any sense to you if you did not recognize all of the reasons as unpleasant characteristics.  One way of appreciating this is to notice that if the original reasoning had been:

• People tend to avoid Harold because he's generous, intelligent, courageous, and trustworthy.

you would have said that this reasoning makes no sense, because these are all positive characteristics.  Anytime we are processing reasons we are searching for a level of generality that helps us to make sense of what is being said.  Most of the time we are totally unaware of this.

11.3  When to separate instances

You'll notice that in the reasoning immediately above we put all of the instances in one box.  This is ok as long as there are no reasons or instances given for the instances themselves.  Whenever an instance or a reason is given for an instance, the instance needs to be put in it's own box, as we explain below.

11.4  Instances of instances

People often give instances of instances.  For example:

• Chuck is way too strict with his kids.  There are just so many perfectly harmless things that he won't allow them to do.  He won't let them eat any junk food, like candy or soda or chips.  And he won't let them watch really good PG-13 movies like Austin Powers and Talladega Nights.

You may have noticed that in the above rationale we did not express the terminal instances as complete sentences.  No harm is done by abbreviating the terminal instances in this way.  But when instances are themselves supported by reasons, it is important to write them out as complete sentences.

11.5  Reasons for instances

Although instances commonly occur at the end of a chain of reasoning, it is not at all unusual for the instance itself to be supported by a further reason.  For example:

• Before they turn into butterflies, caterpillars avoid detection by predators by employing various types of camouflage.   The caterpillar of the Asian swallowtail butterflies, for instance, have black and white spines that make them look similar to bird droppings. Birds that feed on the Asian swallowtail typically do not recognize it as prey in this stage of development.

Whenever you represent a rationale in this manner, be especially sure to check to see that the reason supporting the instance continues to function properly.   Since the example above is an explanation, all of the reasons must function as causes.  You can see that R1' does identify what causes the black and white spines of the Asian swallowtail to serve as camouflage.  However, you may have also noticed that the final sentence: "Birds that feed on the Asian swallowtail typically do not recognize it as prey in this stage of development." does not occur in this reconstruction.  This is because this statement is not serving in an explanatory capacity.  For the sake of providing a unified reconstruction it would have been tempting to represent the reasoning as follows:

But this would be incorrect.  In order to incorporate R1" into the reconstruction, it must simply be identified as a small, separate argument with the instance above as it's conclusion as follows:

11.6  Instances of conclusions

Conclusions may also have instances.  For example.

• Pet owners have long been known to suffer less from stress, but it turns out that proximity to animals brings other health benefits as well.  Kids who are raised with dogs and cats are much less likely to develop asthma when they get older.  People who work with livestock are less likely to develop lung cancer.  And the recent National Health and Nutrition Examination Study found that people who never owned a cats are 40% more likely to experience heart problems. 

11.7  Instances to avoid

Generally speaking you should avoid constructing rationales that have reasons that themselves make explicit reference to reasons, causes or evidence.   For example, this reasoning

• The reason we think Jimmie Dang shot the junkie is because all the evidence points in his direction: Jimmie is a drug dealer; the dead man was dealing drugs on Jimmie's turf; Jimmie owns a small caliber gun like the one that was used in the shooting; and two eye witnesses said they saw a large Asian man fleeing the seen of the crime.

This example is presented in such a way that it would be very natural to characterize each piece of evidence as an instance of one reason as follows:

In the argument above the reason given as evidence that Jimmie Dang is the shooter is that "There is evidence that Jimmie Dang is the shooter."  The problem is the statement that there is evidence is not itself evidence.  (Similarly, the fact that there are causes is not itself a cause.)  Hence, even when people make explicit claims that evidence, causes, or reasons for their conclusions exist, these statements should ordinarily not themselves occur as reasons in the reconstructed rationales. An alternative reconstruction of the argument above would be: