Philosophy consists of arguments. The works of philosophers are not primarily collections of assertions, or compilations of data, or aesthetic productions (although they may be these in part). Rather, they are primarily collections of arguments.
ARGUMENT. Since we are going to spend much of our time talking about arguments, we should have a definition of "argument." In general terms, an argument is a set of premises and a conclusion. What, then, are premises and a conclusion? To begin with, they are statements:
STATEMENT: A sentence that is true or false. This does not mean "a sentence that is known to be true or known to be false." It means only "a sentence that can sensibly be called true or false." Sentences that are not statements include questions, commands, expressions of feeling. Consider these examples:
The first sentence is a question. It makes no sense (or not any obvious sense, at least) to call it true or false. If you ask me what time it is and I respond, "Yes," or "That's true", you will very likely conclude that I didn't understand what you said. Likewise, if you tell me to shut the door and I respond, "That's correct" or "That's false," you will again think I didn't hear what you said. Questions and commands aren't true or false. The last sentence, however, is a statement. If you say that Texas is the largest state in the US and I reply, "That's true," my response makes sense. It would also make sense if I replied, "That's not true."
So far, then, an argument is some statements (the premises) and a statement (the conclusion). What else do we need to say about the relationship between premises and conclusion? One way to express it would be this: arguments are offered as support for their conclusions, or as reasons for believing their conclusions, or in an effort to persuade someone to accept their conclusions. These are all descriptions of the uses to which arguments are put. However, we can't say that an argument is a conclusion together with some premises that support it, or some premises from which it follows, or some premises that give a reason for believing it. The problem is simply that not all arguments are good arguments. If the premises of an argument don't actually support its conclusion, then it will be a poor argument, but we will still allow it to be an argument. We don't want to set a limit on how bad an argument can be before it ceases to be an argument at all, so we will just say that an argument is some premises and a conclusion.
Since we're talking about arguments, we need an example. Here's a philosophical one:
Everything that I believe myself to know has been acquired by means of my senses. However, there have been times in the past when my senses have deceived me. So, it is possible that I am mistaken about anything that I believe myself to know. But if it is possible that I am mistaken about something, then I do not know it. Therefore, I do not know anything. (Roughly after Descartes, Meditations I)
Arguments have a certain overall structure. Every argument is intended to establish, prove, convince you of, or justify some statement. We call that statement the CONCLUSION. An argument accomplishes this by presenting, or assuming, or asking you to accept, some other statements, each of which is called a PREMISE. We can rearrange the example above to show this structure by listing the premises and the conclusion in sequence:
Philosophers sometimes do present arguments in just this way, even numbering the premises. Most of the time, however, even philosophers us more discursive prose. In such cases, we usually find various words that serve to indicate the structure of the argument. One class of such words includes terms that mean, in effect, "here is the conclusion of an argument": therefore is a nice example.
EXERCISE: Look for some examples of arguments, both philosophical and non-philosophical. Some good sources, apart from works by philosophers, are newspaper and magazine editorials, political rhetoric, literary and art criticism, legal defenses or prosecutions. Try to identify the structures of the arguments in them.
An argument is intended to give a reason for believing its conclusion. In very simple terms, you may think of an argument as saying this:
The force of "therefore" amounts to this: if you accept all the premises of this argument, then you must also accept the conclusion. Why would you have to accept the conclusion if you also accept the premises? A very good reason would be this: that it is impossible for the conclusion to be false if the premises are all true. Not every argument has this property. When an argument does have this property--that is, when it is impossible for its conclusion to be false when all its premises are true--then we call it valid. If an argument is valid, then it would be irrational to accept its premises and not accept its conclusion.
You will often find alternative ways of expressing this same notion. When an argument is valid, we also say that its conclusion follows from its premises and that its premises imply or entail its conclusion. Take note that in philosophical usage, "imply" is more likely to have this sense than its more everyday meaning of "suggest without explicitly saying."
The first thing that a good argument should be, then, is a valid argument. However, validity alone is not enough to make an argument a good reason for accepting its conclusion. Validity is, in effect, conditional: if you accept the premises, then you must accept the conclusion. However, if you don't accept all the premises, then the argument gives you no particular reason to accept its conclusion. What would make the argument a good reason for accepting its conclusion is its having true premises. If an argument is valid, and if in addition all of its premises are true, then its conclusion must be true. We call such an argument (a valid argument with true premises) a sound argument.
Determining whether an argument is valid is the task of logic. For philosophers, this is an important point because logic itself has a close connection with philosophy. Important differences of opinion exist among philosophers about precisely which account of validity is the correct one. For this reason, philosophers should know something about logical theory. In most cases, however, philosophical arguments do not involve the kinds of inference that are disputed. Most of the time, when philosophers disagree about the validity of a philosophical argument, they disagree about the way that argument is to be analyzed.
The most straightforward way to show that an argument is invalid is to show that it is possible for its premises all to be true while its conclusion is false. A common way to do that is by describing the circumstances in which just that would happen. A related way of showing that an argument is invalid is by finding an argument of the same form that has obviously true premises and an obviously false conclusion. What is the meaning of of the same form? That, again, is part of the subject matter of logic.
Determining whether the premises of an argument are all true is usually not a matter for logic. Speaking generally, the premises of an argument could be about anything: Sumerian history, molecular biology, Roman law, human anatomy, astrophysics, baseball statistics. If you want to know whether the premises of an argument about baseball statistics are true, what you need is knowledge of baseball statistics, not logic. In philosophical arguments, however, there are ways of determining the truth of premises that are peculiarly philosophical. To begin with, for any interesting philosophical argument, there are usually further philosophical arguments that aim at determining whether its premises are true. We will explore some of the ways philosophers do this later.
Up until this point, we have been discussing deductive arguments. Deductive arguments are those argument for which validity is the appropriate standard of correctness. Since at least the time of Aristotle, philosophers have distinguished another type of argument for which validity is not the appropriate standard: inductive arguments. These are arguments that infer a general claim from a number of instances of it:
Now, this argument (like inductive arguments in general) is invalid: it is not impossible for its premises to be true and its conclusion false. The more dogs we observe to be hairy, the more confident we may be that every dog is hairy, but it still remains possible that, although every dog ever seen is hairy, nevertheless there is some dog somewhere that is hairless.