Wednesday, November 19, 2003

Philosophical Buzzwords: Soundness and Validity

So I had a great idea forever ago, and got as far as one post, then stopped. Fortunately, history hasn't ended yet, so it's not too late for me to continue. A reminder for those of you who haven't been attentively reading this blog from the very beginning: philosophical buzzword posts are designed to explain philosophy concepts to smart people who don't already know them. I think that more philosophical experts read my blog now than did in September... those people will probably not find these posts very interesting, unless I make a mistake. Enough disclaimer. Here's my post. I mean to explain what it means for an argument to be "sound", "unsound", "valid", and "invalid". Arguments start with premises and end in conclusions. The premises in an argument are connected to the conclusion by a series of steps. The validity of an argument depends on whether each step follows logically from the ones before it. So an argument is valid if the statements in each step are logically required by the ones before it. Here are some examples of valid arguments:
  1. (premise)Jonathan has a blog.
  2. (premise)Anyone who has a blog uses the internet. Therefore,
  3. (conclusion) Jonathan uses the internet.
  1. (premise) Emily has a nephew.
  2. (premise) Anyone with a nephew must love children. Therefore,
  3. (conclusion) Emily loves children.
Here is an example of an invalid argument:
  1. (premise) Jonathan's name is "Jonathan". Therefore,
  2. (conclusion) Jonathan is male.
This argument is invalid because it's not a logical requirement that Jonathan be male, given only that Jonathan's name is "Jonathan". It would be valid if we added the premise, "All people named "Jonathan" are male." (That's not the only way we could make it valid. We could, for instance, instead add the premise, "Everyone is male", and the conclusion would follow.) So that's validity. An argument is sound if (1) it is valid and (2) its premises are true. (Intuitively, a sound argument is a good argument -- one that should definitely convince you of its conclusion.) So the blogging argument above is sound in addition to being valid, while the nephew example is valid but not sound. It should be evident from this discussion that soundness implies validity. So technically speaking, Dave's example in a comment to my last post is not valid, although it's pretty close: "We should not permit anyone who is a known terrorist to fly, because known terrorists are likely to try to kill people on the plane." This argument is not strictly valid because, like the naming case, it's missing a (very reasonable premise). If we add in, "We should not permit anything that is likely to cause an attempt to kill people on the plane," then the argument becomes valid. This suggests something else worth recognizing -- there are three conceptually distinct positive qualities an argument might have: (1) validity, (2) soundness, and (3) having a true conclusion. (2) logically implies both (1) and (3), and the other relationships are all logical independence. Now you know about soundness and validi-TEE. Next time won't you blog with me.

No comments:

Post a Comment