Aug 14

The Matrix Objection – Introduction

Now that my workload has lessened a bit and I’ve finally caught up with George R. R. Martin in A Song of Ice and Fire, I should have more time to get back to writing. I’d also like to conduct a number of book reviews, including for Richard Carrier’s latest book on Bayes Theorem and the historical Jesus. So, look for more of that in the near future. Up first is a short introduction to what I call The Matrix Objection, which seems to come about in discussions of using internal evidence as warrant for belief in God (or anything else, really).

Internal evidence is often cited by Christian philosophers, such as Alvin Plantinga’s sensus divinitatis or William Lane Craig’s internal witness of the Holy Spirit. They think this evidence, which is available only to them and cannot be tested or verified by anyone else, is a good reason for belief in God. This internal evidence is also irrefutable, so it cannot be disputed by any external finding or evidence. If you discovered something in the world that contradicted this pre-existing internal belief, like evolution perhaps, then that thing you discovered must be mistaken because your internal sense of God’s truth could never possibly be mistaken.

In response to such ideas, John Loftus posted this quote on his blog:

Self-authenticating private evidence is useless, because it is indistinguishable from the illusion of it.”

I think this quote provides a strong first level response, but our defense of this argument must be able to overcome what I’ll call The Matrix Objection (TMO).

TMO: Real life would be indistinguishable from life in The Matrix, but you believe your life is real based on what is self-evident.

TMO argues that since we use private evidence in some cases, the Christian should also be entitled to use it. It’s a technique to widen the gates of what we consider acceptable evidence.

I will argue, however, that TMO is ad hoc, while the initial illusion objection is not. If I am correct, then these are not analogous cases and TMO is not very successful as a response to the initial quote.

The main difference is TMO—and similar objections, like Plantinga’s argument from other minds—points to merely possible illusory alternatives, while the argument quoted by Loftus points to a plausible alternative. To get a grasp for the difference in force between possible and plausible reasons, consider the following scenario:

You are a football coach trying to decide the right play to run at a crucial moment in the game. You hold a brief conference with two of your assistants. The first assistant recommends a play action pass. His reasoning is that you’ve been running the ball successfully and the defense should “bite” on the fake. The second assistant recommends a running play to the left because aliens might attack the right side of the field.

Now, even if you don’t know much about football, you probably recognize a problem with the second assistant’s reasoning. Which assistant’s advice would you follow? The argument from the first assistant takes available evidence and makes an inference based on probability. If one cared to research football statistics, you would see that it is correct to infer that success in running the football correlates with later success using the play action pass. On the other hand, how do we assign any probability to the wild scenario proposed by the second coach? It is merely an appeal to possibility without anything in favor of its plausibility.

For the same reasons that any good head coach would take the advice of the first assistant over the second, we should also heed the possibility of illusion as Loftus means it as more likely than illusion as TMO means it. So far, I think the reasoning for this should just strike most of you as intuitively correct and even obvious. Next time, though, I’ll dig into the probability a bit more and attempt some actual estimates to be sure that our intuitive estimation is correct and provide a fuller defense against TMO.

Similar Posts:


Skip to comment form

  1. Ryan

    Here’s another:

    Premise 1: Members of different religions claim to have divine knowledge.

    Premise 2: These religions are mutually exclusive.

    Conclusion 1: Not all supposedly divine knowledge is correct.

    Premise 3: There is no method to determine which set of supposedly divine knowledge is correct.

    Conclusion 2: Supposedly divine knowledge is insufficient evidence for any religion.

    I assume premise 3 because, if such a method did exist, then we would not have to waste time on the argument “I know it in my heart to be true; therefore, it is true.” We would instead focus on the method and its conclusions.

    But I’m not sure that arguing against Plantinga and Craig here is worthwhile. That they can hold such a ridiculous belief and feel comfortable with telling others about it strongly suggests that they would not be open to any argument against it. With such a failsafe built into their faith, we can’t make any progress. If I thought that atheists might be fooled by it, I might change my mind, but I have not seen any evidence that “I just know” has compelled reasonable people to change their minds.

    Moreover, if a theist perceives this “divine knowledge” to be the basis of his faith, then the rest of his arguments are distractions. I hate to deal with people who do not argue what they believe for the actual reasons for which they believe it. It is a waste of my time because they are not really invested in the debate.

  2. Mike

    Hey, Ryan. I agree I won’t be convincing the likes of Plantinga or Craig considering their belief about their special sense’s infallibility. But we can still demonstrate to fellow atheists and believers who are more open to external evidence just what is wrong with this argument.

    Also, I’m inclined to think maybe we could test for which is right (premise 3). I think a consistent showing of any sort of special knowledge (i.e., above and beyond a control group) might at least show that it’s more probable that a group has special internal knowledge. So, I’d say they actually fail the only kinds of tests we can imagine. But if we were being charitable, your premise as stated should still work.

  3. Ryan

    That requires that the theist’s divine knowledge be falsifiable. If the issue is whether or not someone can “just know” that his deity exists and has a particular set of desires, then we can’t really test it. If the issue were whether or not someone has knowledge about the physical world that he should not have, then we could test the accuracy of his knowledge, but it would still be unreasonable to conclude that the knowledge comes from a deity and that his religious beliefs are all true.

  4. Ab3

    Hey Mike, how would you determine that it is more plausible that we live in a world of external objects as opposed to the matrix? Isn’t it the philosophical “rub” of these questions that they come with equiprobable prior probabilities (50/50) and consequent probabilities both equal 1? How does one grab a probabilistic foothold for any given hypothesis in such a scenario?

  5. Mike

    In this case, the objection is coming from a position of acceptance that something being self evident counts toward plausibility. The person in Plantinga’s position not only wants what is normally considered self evident (like direct observation) to count as evidence, they want even more to come under that umbrella. So, we’re not defending ourselves against a true skeptic. We are starting with the assumption that some things being self evident count as good evidence. With that starting point, we want to explain how we can reject TMO without widening the gates for any claimed internally verified knowledge. Since we are starting with these assumptions, it allows us to use tools like inference.

    It’s true that there is no overcoming the argument of a true skeptic without begging the question. Luckily, that’s not the opponent here.

  6. mdejess

    Bdfore anything else the author must talk about what is evidence, what is the target of evidence, and how evidence hits its target.

    Otherwise there is nothing about evidence that can be relied upon since one man’s evidence is another man’s non-evidence.

    Before anything else, people who bank on evidence must work on first coming to concurrence with all interested parties on what is evidence etc.


  7. Mike


    I’ll be getting into that a bit more on my next post, but that’s what the discussion is already about at a less foundational level. Plantinga, for example, agrees that certain what you might call direct observations do count as evidence. Further, he’ll say, since that counts as evidence, there are other similar ways of knowing available that should also count.

    Now, I’ve laid out the introduction to an objection that says Plantinga’s addition is not that similar to other basic sorts of knowledge. However, if that objection is to be successful, we’ll need to be able to say more clearly why we might think those are different. That’s what my next post will discuss.

    I’m already starting a level or two into the argument, so I’m not going to start here with the basics of epistemology and everything covered by Plantinga in Warrant and Proper Function. That would take too long and I think it’s helpful to discuss second and third level arguments in addition to more basic points.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

Notify me of followup comments via e-mail. You can also subscribe without commenting.