Dr Ben Holloway, Professor of Philosophy and History of Ideas, Southeastern Baptist Theological Seminary

There are three ways we use the term ‘knowledge.’ Sometimes we use it to describe procedural knowledge. When we use the term this way, we might be saying that someone knows how to ride a bike, speak French, or play a game. Knowing how to do something is different from knowing about it. One can become an expert on fishing without ever having cast a line. On the other hand, one can be very skilled at fishing while only knowing a minimal set of facts about it.

A second way we might use the term is to talk about experiential knowledge. When we use it this way, we are talking about being personally acquainted with someone. For example, we might say that Jim knows Bill.

The use we will be concerned with here is for propositional knowledge. Propositional knowledge is what we are talking about when we say that we know that grass is green, that Parmenides denied change was possible, or that Jesus is God.Someone riding a bicycle

What Counts as Knowledge?

According to a traditional view, knowledge is defined by answering the question, ‘what are the necessary and sufficient conditions for knowledge?’ The answer often given is threefold.

First, the proposition believed must be true. If, at 1pm, I believe that Chelsea will win the cup and, at 3pm, Chelsea does not win the cup, then I did not know that Chelsea will win the cup even though I believed it.[1]

Second, knowledge requires belief. It is true that God exists. But if someone doesn’t believe it, that person cannot know that God exists.

Finally, there must be some rational status for the belief. I can’t know anything by a lucky guess. Suppose without any reason, I guess at 1pm that Chelsea will win the cup. Since I know nothing about football nor what ‘cup’ I am talking about, it seems implausible that I could give a reason for my belief. I am merely guessing. If Chelsea does win the cup at 3pm, you would deny that I knew that Chelsea would win the cup a couple of hours earlier. You would put it down to luck. And you don’t get to know things by guessing.

What is lacking in the scenario is my having any reason to believe that Chelsea will win the cup. Having a good reason would give the belief a good rational status and may render it knowledge. Having a good reason is one way a person may be justified in his belief.

Although there are different views on what justifies a belief, having a good reason being one of them, justification seems to be a good candidate for our third condition. Thus, according to the traditional theory, you and I know something if we believe it, it’s true, and we are justified in believing it. More formally,

S knows p if and only if (i) S believes p (ii) p is true (iii) S is justified in believing p.

The analysis of knowledge isn’t without difficulty. In a famously brief paper, Edmund Gettier constructed examples in which all three conditions are satisfied but don’t count as knowledge. Here is my elaboration of one:

Suppose Smith and Jones have applied for the same job. Their boss’s secretary has reliably informed Smith that Jones has got the job. Disappointed, Smith retreats to the kitchen. Coincidently, Jones is in the kitchen emptying his pockets on the table. Smith counts ten coins on the table and concludes that the person who has got the job has ten coins in his pocket.

In due course, Smith is summoned to the boss’s office. He dutifully attends, expecting to hear that he hasn’t got the job. To his surprise, his boss offers it to him. Delighted, Smith returns home to tell his wife. On the way in, he empties his pockets on the side table. Coincidentally, he has ten coins in his pocket. So, he was right. The man who got the job has ten coins in his pocket!

To see the challenge to the traditional analysis, consider the proposition, the person who has got the job has ten coins in his pocket. Prior to being summoned to the boss’s office and after he had seen Jones in the kitchen, Smith has good reason to believe it. He has good reasons for believing that Jones has got the job and that Jones has ten coins in his pocket. It follows that the person who has got the job has ten coins in his pocket. It also turns out to be true. So, Smith believes it, it is true, and he is justified in believing it. But it is only a matter of luck that it turns out to be true, so we would want to refrain from saying that Jim knows it.

In response to the problem, many philosophers have modified the definition by adding a fourth condition or replacing the third one. Others have taken the inability to provide a precise definition of knowledge as a sign that we ought to abandon its analysis altogether.

One way to solve the problem is to stipulate that any reason for a belief must be a good reason. What counts as a good reason is that it is true. Since one of Smith’s reasons is false, namely, the belief that Jones has got the job, Smith doesn’t have a good reason for the resulting belief, until, of course, he sees that he himself has ten coins in his pocket. Others have suggested that we need a new version of justification more related to the reliability of our cognitive mechanisms than the relation between our evidence and beliefs.

Having difficulty with finding a precise definition should not be cause for alarm. Many concepts are difficult to analyse, but we continue to be able to use them adequately anyway.


Although, I will have more to say about justification in a future post, it is worth making some comments about it now. Justification has some interesting features. For example, one can have good reasons to believe a false proposition. How many times have we followed the directions of a stranger only to find ourselves hopelessly lost?

Justification is also relative to times and persons. Consider your former self. If you examined many of your beliefs, which have since changed, you might find that you had decent reasons for them at the time. Upon the discovery of evidence that pointed the other way and finding it sufficient, you changed your mind. Like your former self, people who believed that the earth was the centre of the universe were justified in believing it (they had good reason). They did what they could with the evidence available to them. Since Copernicus and the telescopes of Galileo, better reasons have since been given to think the sun is at the centre.

It is important to see that opposing views can both be justified. Consider friends or family members with whom we disagree. Like us, they may have amassed reasons for their view. Now, it may turn out that one side has better reasons than the other, but assuming that most people try to obtain the most reasonable beliefs allows us to engage in a much more fruitful discussion than one side assuming the other is entirely irrational.


Justification also comes in degrees. One can have such good reasons to believe a proposition that one has a very high degree of certainty. On the other hand, one can have a belief that is only weakly justified. Assessing the strength of our reasons is part of our job in managing our beliefs properly.

John Feinberg distinguishes between two kinds of certainty.[2] Objective certainty obtains between the reasons and the proposition. A proposition can have a high degree of objective certainty even if no one believes it. Subjective certainty is the degree of psychological commitment a person has to a belief.

Although, in general, it is good to proportion our subjective certainty to the objective certainty of a belief, subjective certainty isn’t only a rational matter. For example, people often have treasured beliefs that they are reluctant to relinquish in the face of sufficient or even overwhelming evidence to the contrary.

Sometimes this phenomenon occurs because of psychological factors unrelated to the merits of the case. Perhaps an atheist’s previous experiences make it very difficult to accept that God loves them. Overcoming these psychological factors is not merely achieved by providing more evidence or better reasons. They may be more practical in nature.

Even when good reasons have been given for our Christian beliefs, there are psychological factors which influence our strength of commitment to them. In his letter to the Ephesians, Paul gives multiple reasons for his claim that Gentile inclusion in the church is part of the plan of God. However, he knows that no matter how good his arguments are, something else is required. Consequently, in two places in the letter, Paul breaks into prayer, asking the Lord to illuminate the mind and strengthen the beliefs of the readers (1:17-19; 3:14-19). Paul recognises that while objective certainty comes from the strength of his reasons (inspired by the Holy Spirit), subjective certainty will have to be aided by the work of the Holy Spirit in the mind of the reader.

Though we should recognise the psychological aspects of belief, we shouldn’t reduce our ministries to moving people psychologically through showing the therapeutic benefits of faith or telling powerful stories. We should also try to show that there are good reasons to believe what we believe is true (1 Peter 3:15).

For an excellent discussion of the difference between objective and subjective certainty, I suggest reading John Feinberg’s Can I Believe it’s True? (Crossway, 2013) from which I drew much of the above material. Epistemology is a core discipline in philosophy and there are some very readable introductions available. I use Richard Feldman’s Epistemology (Pearson, 2002) in my classes at Southeastern, but there are many others. Another good introduction is Noah Lemos’ An Introduction to the Theory of Knowledge (Cambridge University Press, 2007) from where I got the football example for justification. Edmund Gettier’s famous paper, “Is Justified True Belief Knowledge?” is short and worth a read (Analysis, vol. 23, no. 6, June 1963, pp. 121–123).



[1] I am indebted to Noah Lemos for the football example. Noah Lemos, An Introduction to the Theory of Knowledge (New York: Cambridge University Press, 2007), 6.

[2] John Feinberg, Can You Believe It’s True? (Wheaton: Crossway, 2013), 160–168.