Does Cognitive Dissonance Theory Explain Belief in Jesus’ Resurrection?

A reply to Martin Hartke’s YouTube video and blog

Nick Meader
13 min readJan 9


This is a response to Matthew Hartke’s blog article (1) on how cognitive dissonance theory (CDT) explains Christianity. My PhD is in social psychology — so I always enjoy engaging with articles combining psychology and Biblical studies.

Hartke’s article brought back memories. As a final-year psychology undergraduate, I agonised over becoming a Christian. Could psychological theories like CDT explain the disciples’ beliefs? And those of later Christians? It’s fun to tread those paths again.

Summary of CDT

Hartke starts off with an accurate summary of CDT:

The term “cognitive dissonance” refers to the state of mental discomfort or tension that people experience when their beliefs, values, or behaviors come into conflict with their experience of the world, or when they hold two ideas that are psychologically at odds with each other…Cognitive Dissonance Theory, or CDT, is a prediction about how people tend to respond to that state of discomfort, how we avoid mental conflict and try to reduce it when it occurs. (1)

Yet, some further elaboration will help. Vaidis and Bran (2) point out the need for consistent terminology when discussing CDT. It’s important to distinguish the different stages of the theory (see Figure 1):

  • a triggering situation where there is an inconsistency between beliefs and reality (triggering situation)
  • discomfort caused by this inconsistency (cognitive dissonance state)
  • an attempt to reduce discomfort (regulation).

A common issue is that writers applying CDT often focus on regulation strategies whilst neglecting earlier steps in the model (2). Hartke’s article shares this limitation.

Figure 1. Three Stages of Cognitive Dissonance

CDT and extreme inconsistency

Most social psychologists agree that CDT only operates where the inconsistency between the attitude and reality is extreme (3).

For example when I’m on a diet, if I decide to have a biscuit on occasion — this is unlikely to cause me dissonance. I may make a minor change like this without much discomfort. This falls within what psychologists call the latitude of acceptance.

But if I ate the whole packet of biscuits, I’d crossed the latitude of acceptance. I must find an explanation for why I’ve gone so far outside of my original intentions. I could say, “I went for a very long run” or “it’s Christmas!” CDT predicts I need to find a way to resolve this dissonance.

Attributional biases

As a post-Christian, Hartke is attributing causes for beliefs that differ from his own. The social psychological literature highlights the potential risks of attributional biases. The best known is the correspondence bias — where we propose ‘dispositional’ causes for others’ behaviour, even when clear ‘situational’ explanations are apparent.

A classic example is the study by Jones and Harris, (4) where American participants read out speeches written by their fellow students on Fidel Castro.

As you would expect, most participants concluded that pro-Castro students wrote positive speeches about Castro if they had a choice of what to write (dispositional attribution). However, when participants were informed there was a situational cause — that students were instructed what to write — they overwhelmingly continued to make dispositional attributions (positive speeches were written by pro-Castro students).

This pattern of results has been found in many studies across a range of groups. These tendencies are amplified when dealing with members of an outgroup (a group different to ours). Thomas Pettigrew termed this the ultimate attributional error:

…negative outgroup behaviour is dispositionally attributed, whereas positive outgroup behaviour is externally attributed or explained away so that we preserve our unfavourable outgroup image. (5)

“We are unique, they are all the same”

We need to consider one further theory. After two world wars, a generation of psychologists was trying to understand conflict between groups. Social identity theory and social categorisation theory emerged as some of the most influential explanations confirmed over many decades.

These theories predict that group differences are accentuated by the process of social categorisation. Ingroup members tend to be categorised according to positive characteristics. For example, atheists may emphasise their love of scepticism, science, and logic. These positive characteristics are often contrasted with the negative characteristics of outgroups (for example, unquestioning religious people who want to protect their beliefs from science).

A key characteristic of this categorisation process — particularly in the individualist West — is an emphasis on the homogeneity (similarity) of the outgroup.

Jesus’ disciples and the Millerites — “apocalyptic birds of a feather”?

According to Hartke, the paradigm for understanding Jesus’ disciples’ belief in his resurrection can be found in the Millerite response to the Great Disappointment (their leader predicted the end of the world in 1844) and other failed apocalyptic movements. Jesus’ disciples had to adapt their core beliefs to the reality of Jesus’ failure to bring the new heaven and new earth.

The prophesied event is reinterpreted in such a way that what was supposed to have been a visible, verifiable occurrence is seen to have been in reality an invisible, spiritual occurrence. The event occured as predicted, only on a spiritual level. (1)

The solution of failed prophecy for the Millerites follows this pattern. Hiram Edson — taking a walk in a cornfield — received the ‘insight from God’ that 1844 was the time for the cleansing of the heavenly sanctuary, not the earthly one. A spiritualising of Miller’s original prophecy. Ellen Harmon later had a series of similar visions.

Hartke argues this spiritualising of failed prophecy — to regulate the cognitive dissonance state — was exhibited by the early Christians and other similar movements. Figure 2 illustrates some of the similarities and differences between these two groups.

Figure 2 comparing Millerites and Jesus’ disciples

Getting the comparator right

Hartke argues the New Testament accounts of Jesus’ resurrection and ascension are a regulation strategy to reduce cognitive dissonance caused by Jesus’ death. He proposes a dispositional cause (cognitive dissonance) for the disciples’ belief.

Social psychological theories suggest that attributions are not always as neutral as we think. The risk of these biases is accentuated when we use a broad comparator—in Hartke’s case all apocalyptic movements in history, regardless of time or culture. Our tendency to think “we are unique, they are all the same” can unintentionally lead us to be selective in highlighting similarities and downplaying differences.

Being more specific about comparators helps to guard against these potential biases. There is good reason to think Messianic movements occurring at a similar time and place, are more suitable comparators with Jesus’ disciples than a group of 19th Century Americans. A more careful way of testing Hartke’s assumption is to examine:

  • if there were other Messiah candidates at a similar time in history where their leader died
  • if they adopted a similar strategy to Jesus’ disciples.

…many of the messianic movements between roughly 150 BC and AD 150 ended with the violent death of the founder. When this happened, there were two options open to any who escaped death: they could give up the movement, or they could find themselves another Messiah. (6)

These examples suggest that a situational cause for regulating ‘dissonance’ was common. Members abandoned their group when their previous beliefs conflicted with the reality of their founder’s death. We have to account for why the disciples didn’t “give up the movement, or find themselves another Messiah” like other groups. Below we assess the four key similarities between the Millerites and Jesus’ disciples highlighted by Hartke.

1. Resurrection and ascension — regulating response or triggering situation?

CDT posits that cognitive dissonance begins with a triggering event. For Matthew Hartke, Jesus’ disciples expected him to bring the Kingdom of God immediately — instead, Jesus was nailed to a cross. Their dreams were over. So far, we agree.

However, where we differ is that Christians consider that Jesus’ death, resurrection, and ascension triggered disciples to reconsider their beliefs. For Hartke, beliefs about Jesus’ resurrection and ascension are regulation strategies to reduce the discomfort created by the disappointment of his death.

I’m unaware of how he comes to this conclusion. Hartke seems to think this had to be the case: “And as Dissonance Theory shows, this wasn’t much of a choice.” He moves from ‘inconsistency’ between expectations and reality straight to ‘regulation’ of discomfort — what psychologists Vaidis and Bran warned against.

There’s a stage in between. We must first demonstrate the disciples were experiencing a state of extreme inconsistency between their beliefs and reality — exceeding their latitude of acceptance.

2. Now and not yet: suspicious safeguarding?

For Matthew Hartke, the New Testament teaching about the ‘now and not yet’ of God’s Kingdom (the new creation begins with Jesus’ resurrection but is completed when he returns) is suspicious. According to Hartke:

  • the primary function of this doctrine is to protect the disciples’ beliefs from being testable (dispositional attribution).
  • the responses of 1st-century Jews and 19th-century Americans are ‘very similar’ (“we are unique, they are all the same”?) — despite our knowledge that cognitive bias can often operate differently between cultures.

a) Millerites

The problem is, the grounds for belief change are quite different. For the Millerites, ‘new revelation’ comes through visions experienced by Hiram Edson and Ellen Harmon. The Millerite believers had no way to assess the validity of these visions. It’s a message received by these individuals that no one else has direct access to. Nor can we debate their consistency with ‘previous revelation’ — as their message is new. Hartke is right, this ‘new revelation’ is untestable.

b) Jesus’ disciples

How did Jesus’ disciples safeguard their beliefs?

…whenever they talk about the present reality of the kingdom, they talk about it in terms of Christ’s exaltation to the right hand of God in heaven, sending the Spirit, establishing a new covenant, and other largely nonempirical events. (1)

An obvious physical and empirical omission in Hartke’s list above is Jesus’ resurrection:

For he has set a day when he will judge the world with justice by the man he has appointed. He has given proof of this to everyone by raising him from the dead. (Acts 17:31, NIV)

Would this not provide a situational cause for their change in beliefs? Seems at least possible. Was it testable? To produce Jesus’ dead body would have been a straightforward empirical refutation of the disciples claim. Additionally, the New Testament consistently affirms that Jesus was seen by his disciples. Not as a ghostly apparition, but physically.

I’ve briefly summarised a Bayesian approach to evaluating the evidence for the resurrection, building on arguments from Richard Swinburne (7). A further difference, as we’ll see below, is that we can evaluate whether the disciples’ claims that Jesus was the Messiah are consistent — or within the latitude of acceptance — with the Hebrew Bible.

3. Reinterpreting prophecy

This is the most important aspect of Hartke’s argument. He provides two examples of the New Testament reappropriating passages from the Hebrew Bible to reduce cognitive dissonance.

a) ‘Now and not yet’ of the kingdom

The New Testament ‘now and not yet’ understanding of the Messianic kingdom was a Christian innovation. There are no indications that any other pre-Christian group expected this to happen.

Was this new belief, Jesus’ disciples’ response to something they didn’t expect — his resurrection and ascension? Did they search the Hebrew Bible to make sense of these unexpected events? Was their belief in the ‘now and not yet’ kingdom a justified response to their experience? If so, Jesus’ death, resurrection, and ascension were the situational causes of this belief that arose within Judaism — and later developed into Christianity.

Hartke, in contrast, proposes a dispositional cause for this new belief. Jesus’ disciples changed their beliefs to reduce the discomfort caused by their cognitive dissonance — and to maintain their commitment to the movement. How to decide between these explanations? We need to go back to the Hebrew Bible.

b) A suffering gentle servant or a victorious king?

Critics of Christianity, often emphasise the necessity that the Messiah must fulfil all prophecies at once. This was a common position within Jewish tradition — but far from universal. For millennia, there has been an understanding that the Hebrew Bible included two very different portraits of the Messiah — a humble suffering servant and a victorious king. I discuss this in more detail here.

How do we reconcile these two depictions? Jewish tradition has offered many alternative solutions including:

  • The Messiah will fulfil both of these roles (e.g. Pesikta Rabbati)
  • There will be two Messiahs: Mashiach ben Yosef (Messiah son of Joseph) the suffering servant and Mashiach ben David (Messiah son of David) the victorious king (e.g. Ibn Ezra and Radak considered Zechariah 12 and 13 to be about Mashiach ben Yosef)
  • Reinterpret passages on the Messiah to edit out his suffering (e.g. Targum of Jonathan on Isaiah 52:13–53:12)
  • The type of Messiah will depend on the obedience of the Jewish people — he will be gentle if they are obedient or fierce if they are not (Sanhedrin 98a)

Some of these solutions resolve the tension by downplaying or reinterpreting some passages whilst emphasising others — which is problematic. Others provide no solution to the tension between the two roles of the Messiah. Is it possible for the Messiah to both die and be a victorious king at the same time?

The disciples’ solution was that Messiah would return a second time. He comes first as a suffering servant (Psalm 22, Zechariah 12–13, Isaiah 52:13–53:12) then as a victorious king (Isaiah 66, Daniel 12). In my view, this explanation lies within the latitude of acceptance of what the Hebrew Bible taught. It makes sense of the tension between the ‘suffering servant’ and ‘victorious king’ passages.

c) Psalm 110

But what about the post-hoc ‘reinterpretations’ of passages that Hartke thinks the New Testament includes as a mechanism for overcoming cognitive dissonance?

We agree, Psalm 110 is ‘a promise of victory made to the Davidic dynasty’ (1). He is also right that a several passages in the New Testament cite this as a Messianic Psalm, where the Messiah sits at the right hand of the Father. Is that going too far?

Of course, one of the promises to the Davidic dynasty is that the Messiah will come from this line (2 Samuel 7). So it’s not unreasonable to conclude a Psalm applied to David points to the Messiah. A common inference in rabbinic and Christian literature. Additionally, some scholars have argued that the Psalms were redacted by the editor of the Psalter to suggest a Messianic interpretation of the royal Psalms (8).

There is clear evidence for the Messianic interpretation of Psalm 110 in the rabbinic literature from the second half of the third century (8). For example in the Talmud (Avot D’Rabbi Natan 34:4).

The Eternal has sworn and will not change His mind; you will be a priest forever, [the rightful king that I have chosen]” (Psalms 110:4). From this verse we know that the messianic king is even more beloved than a rightful priest.

There is debate among scholars whether Messianic interpretations of the Psalm existed in the first century. One potential source of evidence is 1 Enoch — probably written by Jewish mystics between 100 and 200 BCE.

The account of the son of man sitting on a throne in heaven is dependent on Psalm 110 and Daniel 7 (9). There are several overlaps with Jesus’ statement in Mark 14:62.

“And you will see the Son of Man sitting at the right hand of the Mighty One and coming on the clouds of heaven.” (NIV)

The evidence suggests the New Testament interpretation of Psalm 110 was far less radical than suggested by Hartke. We can see an evolution in the disciples’ views about the Messiah. But it remains within standard exegesis of the Psalm and coherent with other interpretations existing in the first century and beyond.

d) Jeremiah 31:33

Hartke’s second example is Jeremiah 31:33, the promise that God would make a new covenant with his people. Again, I don’t have much problem with his suggestion that this passage from Jeremiah is about the spiritual renewal of Israel.

However, rabbinic literature has also recognised the connection between this promise of the new covenant and the coming of the Messiah in this passage. Ramban’s (aka Nachmanides) famous commentary makes this precise point.

Christians make a similar claim. That Jeremiah 31:33 is promising a new covenant, a time of great spiritual renewal and the starting of the kingdom, at the coming of the Messiah. Fulfilling the original promise to Abraham that all peoples will be blessed through him (e.g. Genesis 12:3).

4. Admitting our mistakes is suspicious?

Fourth and finally, just as the Millerites justified all of these reinterpretive moves by concluding that they had simply misunderstood the true meaning of the prophecies, so too with the early Christians. (1)

Changing your mind isn’t necessarily evidence of cognitive dissonance. As the economist John Maynard Keynes is thought to have said: “When the facts change, I change my mind. What do you do?” (Although see here!) We must again be careful not to jump to dispositional attributions.

Hartke’s four comparisons between Jesus’ disciples and the Millerites have significant limitations.

  • Social psychological theories predict we commonly attribute dispositional causes to others’ beliefs. So it’s not surprising that Hartke finds cognitive dissonance a convincing explanation for the disciples beliefs.
  • Social psychology gives us reason to expect that disparate outgroups can appear unreasonably similar to us — we can unintentionally overemphasise overlaps between outgroups like 1st-century Jews and 19th-century Americans without spotting key differences.
  • Applying CDT as an explanatory framework for a group’s beliefs first requires evidence of a cognitive dissonance state. Hartke’s account focuses on the evolution of the disciples’ beliefs (‘regulation’) without sufficiently demonstrating that they were experiencing cognitive dissonance.


  1. Hartke M. An Unshakable Kingdom: How Cognitive Dissonance Explains Christianity.
  2. Vaidis DC, Bran A. Respectable Challenges to Respectable Theory: Cognitive Dissonance Theory Requires Conceptualization Clarification and Operational Tools. Frontiers in Psychology 2019; 10: 1189.
  3. Fazio RH, Zanna MP, Cooper J. Dissonance and self-perception: An integrative view of each theory’s proper domain of application. Journal of Experimental Social Psychology 1977; 13(5): 464–479.
  4. Jones EE, Harris VA. The Attribution of Attitudes. Journal of Experimental Social Psychology 1967; 3: 1–24.
  5. Hogg GM, Vaughn M. Social Psychology. Pearson Press.
  6. Wright NT. Resurrection of Son of God. Fortress Press.

7. Swinburne R. Resurrection of God Incarnate. Oxford University Press

8. Pao DW, Schnabel EJ. Luke. In: Beale G, Carson D, eds. Commentary on the New Testament Use of the Old Testament. Michigan: Baker Academic.

9. Boyarin D. Beyond Judaisms: Metaṭron and the Divine Polymorphy of Ancient Judaism. Journal for the Study of Judaism 2010; 41: 323–365.



Nick Meader

My background is in psychology, epidemiology and medical statistics. I’m mainly discussing here theology, philosophy of religion and mental health.