S15. Seeing the Neural Footprints of Prayer with fMRI

4 hours ago
9 min read

Have you ever had the feeling, while praying or silently repeating something inside, that the world around you slowly becomes quieter, as if the sounds are still there but somehow further away?

Scientists have actually put this invisible inner experience into an MRI scanner, slice by slice, to see what kind of traces it leaves in the brain. This study specifically looked at how the brains of Sahaja Yoga practitioners “switch mode” when they are praying, moving into a more inwardly focused and concentrated state.

Who went into the scanner?

The study invited 16 experienced Sahaja Yoga meditators to take part, including 10 women, aged between 24 and 63, with an average age of about 48 years. Each participant had more than 7 years of daily meditation practice, with an average Sahaja Yoga experience of around 20 years, and some had been practicing for over 30 years. In their daily lives, they spent roughly 20 to 120 minutes meditating per day, with an average of about 47 minutes, and they frequently reported experiencing a state of “no extra thoughts, yet fully awake and present.”

In terms of religious belief, they showed very strong faith in God (everyone rated this as 5 out of 5) and a very high level of trust that God responds to their prayers (on average about 4.8 out of 5). By contrast, when asked “Do you believe Santa Claus exists?”, the average rating was only about 2.1, clearly showing they distinguish between “the Divine” and a fictional character.

They were also very familiar with the prayers commonly used in Sahaja Yoga:

They recited the Lord’s Prayer on average about 4.5 times per week.
They prayed to the Divine Mother even more often, about 5.7 times per week on average.

In other words, the researchers did not randomly pick a few people off the street to “experiment with prayer”; they worked with serious, long‑term practitioners who regularly pray in this way.

What did they do inside the scanner?

The researchers used a 3‑Tesla functional MRI scanner (fMRI), which measures changes in blood‑oxygen levels (BOLD signal) to indirectly reflect how active different parts of the brain are during certain tasks. For each participant, 420 whole‑brain images were collected while they lay in the scanner with their eyes closed, to avoid visual distractions.

Before the main tasks started, there was about 7 minutes of structural scanning, during which the participants were asked to enter a meditative state. This helped them get used to the noise and setting of the scanner and also simulated the way they would normally be praying during meditation in their regular practice. Then the actual experiment began—throughout the entire run, everything had to be done “silently in the mind”, using only inner speech and no vocal sound.

The experiment included five conditions. Each condition lasted 26 seconds per block and was repeated 6 times, woven together into a roughly 14‑minute run. Before each block, the participants heard a brief 2‑second audio cue (for example, “Lord’s Prayer”, “Poem”, “Prayer to Divine Mother”), indicating which inner task they should perform next.

The five tasks were:

Formal religious prayer (Father)Reciting the Christian Lord’s Prayer silently in the mind. This is a highly structured, memorized prayer.
Improvised religious prayer (Mother)Silently offering a spontaneous prayer to the “Divine Mother”, speaking from the heart in their own words, asking, thanking, or sharing—without reading or repeating a fixed text.
Formal secular speech (Poem)Silently reciting a well‑known poem. In form, this is very similar to the Lord’s Prayer: rhythmic, fixed wording, something you recite from memory.
Improvised secular speech (Santa)Silently making spontaneous wishes to “Santa Claus” in their own words, saying what they would like to receive or see happen—but directed at a figure they do not actually believe exists.
Control condition: counting backwardsSilently counting backwards from 100. This is a purely cognitive, non‑emotional task used as a basic comparison, instead of “doing nothing”. A simple rest condition could itself generate a lot of spontaneous inner thoughts, which would complicate the analysis.

The study design followed a neat “2 × 2” structure:

Domain: religious vs. secular
Speech act: formal (memorized) vs. improvised

So the four main conditions can be laid out like this:

Formal religious: Lord’s Prayer (Father)
Improvised religious: prayer to Divine Mother (Mother)
Formal secular: reciting a poem (Poem)
Improvised secular: wishes to Santa Claus (Santa)

The clever part is this:The researchers did not simply compare “prayer vs. doing nothing”. Instead, they compared prayer with secular language tasks that look very similar on the surface, in order to separate what is “special about prayer” from what is just “using inner speech”.

Before looking at brains, meet a few key regions

If you imagine the brain as a city, this study is particularly focused on a few “landmarks”:

Thalamus – the sensory relay hub

The thalamus is like the brain’s main relay station. Most sensory information from the eyes, ears, and skin passes through here before being sent to the appropriate cortical areas. When the thalamus is highly active, the brain is busy processing signals from the outside world; when its activity drops, we can think of this as turning down the “gain” on external input.

The medial prefrontal cortex (mPFC) is involved in many “self‑and‑others” functions, such as:

Imagining what others are thinking
Evaluating our own behavior and social image
Moral emotions and social judgement
Integrating feelings of elevation or inspiration

mPFC is a key region in many studies on religious experiences and social cognition.

Dorsolateral and ventrolateral prefrontal cortices (dlPFC, vlPFC)

These areas are often grouped under “executive control”:

Focusing attention
Holding and manipulating information in working memory
Supporting language production and behavioral control

When we speak spontaneously, organize complex sentences, or deliberately steer our thinking, these regions are particularly active.

Finding 1: The brain’s “external volume knob” is turned down

When the researchers compared the two prayer conditions (Lord’s Prayer and improvised prayer to the Divine Mother) with their secular counterparts (poem recitation and wishes to Santa), one striking effect appeared: activity in both thalami was significantly reduced during prayer.

In other words, whether the participants were reciting the Lord’s Prayer or praying freely to the Divine Mother, as long as they were in a religious prayer mode, this sensory relay hub became quieter than in the comparable secular language tasks.

In experiential terms, you could say:

It is as if, during prayer, the practitioners’ brains turn down the volume of the outside world, pulling mental resources away from external distractions and back toward the inner space—not in a sleepy way, but as a focused form of not being pulled around by what’s happening outside.

Interestingly, this echoes findings from studies on other types of meditation—such as yoga, transcendental meditation, and mindfulness—where decreased thalamic activity has been linked to reduced pain perception and reduced responsiveness to external stimuli. In this study, we see that for experienced Sahaja Yoga meditators, prayer itself already begins to activate this “turn down the external noise” neural pattern, like a preparatory stage at the brain level before moving into deeper meditation states.

In daily life, our frontal lobes are almost constantly at work: thinking through to‑do lists, worrying about how others see us, choosing the right words, weighing risks and consequences.

But in these Sahaja Yoga practitioners, the frontal lobe story during prayer looked quite different: many frontal regions became quieter.

Prayer vs. secular speech: frontal deactivation

When the researchers compared the two kinds of prayer with their parallel secular language tasks, they found lower activity during prayer in several regions, including:

Left dlPFC (linked to executive control and focused attention)
Left vlPFC (linked to language and cognitive control)
Supplementary motor area and parts of the dorsal anterior cingulate cortex (linked to initiating and monitoring actions)

This is intriguing. Intuitively, one might assume prayer is a “high‑effort” mental task that would make the frontal lobes work harder. Yet here, prayer actually made these “control and management” regions less active than secular language tasks.

The researchers’ interpretation is that this style of prayer involves a posture of surrender. The practitioners are not laboriously “constructing” each sentence or tightly controlling their inner monologue. Instead, the language of prayer flows more naturally, and the frontal lobes loosen their grip.

mPFC: prayer and formal recitation both quiet it down

When they looked more closely at the mPFC, they found a very interesting pattern:

In both prayer conditions (Lord’s Prayer and improvised prayer to Divine Mother), mPFC activity decreased.
In the formal secular condition (reciting the poem), mPFC activity also decreased.
Only in the improvised secular condition (wishes to Santa Claus) did mPFC activity clearly increase.

This combination is quite revealing. Improvised wishes to Santa are emotionally and cognitively similar to a playful “social imagination game”: we know Santa isn’t real, but we imagine talking to a character, picturing his reactions and our own image in front of him. That naturally recruits the social cognition and self‑evaluation functions of the mPFC.

In contrast, during religious prayer and formal recitation, the mPFC becomes quieter, as if letting go of “How do I look?” and “What does the other think of me?” The researchers suggest this reflects an inner attitude of genuine surrender and trust in the context of Sahaja Yoga prayer, rather than a kind of social performance or impression management.

Finding 3: Formal vs. improvised speech—left and right hemispheres share the work differently

The study also compared formal, memorized speech with improvised expression, regardless of whether the content was religious or secular, to see how the form of language itself shapes brain activity.

Improvised speech: left‑hemisphere language regions work harder

When the researchers contrasted “improvised speech” (Mother + Santa) with “formal recitation” (Father + Poem), improvised speech produced higher activity in regions such as:

Left vlPFC (including Broca’s area)
Left dlPFC
Supplementary motor area and dorsal anterior cingulate cortex
Frontal eye fields and premotor cortex

This fits nicely with what we know about spontaneous speaking: when people have to “speak on the fly”, the left‑hemisphere language network becomes more active, because it has to construct sentences, choose words, and adjust phrasing in real time.

Formal recitation: right frontal and parietal regions join in

Looking at the reverse contrast, “formal recitation” versus “improvised speech” showed relatively higher activity in:

Right dlPFC
Right inferior parietal cortex (including supramarginal gyrus and angular gyrus)

These areas are thought to be involved in retrieving and rehearsing well‑learned material—like pulling out a memorized text from long‑term storage, relying more on established patterns rather than creating new sentences every second.

The researchers propose an intriguing idea: formal recitations tend to recruit more right‑lateralized networks, while improvised language relies more on left‑hemisphere language areas. This is not yet a firm rule, but it opens an interesting path for future research into how the two hemispheres share the workload in religious language.

How does this differ from “typical” Christian prayer in the scanner?

This study also drew on earlier fMRI research with Christian participants using similar paradigms. Those studies often found that:

During personal, improvised prayer, regions such as mPFC, temporo‑parietal junction (TPJ), and precuneus—key areas for social cognition and imagining interactions with others—were strongly activated.

In that context, prayer looks, neurologically, very much like an intense interpersonal dialogue with an invisible “Other”.

In contrast, with the Sahaja Yoga practitioners in this study, the picture is different:

Frontal regions, including the mPFC, tended to deactivate during prayer rather than light up.
The thalamus clearly deactivated, suggesting decreased processing of external sensory input.

From a neural perspective, Sahaja Yoga prayer in this sample seems more like a shift away from external and social‑evaluative processing, and toward inner quietening and surrender. It is as if the brain is gradually withdrawing attention from the outside world and from self‑monitoring, making room for a deeper inner stillness.

This is not about ranking which type of prayer is “better”. Rather, it reminds us that different traditions may engage different neural mechanisms when people pray.

Limitations: not the final word, but highly thought‑provoking

As with any rigorous study, the authors openly discuss their limitations:

Small sample size – Only 16 participants, all of them highly experienced, long‑term practitioners. The findings may not directly apply to beginners or to the general population.
No pure rest condition – The baseline task was counting backwards, which is cognitively demanding. This might have reduced the contrast between conditions. Including a pure rest condition might reveal additional differences.
No real‑time performance measures – Because all tasks were done silently in the mind, the researchers could not monitor performance moment‑by‑moment during scanning. They had to rely on self‑report after the scan, where participants rated how well they felt they had performed each task (on average between “well” and “very well”).

Even with these limits, the study gives a clear and inspiring picture:

Prayer is not just a psychological or religious label; it has specific neural signatures in the brain.
In long‑term Sahaja Yoga practitioners, prayer seems particularly effective at turning down external sensory processing and relaxing frontal control systems, preparing space for a quieter inner state.

Prayer, seen as a “neural texture”

Putting someone into an fMRI scanner and asking them to pray in silence might sound clinical and cold. Yet the neural images that emerge tell a very human story:

A group of long‑term practitioners, bringing their trust and familiarity with sacred prayers into the noisy environment of a scanner, still manage to enter a quiet inner space.
Inside their brains, the relay hubs for outer sensory information gradually quieten, and the control‑and‑evaluation centers of the frontal lobes loosen their grip.

From the outside, they appear to be lying perfectly still in a narrow tube.From the inside, something subtle is unfolding:a turning down of sensory input, a letting go of control, a drawing of attention inward—a prayer state that these Sahaja Yoga practitioners know well, now rendered visible for the first time in detailed brain images.

For those who already practice meditation or prayer, these results may feel like putting familiar inner experiences into scientific language.For those who have never tried, it may serve as an intriguing invitation: the brain does, in fact, seem to “see” a difference when we pray—and your own experience might be the most meaningful experiment of all.

Further Reading:

Research Articles: Original Articles and Research Papers