Aphantasia as a functional disconnection

One of the quiet frustrations of the first ten years of aphantasia research has been that nobody could say with any confidence why an aphantasic brain does what it does. The visual cortex is there. The hippocampus is there. The frontal-parietal networks that ordinarily produce mental imagery are all intact. And yet, when you ask someone with aphantasia to picture an apple, nothing arrives.

In a 2025 review paper in Trends in Cognitive Sciences, Jianghao Liu and Paolo Bartolomeo propose the most coherent framework for this to date. Their argument, in plain terms: aphantasia is a functional disconnection. The imagery machinery is physically present; what is quieter than usual is the communication between the regions that would normally cooperate to produce imagery.

What they mean by functional disconnection

Ordinary mental imagery involves at least two groups of brain regions working together:

• Frontal-parietal control networks — the “executive” circuitry that decides you want to picture something and directs attention accordingly.
• Early visual cortex — the regions at the back of the brain that ordinarily process what you are seeing, and which also get recruited, in a top-down way, when you imagine something.

Liu and Bartolomeo argue that in aphantasia the second group is structurally unchanged; what is atypical is the first group’s ability to call on the second. Neither piece is broken. The two do not hand off to each other the way they do in a typical imager.

Their evidence is drawn from across a decade of fMRI, EEG, and behavioural studies — including the binocular-rivalry work of Keogh and Pearson, the hippocampal studies of Monzel and colleagues, and more recent 7-Tesla fMRI imaging. No single study proves the framework; its strength is that it is the one hypothesis consistent with all the observations at once.

Why this matters to a non-specialist reader

Three practical consequences:

1. Nothing is broken. If you have aphantasia, the imagery hardware is not missing. This is worth knowing, because the most common fear of someone newly identifying with the term is “something is wrong with my brain”. On the current best theory, nothing is — it is a difference in how the parts talk to each other, not in the parts themselves.
2. It explains the heterogeneity. People with aphantasia vary enormously in how they experience the condition — no visual imagery but vivid dreams, no voluntary imagery but occasional hypnagogic flashes, some senses affected but not others. A single-switch “imagery on or off” model cannot explain any of that. A connectivity model can: different pairs of regions may connect or fail to connect to different degrees in different people.
3. It does not (yet) imply a treatment. “Disconnection” sounds like the sort of thing that might be reconnectable — and it is tempting to read the paper as promising future interventions. Liu and Bartolomeo are careful not to. No manipulation has been shown to reliably produce imagery in someone who lacks it. Be suspicious of anyone selling you otherwise.

What this does not change

The paper is a review and a framework. It does not settle the genetic question. It does not settle the developmental question. It does not explain why some brains end up with this connectivity pattern and others do not. The authors are explicit that this is a direction for future research, not a conclusion.

For a reader of this site, the best way to hold this piece of research is as a confidence boost on a claim the cornerstone guide already makes: aphantasia is not damage, not disease, not deficit. It is a configuration.