Science · · 3 min read
Study maps how the brain rebuilds knowledge from scrambled lessons
Research identifies two stages of brain activity involved in turning temporally disordered learning materials into an organised understanding of complex physics.
A study published by Springer Nature has identified a two-stage pattern of brain activity that helps people assemble coherent knowledge from lessons presented out of order.
The research examined how learners responded to video lectures on complex physics concepts when the material was temporally scrambled. Two separate groups took part: one was assessed through behavioural testing, while the other learned during functional magnetic resonance imaging (fMRI). Across both groups, the findings indicated that learners did not immediately reconstruct the organised structure of the subject. Instead, they first needed to gather enough information before their understanding began to resemble the structure defined by experts.
The results, reported in Communications Biology, point to different roles for two brain regions. The posterior cingulate cortex became more closely aligned with the developing organisation of the knowledge while learners were revising their understanding. The precuneus, by contrast, maintained a stronger relationship with that structure after the reorganisation had taken place.
Building order from disruption
Learning normally allows information to accumulate in a sequence that can support a larger explanation. In this study, however, the videos disrupted that sequence. Participants therefore had to infer how separate pieces belonged together rather than receiving the concepts in an already organised progression.
The researchers found that this reconstruction was not simply a matter of matching each new fragment to a stable mental framework. A learner’s internal representation had to change as additional information arrived. The results suggest that people can adapt to disordered input, but that the process depends on reaching a point where enough evidence has been collected to support a more coherent arrangement.
The posterior cingulate cortex was particularly associated with this updating period. Its stronger correspondence with the structure of the subject during learning suggests a role in bringing incoming information into line with an emerging understanding. The precuneus showed a different pattern, with its relationship to the knowledge structure lasting beyond the main reorganisation phase.
Together, the findings describe a delayed progression rather than an instant recovery of order. First, the learner accumulates and revises information. Later, the reorganised understanding is represented in a more sustained way.
Immediate learning and later limits
Communication between the two regions also mattered. Functional coupling between the posterior cingulate cortex and the precuneus predicted how successfully participants learned from the scrambled material in the immediate assessment.
That relationship did not completely predict strong retention at a later point. In other words, a learner could perform well soon after encountering the disordered lesson without necessarily preserving the reconstructed knowledge as robustly over time. The distinction gives the findings an important qualification: the brain may be capable of making sense of fragmented material, but successful short-term reconstruction is not identical to durable learning.
The study therefore presents disordered instruction as having two sides. It reveals the brain’s capacity to reorganise incomplete or poorly sequenced information, while also indicating a possible cost that becomes visible when knowledge must be retained after the initial learning episode.
The work was carried out by researchers affiliated with the State Key Laboratory of Cognitive Neuroscience and Learning and the IDG/McGovern Institute for Brain Research at Beijing Normal University, along with the Institute of Brain and Psychological Sciences at Sichuan Normal University. The paper was authored by Xu, He, Zhou and colleagues and published in 2026.
The research received support from the National Natural Science Foundation of China under grants 62293550 and 62293551. The article is available through Springer Nature under a Creative Commons Attribution-NonCommercial-NoDerivatives licence, which allows non-commercial sharing with attribution but does not permit distribution of adapted versions.