Exploring the experiences of peer counsellors providing remote behavioural activation therapy for antenatal depression: a qualitative study.

What makes a novice into an expert? This longstanding question sits at the intersection of neuroscience, cognitive science, and artificial intelligence. Across diverse domains such as mathematics, language, motor skills, and board games, extensive experience equips individuals with domain-specific knowledge, which serves as the foundation for fast, accurate, and flexible decision-making. Decades of cognitive research have established how such knowledge supports expert performance at the behavioural and computational levels. Yet, we still lack a precise account of its neural basis: specifically, information expert representations encode (the ), that information is organised for efficient readout (the ), and in the brain such representations are expressed (the neural ). Research on expertise has revealed that expert performance rests much more on knowledge than on raw general ability. Years of focused exposure tune perception and memory to a domain’s recurring structure. Co-occurring features are grouped into chunks and richer templates that bind surface cues to deeper relations and typical courses of action. When familiar constellations appear, they trigger these stored structures, which rapidly unpack the relevant details and predictions, functioning like an extension of working memory. This enables experts to quickly grasp a situation’s underlying structure while simultaneously preparing appropriate responses. Novices, in contrast, organise by appearances and get stuck on particulars. Expert cognition is therefore qualitatively different: recognition-driven and relational, powered by knowledge that supports both a quick recognition and fine-grained analysis when needed. Neuroimaging studies confirm that expertise involves a functional reorganisation of neural resources toward domain-specific and memory-based processing, which enables access to knowledge for fast, accurate decisions. Across domains, these systems are engaged more strongly (and often bilaterally) in experts than in novices. The precise loci vary with stimulus type, mapping where domain knowledge is stored and read out. In chess, experts preferentially recruit lateral temporal and parietal regions for object identities and functions but invoke scene/navigation systems for pattern recognition. Radiological expertise, with its holistic processing at its core, is characterised by ventral occipito-temporal selectivity. In higher-level calculation, professional mathematicians engage a bilateral frontoparietal-temporal network, while mental calculators draw on episodic memory components in the temporal lobe. When computation is externalised, as with the abacus, experts rely on visuospatial and premotor simulation circuits related to instrument manipulation. Finally, in motor expertise, a premotor-parietal action-observation network (AON) allows for efficient anticipation from kinematic cues.