LN regions (comprising 10s of voxels, not “tens of thousands”, as D. et al. state) and voxels show consistently high responses to language. Here is a sample LN voxel's response (in a sample brain) to 500 sentences (GretaTuckute; see the paper for a fROI-level profile): 10/n

Wrt concepts, many🧠areas show semantic tuning (as AlexHuth showed back in 2016). We recently found such tuning in single cells outside of the LN (pubmed.ncbi.nlm.nih.gov/38961302/). However, even semantically tuned LN voxels/neural populations are only driven by linguistic inputs. 6/n

1️⃣: Drijvers et al. argue that many parts of the brain respond to language. We agree. A large section of our NRN review is dedicated to non-language-selective networks that contribute to aspects of language processing. 3/n