On the referential capacity of language models: An internalist rejoinder to Mandelkern & Linzen. (2026). Computational Linguistics, 52(3). https://cljournal.org/index.php/cljournal/article/view/3097