- First thing: go over the Generating Philosophy ideas in [this ChatGPT conversation](https://chatgpt.com/g/g-p-6863cdae32988191b6ef486a436ee4e5-nick/c/6a2558d7-d9c4-83ed-a3b2-d4565b7e8bf5). https://chatgpt.com/g/g-p-6863cdae32988191b6ef486a436ee4e5-nick/c/6a26f5ae-2a00-83eb-8582-9688ba815b4d ## report on hearing meanings paper 8 Jun 2026 The revised manuscript is markedly improved. The author has corrected the earlier framing of the debate as an introspective stand-off, has acknowledged the continuity with context-sensitivity arguments, has made Brogaard's own cross-linguistic observation explicit, and has separated the systematic commitment in ordinary bilingual overlap cases from the pervasive commitment in the idealised A/B case. The paper is now clearer and fairer to the existing literature. Two of the author's responses seem to me fully satisfactory. First, the paper now acknowledges the shared starting point with Brogaard's own cross-linguistic observation. The revised manuscript notes that "Brogaard (2018) also observes that two different languages may share the same speech sounds while nonetheless associating different meanings with those sounds," and asks a different question of it: "what follows once that observation is placed within a more specific bilingual model of online lexical access and selection?" This explains why the author and Brogaard draw different conclusions from a similar starting point. Second, the ambiguity between systematic and pervasive cognitive penetration has been resolved. The manuscript now states that "the present argument does not require the stronger claim that penetration is continuously present throughout bilingual experience," and that "the core claim is a systematic one," with pervasive penetration arising only in "the idealized maximally overlapping A/B limit case." The ordinary bilingual commitment is now described as systematic rather than pervasive. I have two remaining points. 1. The target should be specified more clearly: linguistic meaning, utterance meaning, or speaker meaning. In my previous report I said that, in addition to appealing to introspection, Brogaard distinguishes linguistic meaning from utterance meaning. The revised manuscript improves the surrounding discussion, but this distinction is still not addressed. The paper now ties its argument more closely to context-sensitivity objections. In §3 the author writes that those objections show that "utterance understanding is systematically sensitive to background information about speaker intentions, discourse setting, and world knowledge." This leaves it unsettled whether the target is utterance meaning, speaker meaning, or sentence-level linguistic meaning. Elsewhere the manuscript targets sentence meaning. The abstract says that "distinct sentence meanings can themselves figure in the phenomenal character of speech perception," and §2 says that the strong view holds that "distinct sentence meanings themselves can make a difference to auditory phenomenal character." The author should make the target explicit. If the argument is aimed at sentence-level linguistic meaning, as the Lena and Amy case indicates, the paper should say so, and should state how that target stands to the context-sensitivity literature, much of which concerns richer forms of utterance understanding. A defender of Brogaard can otherwise reply that the inferential, evidence-sensitive recovery of utterance meaning is compatible with a non-inferential perceptual account of linguistic meaning, and that the context-sensitivity material therefore leaves the perceptual experience of linguistic meaning untouched. This reply does not defeat the paper, but the author should pre-empt it by saying which kind of meaning the argument concerns, and by claiming that, in bilingual overlap cases, the selection even of linguistic sentence meaning depends on language-identity cues, beliefs, and expectations. 2. The reply to the attention/routine response should be sharpened. In §4.2 the author states that a defender of Brogaard might hold that "fluent bilingualism installs two perceptually learned routines, one associated with each language," and that contextual cues "merely bias attention toward the appropriate routine," so that the case involves "non-penetrating guidance of already learned perceptual processing" rather than cognitive penetration. I think this is a good addition. However, the reply needs a further step, because the charge that the attention/routine response "merely redescribes the selection problem" is too quick. The author writes that, if sentence meaning figures in auditory phenomenal character, then "language-identity information is not merely facilitating processing upstream" but "helps determine the phenomenal outcome itself." The trouble is that the defender grants exactly this. On the attention/routine reply, attention selects which learned routine is operative, and so does help determine the outcome, but without belief thereby penetrating perception. The author's claim that this redescribes rather than answers the problem therefore assumes what is in dispute, namely that the relevant selection is penetration rather than non-penetrating attentional modulation. The author should say why belief-guided selection between learned language routines satisfies Brogaard and Gatzia's restrictive criterion, instead of falling under the non-penetrating top-down modulation that Brogaard allows. If Amy is told that the language is B rather than A, "she will revise her belief about which language is being spoken, and her auditory experience will shift." This should be connected to the attention/routine reply. The open question is whether the evidence-sensitive shift occurs within perceptual experience, or whether it can still be understood as evidence-sensitive selection of a learned routine prior to perceptual processing. A full processing model is not required, but the text should say why the evidence-sensitivity of the shift excludes the non-penetrating attentional explanation. These two points are connected. If the target is linguistic sentence meaning, rather than utterance meaning or speaker meaning, the author should say so explicitly. But this clarification also affects the reply to the attention/routine objection. The author's claim would then be that even the selection of linguistic sentence meaning in bilingual overlap cases is belief-sensitive in a way that reaches auditory phenomenology itself. A Brogaard proponent may deny this, arguing instead that beliefs about language identity merely select, by attention or task-set, which learned language routine runs. The paper should therefore explain why the relevant belief-guided selection counts as cognitive penetration, rather than as non-penetrating selection among learned routines. This could be done by adding a brief clarification to §2, where the strong hearing-meanings view is defined in terms of sentence meanings, and then drawing on that clarification in §4.2, where the attention/routine objection is discussed. There is some weirdness with this reference which should be fixed: The manuscript cites Gasparri and Murez as 2019 in the introduction and as 2021 in §2 and §3, and the reference list gives "Gasparri, L., & Murez, M. (2019). Hearing Meanings: The Revenge of Context. Synthese, 198(6), 5229–5252."