This is a paper on the kirtana, a musical genre of Bengal and the author of this paper explored his distant past with the narrative of his subjective experience of musical exposure in the context of the cultural explosion of industrial noise, unintended sounds or non-discursive sonority of his surroundings. The author started with his understanding of kirtana as bhana, which, according to Bharata, the initiator-commentator of bharatiya natyasastra, is an individual performance of an actor, who at a time plays many roles as a self and as many others. Secondly, the author elaborated the problem of folk-classical songs dichotomy that, according to the author, is socio-politically motivated. Thirdly, he tried to understand the codes of kirtana as well as code switching in the discourse of kirtana in reference to non-formal linguistics without being bothered about the fragmentation of externalized arbitrary signs or atomic levels of linguistic analysis. An n-glossic situation was observed in the discursive formation of kirtana. This code-analysis ended with a discussion on the difference between speaking and musicking. One of the focuses of this paper is the particular part of the kirtana: akhar, which is betwixt or in between speaking and musicking. There are several steps in the kirtana: speaking, musicking, dialoguing, rhythmic gaps, well-constructed pauses or silencemes, simultaneous dancing, acting etc. and akhar is at a time an insider and an outsider. Thus, akhar is a liminal or threshold point of the song, which is characterized by ambiguity, openness, and indeterminacy. Moreover, the complicated role-playing of single interlocutors had made the author think about the preparation of a kirtaniya in the rehearsal room. Though Volosinov (1986) found this type of multi-layered performance by a single reader/performer is difficult in the context of Russian narratives, the Bengali kirtaniyas showed the path by performing such difficult text with professional precision. The reporting of the reported speech in the bhana of kirtana had become quasi-direct discourse with the full non-authoritarian participation of the three: composer, performer and the audience. If linguistics is considered to be a “discipline” for establishing dialogue without manipulation, the performance of kirtana as an open text might be cited as an example of such dialogue. And lastly, the author raised a problematic question regarding the use of kirtana in the parody-songs: how did such performance had become the vehicle of parody in Bengali culture of the 20th C.?