This paper examines honorific mismatches between the allocutive marker and a 2P pronoun within a clause in Japanese. Unlike allocutive languages such as Tamil, Punjabi, Magahi, among others, which require obligatory matching in honorificity (and phi-features) between the allocutive marker and a 2P pronoun in a clause, Japanese allows the honorific allocutive marker to also occur with a non-honorific 2P pronoun. These mismatching structures can be interpreted both literally and sarcastically. We provide a two-step analysis to explain such mismatches: first, we claim that 2P pronouns in Japanese are lexical items, which cannot participate in syntactic binding to inherit their honorific features from a left-peripheral honorific head (e.g., c). Instead, they enter syntax with pre-specified honorific information. Allocutive markers, by contrast, obtain their honorificity in syntax, typically via agreement. Consequently, nothing in syntax forces matching between 2P pronouns and allocutive markers. Secondly, mismatching instances are also not ruled out in the pragmatic component because the meanings encoded by the 2P pronouns and the allocutive marker in Japanese are not comparable, and contribute independently to the meaning of the entire sentence. The honorific information encoded in pronouns is the speaker’s positive/negative attitude towards the addressee based on his/her evaluation of the addressee, while the allocutive marker encodes the speaker’s intent to be polite, irrespective of the evaluation of the addressee, who may even remain unidentified. It is possible for these two meanings to be expressed simultaneously in select conversational set-ups, leading to felicitous mismatching structures in the language.