This study with 54 participants (27 dyads) investigates whether a bilingual speaker’s language production is influenced by the comprehension of another speaker’s language. Dyads of bilinguals completed a “joint” picture naming task in which speaker turns were indicated by picture position (left/right). Speakers either engaged in monolingual language production, naming pictures exclusively in L1 or L2, or in bilingual production, switching between L1 and L2 within a single block. In monolingual production blocks, comprehension regarding the other speaker’s language was either monolingual (same language) or bilingual (the other language). In bilingual production blocks, language switched every two pictures, regardless of speaker turns. During monolingual production speakers showed longer naming latencies for L1 compared to L2 (“L1 slowing”) when exposed to bilingual comprehension but not for monolingual comprehension. Moreover, speaker switches elicited longer naming latencies in bilingual comprehension compared to monolingual comprehension. Both results indicate that even if speakers did not switch languages themselves, comprehending the other speaker’s language influenced their own language production. During bilingual production speakers showed language switching costs when the speaker stayed the same, but these costs were abolished when speakers switched. As in the monolingual production blocks, speaker switches during bilingual comprehension seemed to increase L1 slowing. These findings suggest that speakers employ “reactive” control processes (resulting in switch costs) only in response to the language switching requirements in their own speech, while more proactive language control processes (resulting in L1 slowing) are triggered by comprehending the other speaker’s language.