Abstract The current study integrates studies of priming and communication to investigate whether an interlocutor is necessary for successful communication, and whether language intent can be non-consciously primed to be helpful. Previous research has shown social behaviour to be automatically activated using the methodology of priming (Herr 1986; Bargh 1990; Ferguson & Bargh 2004; Chartrand 1996; Epley & Goldvich 1999). Priming of linguistic production, an essential feature of social interaction has been found to facilitate communication (Bock 1980; Haywood, Pickering & Branigan, 2000). Word search tasks (Bargh, Gollwitzer, Lee-Chai, Barndollar &Trotschel, 2001) were employed as the priming manipulation containing helpful, unhelpful or neutral words. The subsequent task, utilised methodology of Clark and Wilkes-Gibbs (1986) consisting of a referential communication task using a sequence of complex figures, which participants described in monologue. We explored whether or not there was a difference in language produced in three primed conditions; helpfulness, unhelpfulness and neutral. Our results found contrasting results to previous social priming research; the priming manipulations did not have a significant effect on language intent. However, we interpret these results as support dialogue being necessary for successful communication and understanding (Garrod & Pickering 2004; Pickering & Garrod 2004; Clark and Wilkes-Gibbs 1986; Garrod & Anderson 1987).