Scholars have disagreed about whether false implicatures can be considered lies, with some claiming they can (e.g., Meibauer, 2014), and others arguing that they cannot (e.g., Dynel, 2011; Saul, 2012). Some in the second camp go as far as saying that whether an utterance can be considered a lie is the very test for whether content is said explicitly, since those judgments are more consistent than true/false judgments (Michaelson, 2016).

This work, part of a larger experimental project investigating lying and the notion of what is said, looks at people’s predispositions to disagree with the person who delivers the utterance in question. I investigated disagreement based on political affiliation, which has been demonstrated to influence on-line language processing in politically relevant contexts (von der Malsburg, Poppels, and Levy, forthcoming), and how this individual-level difference can affect lie judgments. Participants in this study (N = 445) read one of three hypothetical scenarios in which either President Obama or President Trump utters a false implicature, such as “Some committee members have been accused of corruption” after the pre-utterance context reveals that all have. After giving their lie judgment on a 1-7 linear sliding scale, participants answered whether they (1) generally support President Trump and (2) generally did support President Obama.

​

Lie judgments of these utterances differed significantly based on whether the person judging the key utterance agrees with the speaker who uttered it (p < .001), with lie ratings from people who disagree with the speaker (mean = 3.85) significantly higher than lie ratings from those who agree with the speaker (mean = 2.87). The predilection for agreeing with the character caused participants to consider a more literal utterance meaning in their lie judgment.

​

These results are taken as evidence in support of the notions of e.g., Ariel’s (2002) Privileged Interactional Interpretation (PII) or Jaszczolt’s (2009) primary meaning (PM). What is taken as “said” by an utterance may not be consistent across all contexts and individuals, and different levels of linguistic meaning may emerge as the PII/PM based on extralinguistic factors. Specifically, a predisposition to agree politically has been demonstrated to be a contextual factor that urges people towards a bare linguistic conception of what is said, whereas a predisposition to disagree pushes people to count the GCI (in Neo-Gricean terms) or explicature (in Relevance Theory terms) as what is said. The results also indicate that a Lying Test, such as that proposed by Michaelson, is a misguided way to determine what is “said” by an utterance, due to widespread variation in lie judgments.