Complexity vs. salience of alternatives in implicature: A cross-linguistic investigation

Scalar implicature depends on the activation of alternatives. For instance, in English, finger implicates 'not thumb', suggesting that thumb is an activated alternative. Is this because it more specific (Quantity) and equally short (Manner)? Indeed, toe doesn't imply big toe', perhaps toe is longer. As L. Horn points out, Quantity/Manner explanation predicts if English had simplex Latin word pollex meaning 'thumb or then asymmetry would disappear. But suffice for to exist language, also have be sufficiently salient? We explore question four languages are sometimes said lack a single-word alternative thumb: Spanish (which does pulgar 'thumb toe' (< pollex), though non-colloquial form), Russian, Persian, Arabic. To gauge salience various ways describing digits, we use fill-in-the-blank production task. measure availability implicatures using forced choice comprehension find cross-linguistic differences implicature, moreover calculation tracks probabilities closely than structural complexity A comparison between two Rational Speech Act models—one which speaker replicates our data standard one chooses based cost/accuracy trade-off—shows tied probability alternatives.