This dissertation represents the first attempt to integrate typological, semantic, and psycholinguistic perspectives to elucidate a semantically "bizarre" and "illogical" phenomenon called "expletive negation" (henceforth, EN) which is well known in Romance languages but has so far attracted little attention outside of Romance linguistics and in psycholinguistics. The dissertation starts with a typological survey of the occurrence of EN in a 1,142 language sample (covering 341 linguistic genera) which reveals that EN is attested in every linguistic macro-area and that there is considerable overlap of EN-triggers across unrelated languages. A comprehensive study of EN is then conducted in French and Mandarin and a near-exhaustive list of EN-triggers collected from the two languages is provided: EN-triggers attested in both languages are almost identical. To further test the applicability of this near-exhaustive list of EN-triggers, an investigation of EN in two understudied languages, Januubi Arabic and Zarma-Sonrai, is conducted via elicitation. Altogether, the striking similarity of EN-triggers across four different languages (and four different language families) provides strong support for the hypothesis that EN has universally-shared cognitive underpinnings. A language production model is then proposed: EN arises from the concurrent activation of a negative proposition ¬p entailed (or strongly implicated) by an EN-trigger alongside the intended complement proposition p; it is the erroneous production of the negative proposition ¬p rather than the intended proposition p that leads to the occurrence of EN in speech production. EN may start out as a speech error but it can get highly entrenched (or grammaticalized) through repetition in some languages. This model predicts (i) the existence of EN in the same range of EN-triggering contexts in languages where EN is believed not to exist as in French or Mandarin (e.g., English), and (ii) the difference in frequency of EN being produced across triggers in a particular language as well as frequency of overall EN uses across languages. Both predictions are tested and confirmed in an English and Mandarin corpus study. Last, the comprehension of EN is explored: How native speakers of different languages (with different degrees of overall EN entrenchment) understand EN? The model predicts uniformity as well as variability across languages in speakers' comprehension of EN. Three similar semantic interference comprehension experiments are run in English, Mandarin, and French. The results of these experiments show that speakers of all three languages sometimes understand expletively a negator in the scope of an EN-trigger although speakers of the three languages show a different propensity to do so. In addition, for both English and Mandarin, the more a negator is used expletively in the complement of an EN-trigger in corpora, the more likely comprehenders are to interpret expletively a novel occurrence of a negator in the complement of that EN-trigger. Overall, the typological investigation, semantic and production account of EN, and the corpus and experiment results support the claim that "illogical" expressions that recur across languages of the world reflect universal properties of the language production system. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com.bibliotheek.ehb.be/en-US/products/dissertations/individuals.shtml.]