This study addresses the French morphological construction [VN/A/Adv/P]N/A. The main objectives are to posit a single rule for its formation and to question the implications of the agent polysemy. The theoretical framework is lexeme-based morphology, which adheres to weak lexicalism. The first part of our analysis is qualitative and concerns the availability aspect of productivity. The method is introspective. The internal semantic patterns of the French construction are examined. Our results give evidence for the claim that a single morphological construction rule, [VN/A/Adv/P]N/A, is responsible for the cases where the first constituent is a verb stem, and the second constituent may correspond to an internal argument, an external argument or a semantic adjunct. All cases manifest the same patterns, which are related to the denotative meanings included in the agent polysemy: Agent, Instrument, Locative, Action, Result and Cause. Our contrastive analysis shows that the same patterns are found in the four Swedish agentive formations, [N/A/Adv/PV-are]N, [N/A/Adv/PV]N, [N/A/Adv/PV-a]N and [VN]N, which correspond to the French [VN/A/Adv/P]N/A construction and which also contain a verbal constituent and its internal or external argument, or an adjunct. The second part of our analysis is quantitative and concerns the profitability aspect of productivity. The method is inductive. The aim is to explore the polysemy of agent and its assumed hierarchical structure, in synchrony and diachrony. Four French agentive formations, [VN/A/Adv/P]N/A compounds and V-ant, V-eur and V-oir(e) derivations, are included in order to examine semantic competition and blocking effects. Our results give evidence for the existence of an agent polysemy but deny that it has a hierarchical structure valid for every agentive formation. The meanings in the agent polysemy are more or less profitable according to formation type: blocking effects could explain this behaviour.

This study addresses the French morphological construction [VN/A/Adv/P]N/A. The main objectives are to posit a single rule for its formation and to question the implications of the agent polysemy. The theoretical framework is lexeme-based morphology, which adheres to weak lexicalism. The first part of our analysis is qualitative and concerns the availability aspect of productivity. The method is introspective. The internal semantic patterns of the French construction are examined. Our results give evidence for the claim that a single morphological construction rule, [VN/A/Adv/P]N/A, is responsible for the cases where the first constituent is a verb stem, and the second constituent may correspond to an internal argument, an external argument or a semantic adjunct. All cases manifest the same patterns, which are related to the denotative meanings included in the agent polysemy: Agent, Instrument, Locative, Action, Result and Cause. Our contrastive analysis shows that the same patterns are found in the four Swedish agentive formations, [N/A/Adv/PV-are]N, [N/A/Adv/PV]N, [N/A/Adv/PV-a]N and [VN]N, which correspond to the French [VN/A/Adv/P]N/A construction and which also contain a verbal constituent and its internal or external argument, or an adjunct. The second part of our analysis is quantitative and concerns the profitability aspect of productivity. The method is inductive. The aim is to explore the polysemy of agent and its assumed hierarchical structure, in synchrony and diachrony. Four French agentive formations, [VN/A/Adv/P]N/A compounds and V-ant, V-eur and V-oir(e) derivations, are included in order to examine semantic competition and blocking effects. Our results give evidence for the existence of an agent polysemy but deny that it has a hierarchical structure valid for every agentive formation. The meanings in the agent polysemy are more or less profitable according to formation type: blocking effects could explain this behaviour.

Artificiell intelligens (AI) har integrerats i stor uträckning, framförallt inom bilindustrin där autonoma bilar blir allt vanligare. Då allt fler bilar får autonoma funktioner förändras bilförarnas tidiga roll. Förare går från att manuellt manövrera bilen till att övervaka så att de autonoma funktionerna fungerar. Då förarens roll förändras tillkommer utmaningar kring förarnas upplevelser för autonoma funktioner. Syftet med studien är att identifiera dessa utmaningar vilket leder till frågeställningen: Vilka utmaningar innebär design av initiativskifte i autonoma bilar i förhållande till upplevt förtroende? För att skapa förståelse har en kvalitativ studie genomförts där en designetnografiskt ansats legat till grund för den presenterade illustrationen. Deltagarna i studien har varit personer som äger en bil med autonoma funktioner. Studien bidrar med kunskap om hur upplevt förtroende i relation till det initiativskifte som sker i autonoma bilar vilket presenteras i en illustration. I slutsatsen reflekteras samhälleliga och etiska utmaningar vad gäller designande av autonoma bilar / Artificial intelligence (AI) has been widely integrated, especially in the automotive industry where autonomous cars are becoming more common. As more and more cars get autonomous functions, the drivers' previous role changes. Drivers go from manually operating the car to monitoring it, so that the autonomous functions work. As the driver's role changes, challenges arise for drivers' experiences for autonomous functions. The purpose of the study is to identify these challenges that lead to the question: What challenges does the design of initiative shift in autonomous cars in relation to perceived trust entail? To create understanding, a qualitative study has been conducted where a design ethnographic approach was used as the basis for the presented illustration. The participants in the study were people who own a car with autonomous functions. The study contributes with knowledge of how perceived reliance in relation to the change in initiative that takes place in autonomous cars, which is presented in an illustration. The conclusion reflects societal and ethical challenges regarding the design of autonomous cars.