An official website of the United States government
This content will become publicly available on August 1, 2025
Turkish Delights: a Dataset on Turkish Euphemisms
- Award ID(s):
- 2226006
- PAR ID:
- 10565061
- Publisher / Repository:
- Association for Computational Linguistics
- Date Published:
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
We present a crowd-driven adjudication system for rejected work on Amazon Mechanical Turk. The Mechanical Turk crowdsourcing platform allows Requesters to approve or reject assignments submitted by Workers. If the work is rejected, then Workers aren’t paid, and their reputation suffers. Currently, there is no built-in mechanism for Workers to appeal rejections, other than contacting Requesters directly. The time it takes Requesters to review potentially incorrectly rejected tasks means that their costs are substantially higher than the payment amount that is in dispute. As a solution to this issue, we present an automated appeals system called Turkish Judge which employs crowd workers as judges to adjudicate whether work was fairly rejected when their peers initiate an appeal. We describe our system, analyze the added cost to Requesters, and discuss the advantages of such a system to the Mechanical Turk marketplace and other similar microtasking platforms.more » « less
-
This study investigates the processing of Turkish negative polarity items (NPIs) using a self-paced reading experiment with end-of-sentence acceptability judgements. Our participants included adult Turkish monolinguals, as well as Turkish-German early (i.e. heritage speakers) and late bilinguals. We explored whether intrusion effects from illusory NPI licensors extended to bilingual Turkish speakers who had acquired German either early or late in their lives. Stimuli included 30 sets of sentences in six experimental conditions, with the presence of both an NPI and of a suitable licenser (verb negation) systematically manipulated. Our results indicate that bilingual Turkish readers show intrusion effects in their processing of NPIs. Our findings suggest that the structural conditions for NPI licensing in Turkish might be degraded or less stable in heritage bilinguals.more » « less
-
Abstract This study examined putative benefits of testing and production for learning new languages. Undergraduates (N= 156) were exposed to Turkish spoken dialogues under varying learning conditions (retrieval practice, comprehension, verbal repetition) in a computer‐assisted language learning session. Participants completed pre‐ and posttests of number‐ and case‐marking comprehension, a vocabulary test, and an explicit awareness questionnaire. Controlling for nonverbal ability and pretest scores, the retrieval‐practice group performed highest overall. For number/case marking, the comprehension and retrieval‐practice groups outperformed the verbal‐repetition group, suggesting benefits of either recognition‐ or recall‐based testing. For vocabulary, the verbal‐repetition and retrieval‐practice groups outperformed the comprehension group, indicating benefits of overt production. Case marking was easier to learn than number marking, suggesting advantages for learning word‐final inflections. Explicit awareness correlated with comprehension accuracy, yet some participants demonstrated above‐chance comprehension without showing awareness. Findings indicate the value of incorporating both practice tests and overt production in language pedagogy.more » « less
