thumb|265x265px|The requisition of grain from "wealthy peasants" (kulaks) during state Collectivization in the Soviet Union|collectivization in [[Timashyovsky District, Kuban, Soviet Union, 1933]] A kulak ( ; ; plural: кулаки́, kulakí, 'fist' or 'tight-fisted'), also kurkul () or golchomag (, plural: ), was a peasant who owned over of land in the times near the end of the Russian Empire. In the early Soviet Union, particularly in Soviet Russia and Azerbaijan, kulak referred to property ownership among peasants who were considered hesitant allies of the Bolshevik Revolution. In Ukraine during 1
thumb|265x265px|The requisition of grain from "wealthy peasants" (kulaks) during state Collectivization in the Soviet Union|collectivization in [[Timashyovsky District, Kuban, Soviet Union, 1933]] A kulak ( ; ; plural: кулаки́, kulakí, 'fist' or 'tight-fisted'), also kurkul () or golchomag (, plural: ), was a peasant who owned over of land in the times near the end of the Russian Empire. In the early Soviet Union, particularly in Soviet Russia and Azerbaijan, kulak referred to property ownership among peasants who were considered hesitant allies of the Bolshevik Revolution. In Ukraine during 1930–1931, there also existed a term of podkulachnik (almost wealthy peasant); these were considered "sub-kulaks".
Kulaks referred to former peasants in the Russian Empire who became landowners and credit-loaners after the abolition of serfdom in 1861 and during the Stolypin reform of 1906 to 1914, which aimed to reduce radicalism amongst the peasantry and produce profit-minded, politically conservative farmers. During the Russian Revolution, kulak was used to chastise peasants who withheld grain from the Bolsheviks. According to Marxist–Leninist political theories of the early 20th century, the kulaks were considered class enemies of the poorer peasants. Vladimir Lenin described them as "bloodsuckers, vampires, plunderers of the people and profiteers, who fatten themselves during famines", declaring revolution against them.
Discovered by embedding cosine similarity (sentence-transformers MiniLM, 384-dim).