Kamil Stachowski

The paper discusses a group of eleven words with similar phonetic shapes and somewhat similar semantics: jagu-, jak- ‘to come near’; jan- ‘to turn back’; jaguk, jakȳn ‘close, near’; jāk, jān ‘side’; jāna- ‘to sharpen’; jaŋak ‘cheek’; jaŋy ‘new’; and jaka ‘edge’. All have been suspected to belong to the same family, at the heart of which, most probably, would be the verbal root *jā-. Some of the problems associated with this idea were known previously, whereas some are newly identified here. The paper considers various constraints and proposes a scheme centred around *jā ∼ *ja- ‘to be near, …’, which may or may not be connected to MaTung. daga ‘id.’ and Mo. daga- ‘to follow’.

This paper argues that automatic phonetic comparison will only return true results if the languages in question have similar and comparably lenient phonologies. In the situation where their phonologies are incompatible and / or restrictive, linguistic knowledge of both of them is necessary to obtain results matching human perception. Whilst the case is mainly exemplified by Levenshtein distance and Russian loanwords in Dolgan, the conclusion is also applicable to the approach as a whole.

The paper proposes a multi-dimensional, phonologically-aware numeric encoding of Turkish for use with neural networks. The system is evaluated and compared to PatPho (Li/MacWhinney 2002) in a test in which the network computes the shape of the past tense suffix.
 

Artykuł jest przyczynkiem do dialektologii percepcyjnej Polski, ale przede wszystkim dyskusją wyników eksperymentu, w którym dwie grupy poproszono o wypełnienie map. Jednej przedstawiono mapy z zaznaczonymi głównymi miastami, drugiej – z głównymi regionami geograficznymi. Mapy powstałe przez scalenie ich odpowiedzi okazały się diametralnie różne, co sugeruje, że ten jeden szczegół w sposobie przygotowania kwestionariusza może bardzo istotnie wpłynąć na wyniki badania, i jako taki winien być starannie przemyślany i dokładniej przebadany.

There is a sizeable group of words in Turkish whose ultimate origin is known to be Arabic but whose direct donor language is unclear. The paper analyses 69 such words, and compares the phonetic adaptations present in them, to those attested in Arabisms as well as to those found in Farsisms, in order to determine the probability of them belonging to one group or the other. The results are compared to the opinions of the main etymological dictionaries of Turkish, splitting them into two camps.

A specialist in Middle Eastern languages will likely be quick to associate Pol. mamuna ‘an ape-like mythological creature’ with Ar./Pers./Tkc. majmun ‘ape’. It is possible and indeed probable that this name is an Oriental borrowing applied to an ancient native belief, but a closer inspection reveals at least several other possibilities tangled in an ethnolinguistic web of potential conflations and contaminations. This paper presents the ethnographic background and some etymological ideas, though without as yet a definite answer.

A specialist in Middle Eastern languages will likely be quick to associate Pol. mamuna ‘an ape-like mythological creature’ with Ar./Pers./Tkc. majmun ‘ape’. It is possible and indeed probable that this name is an Oriental borrowing applied to an ancient native belief, but a closer inspection reveals at least several other possibilities tangled in an ethnolinguistic web of potential conflations and contaminations. This paper presents the ethnographic background and some etymological ideas, though without as yet a definite answer.

The article presents a – to the best of the author’s knowledge – new method of preparing data for quantification of loanword adaptation, together with two of its possible uses. The method is particularly fit for poorly investigated languages where a great deal of data, especially socio-linguistic, are missing. It is illustrated with the example of Russian loanwords in Dolgan. The result is an attempt to measure the commonness and meaningfulness of adaptations, and an index of loanword nativization

This paper is an edition of an article by Władysław Kotwicz (1872–1944) entitled Les voyelles longues dans les langues altaïques, which the author could not publish himself during wartime and did not live to publish after the War was over. The edition is designed to read almost as if published by Kotwicz, but without falsifying the actual manuscript. Also, a brief archival description is provided and the history of the last four years of the text has been reconstructed, based mostly on Kotwicz’s correspondence.

This paper is an edition of an article by Władysław Kotwicz (1872–1944) entitled Les voyelles longues dans les langues altaïques, which the author could not publish himself during wartime and did not live to publish after the War was over. The edition is designed to read almost as if published by Kotwicz, but without falsifying the actual manuscript. Also, a brief archival description is provided and the history of the last four years of the text has been reconstructed, based mostly on Kotwicz’s correspondence.

This paper is an edition of an article by Władysław Kotwicz (1872–1944) entitled Les voyelles longues dans les langues altaïques, which the author could not publish himself during wartime and did not live to publish after the War was over. The edition is designed to read almost as if published by Kotwicz, but without falsifying the actual manuscript. Also, a brief archival description is provided and the history of the last four years of the text has been reconstructed, based mostly on Kotwicz’s correspondence.

This paper is an edition of an article by Władysław Kotwicz (1872–1944) entitled Les voyelles longues dans les langues altaïques, which the author could not publish himself during wartime and did not live to publish after the War was over. The edition is designed to read almost as if published by Kotwicz, but without falsifying the actual manuscript. Also, a brief archival description is provided and the history of the last four years of the text has been reconstructed, based mostly on Kotwicz’s correspondence.