Dear list,
In my research on the ancient Syriac Bible translation (Peshitta), I use this he concept of Cognate Interference as known in translation studies, that is: the tendency in translations where the source text language and the target language are cognate to prefer
cognate over non-cognate alternatives.
[See, e.g., for modern Hebrew and Arabic: Ordan, Noam, Hershberg, Nimrod and Shlesinger, Miriam. "Translational conflicts between cognate languages: Arabic into Hebrew as case in point" Corpus Linguistics and Linguistic Theory, vol. 6, no. 2, 2010, pp. 217-239.
https://doi-org.vu-nl.idm.oclc.org/10.1515/cllt.2010.008]
The cognate interference does not mean that the cognate alternatives are wrong (at least not by definition), but rather that they occur in a significant higher frequency in translated texts than in non-translated texts. In the case of Hebrew and Syriac I found
a higher frequency of the bipartite nominal clause (as against the tripartite construction which is more common in Syriac non-translated texts) and of the conjunction waw as against other conjunctions in non-translated texts.
I can well image that the same phenomenon of Cognate Interference occurs in ancient translations of the Hebrew Bible into Arabic. Any data, experiences or bibliographical references are welcome!
Best wishes,
Willem
Prof. dr. Willem Th. van Peursen
School of Religion and Theology, ETCBC
Vrije Universiteit Amsterdam
De Boelelaan 1105, 1081 HV Amsterdam
X: @PeursenWTvan
Eep Talstra Centre for Bible and Computer