This article investigates the role of script choice in bilingual
writing, drawing on classified advertisements and other texts written for
and by Russian-speaking immigrants in New York City. The study focuses on
English-origin items that appear in Russian texts, which are found to be
written either in roman or Cyrillic script. Through an investigation of
categorical and variable constraints on this variation, it is found that
script choice relates to the distinction between lexical borrowing and
single-item codeswitching. It is argued that writers may, consciously and
on a token-by-token basis, choose the Cyrillic script to mark a word as
borrowed or the roman script to mark it as foreign. However, they may also
avoid this choice, as hybrid forms attest, especially when the use of
characters shared by both alphabets allows ambiguous readings. The
findings thus have implications for understanding notions of language
boundaries in bilingual language use.Versions of this article were presented at two conferences:
“Alphabetics: Interpreting letters” at Harvard
University, 26–27 April 2003, and NWAVE 32, Philadelphia,
9–12 October 2003. I thank the audiences for their valuable insights
and observations, especially Erika Boeckeler and Daniel Kokin. I also owe
thanks to Katya Korsunskaya, Vladislav Rapoport, Doris Stolberg, Mario
Geiger, and Tobias Kuhn. I am grateful to John Victor Singler, Mark Sebba,
and Jannis K. Androutsopoulos, as well as to Jane H. Hill and to two
anonymous reviewers for their valuable comments and suggestions. All
errors and omissions remain my own.