To fully understand consequences of bilingualism it is important to consider how language use intersects with characteristics of a language. My approach to this issue is to examine the consequences of writing system characteristics for how biliterate bilinguals perceive speech sounds in each language. Hindi-English speakers listened to cross-language homophones presented in separate blocks per language. Upon hearing each word, participants were instructed to take away “the first sound” and say aloud what remained of the word. How participants construe “the first sound” was expected to be influenced by their knowledge of how the words are written in Hindi vs. English. Specifically, it was hypothesized that a grapheme-based representation of sounds in English would favor a phoneme-based interpretation of units of sound whereas an akshara-based representation of sounds in Hindi would favor a more syllable-based interpretation. The results supported this hypothesis. For example, on hearing the word “hum” in an English list, participants deleted the initial phoneme and reported “um” as what remained, whereas for the same word presented in a Hindi list, the initial sound was construed as the syllable “hu” leaving “m” as the remaining sound. This pattern characterized bilinguals whose first learned orthography was Hindi. When Hindi was the second learned script, phoneme deletion was the prevailing response for English and Hindi alike, though with increasing study of Hindi, a syllable-based construal emerged for Hindi words. Thus, consequences of bilingualism differ depending on the type of knowledge of orthographic characteristics of the languages and on the circumstances of language acquisition and use. The same word when heard as a word in Language A vs. Language B will be perceived differently depending on the extent to which the bilingual is literate in both languages, and on how sounds are represented in a given orthography.