|dc.description.abstract||This study investigates how Spanish heritage language speakers interpret two types of compound constructions in Spanish: head-initial [N+N] N (e.g., obra cumber ‘masterpiece,’ lit. ‘work summit’) and exocentric [V+N] N (e.g., pelagatos ‘poor man,’ lit. ‘peel+cats’). It was hypothesized that the greater the exposure of participants to English (that has right-headed compounds) the less Spanish-like their pattern of interpreting N-N compounds as left-headed. Further, it was hypothesized that Spanish heritage language speakers who acquire the V-O construction early (prior to age 12) and are thus more likely to be familiar with the conventional mechanism of word formation with the V-N configuration would be more accurate in interpreting [V+N] N than [N+N] N compounds, regardless of their degree of English-exposure. Finally, it was predicted that compounds that are higher in semantic transparency/opacity would be interpreted more accurately than those lower in opacity.
Spanish-English speakers were administered a compound interpretation task in which participants were to select the correct definition for 40 compound words in Spanish. Participants were classified into three groups: late sequential bilinguals who acquired Spanish monolingually in Mexico and learned English after age 12, early sequential bilinguals who acquired Spanish monolingually in the home but came into contact with English at approximately age 6 when they started school, and simultaneous bilinguals who acquired both languages early at home and for who English has been the language of instruction and the dominant language in most social contexts. The performance of these groups was compared to that of a control group consisting of Spanish-dominant speakers.
The findings showed support for the first hypothesis: individuals who acquired English late were better in interpreting the compounds that those who had more exposure to English. The second hypothesis was also supported in that all groups bilinguals interpreted [V+N] N more accurately than [N+N] N compounds regardless of their degree of English-dominance. Finally, compound transparency affected interpretation accuracy. These findings indicate that bilinguals’ performance on compound interpretation in one of their languages in affected not only by linguistic factors (headedness, transparency) but also by bilinguals’ context of acquisition and use of their two languages.||