This study asks whether and how the features that define a language variety co-vary within the communities and speakers said to be representative of it. Of particular interest is the relationship between multiple variables in a setting known to promote contact-induced language change. The central idea that emerges here is that less salient linguistic variables are more likely to co-vary, that is, to be uniformly influenced by the contact setting, than are variables of higher salience. This claim is supported by an analysis of five variables in the speech of four Spanish-speaking adults, two of whom have lived their entire lives in the contact setting and two who are recent arrivals to it. The variables are (1) filled pauses, (2) the presence vs. absence of subject pronouns, (3) subject pronoun position (i.e., pre- vs. post-verbal), (4) general subject position (the pre- or post-verbal position of non-pronominal subjects, e.g. lexical NPs, clauses, etc.), and (5) coda /s/ weakening, examined in terms of rates of deletion as well two acoustic parameters. It is only with respect to the last of these features, which is highly salient sociolinguistically, that strong regionally delineated continuity in the Spanish of the U.S. born speakers is clearly observed. The four lower salience features have shifted in parallel, increasing in similarity to the use of analogous features in English. These results indicate that in a setting characterized by language contact, the fate of socio-linguistic variables is mediated by salience. Low salience features are more susceptible to the influence of the contact setting and are more likely to be uniformly reshaped by it. High salience features, in contrast, are differentiated by speakers’ greater awareness of their social signaling potential and are more likely to unfold along autonomous and individuated trajectories.