Variation between Present Perfect and (Perfective) Past forms has been mostly studied on English [1-2], but lately a crosslinguistic perspective has also considered some other different European languages [3-4-5]. However, these studies usually take a macrovariation perspective and fail to recognize that these languages are not uniform entities, and that dialectal microvariation might show relevant constraints at play in the alternation between these forms.

One case of this alternation is seen in Spanish, where some past actions are expressed with a Perfect, the Pretérito Perfecto Compuesto (PPC: Ana ha vivido en París ‘Ana has lived in Paris’), and some past actions are expressed with a Perfective, the Pretérito Indefinido (PI: Ana vivió en París ‘Ana lived in Paris’). The distinction between these forms is not clear and the constraints on their distribution are not well-understood. A few studies have shown that Spanish uses the PPC not only for typical Perfect readings (resultative, continuative), but also for Perfective meanings, such as the expression of bounded events, as long as they are contained in the hodiernal past [6]. However, most research studies Peninsular dialects, which are spoken by only one tenth of the Spanish speaking population [6-7].

To bridge the gap between macro and microvariation, we analyze crossdialectal variation of these markers across three Spanish dialects: Peninsular, Rioplatense, and Mexican. We use a new methodology in which parallel corpora are analyzed to compare crosslinguistic data [5]. We gather our data from three dialect-specific translations of Harry Potter and the Philosopher’s Stone. This method allows to do a detailed comparison since the context of use is kept stable, and the translator needs to decide which form to use.

Out of 44 English Present Perfects, and contrary to previous studies [8], we find that PPC use is greater in the Mexican variety (n=37), than in Peninsular (n=32) and Rioplatense (n=18) dialects, which show more use of PI. In cases in which the English original displays a Simple Past (n=775), most tokens are translated to the Spanish PI across dialects, but once again Mexican Spanish shows more PPC tokens (n=17) than its Peninsular (n=6) and Rioplatense (n=1) counterparts.

In line with previous work on dialect variation [9-10], we show that the contextual constraints at play in selecting each form are diverse across different Spanish dialects. While in Peninsular Spanish considerations about hodiernality are relevant, Mexican Spanish makes a more extended use of the Perfect form, including cases of a more distant past that are made “currently relevant” from the speaker’s point of the view. Rioplatense Spanish, in contrast, presents only the prototypical Perfect readings, and defaults to the use of PI otherwise. In sum, we show that cross-dialectal variation becomes crucial for the proposal of cross-linguistic generalizations. Data like ours is better explained from a bottom-up perspective in which specific tokens are not fitted into predefined categories, but the richness of dialectal variation is wholeheartedly recognized.