Nominal reduplication used to be the normal (though variable) strategy to mark plurality, as attested in late 19th-century accounts of Diu Indo-Portuguese (Schuchardt 1883). The examples recently encountered were recorded in the speech of an old lady and a poorly educated middle-aged man, and they were rare even in the speech of these consultants. All examples encountered involved animate (human and non-human) referents, but the truth of the matter is that the corpus of occurrences (both 19th-century and modern data) is too reduced to discard the possibility that pluralizing reduplication has ever applied to inanimate referents.
Source: Cardoso 2009: 120
Source: Cardoso 2009: 175