The earliest reference to the apparition occurred in 1648, 117 years after its purported occurrence, in a highly wrought theological meditation by a Franciscan, Miguel Sanchez, who said he was relying on oral tradition since there was no prior written record of the apparition. This lack of earlier evidence was an embarrassment to those promoting the cult of Juan Diego, so a document was produced in 1649, a flowery account of the apparition written in the local language, Nahuatl, and supposedly composed shortly after the date (1548) when the phantom Juan Diego died. Outstanding experts on sixteenth-century Mexico have long treated this as a forgery - Juan Bautista Munoz in the eighteenth century, for instance, and Joaqin Garcia Icabalceta in the nineteenth century - and three modern specialists in Nahuatl have demonstrated that it depends on (and therefore post-dates) the 1643 work by Miguel Sanchez. Even the Cambridge historian, D. A. Brading, who is sympathetic to the symbolism of the Lady of Guadalupe, calls this last investigation “a devastating criticisms of the received linguistic and textual foundations of the tradition.”