According to Brant Gardner, even without records, the three hundred years of separation between the Nephites and Zarahemlaites is not likely to be sufficient for mutual unintelligibility as long as they kept speaking the same language. What is more probable is that without records, and in the midst of another culture, there was no reason for the Mulekites (and thus the people of Zarahemla) to hold to a language that was not a benefit to communicate. If the Mulekites landed among the Olmecs, they would have learned common Zoquean or common Mixean. Given the probable location of the local land of Zarahemla in the Grijalva river basin, this ultimately places the people of Zarahemla in the territory historically associated with Zoque speakers. [Brant Gardner, "Book of Mormon Commentary," [http://www.highfiber.com/~nahualli/] LDStopics/Omni/ Omni1.htm, p. 34]
“They Had Brought No Records with Them”
John Sorenson notes that "the largest archaeological site on the upper Grijalva in an appropriate position to qualify as Zarahemla is Santa Rosa." [John L. Sorenson, An Ancient American Setting for the Book of Mormon, p. 153]
According to Brant Gardner, linguistic research tells us that the upper Grijalva lay at the juncture of two major areas where long-established peoples and their languages existed. A couple of thousand years ago the Mayan languages probably extended throughout much of Guatemala to about the mountainous wilderness strip that separates the highlands of that nation from the Grijalva River valley (Chiapas Depression). Downstream, from near Chiapa de Corzo and extending north and westward, were speakers of Zoque dialects; in the isthmus proper was the closely related Mixe language. Both blocs, the Mayan speakers on the Guatemalan side and groups using tongues of the Mixe-Zoquean family on the isthmian side of Santa Rosa (the proposed site for Zarahemla), had been there for a long time.
Ancestral Mixe-Zoquean has been shown to be the probable language of the Olmecs of the Gulf Coast, while Mayan speakers likely had been in the Cuchumatanes Mountains of northern Guatemala since well before 1000 B.C. (Evidence is uncertain, however, whether Mayan languages were spoken until post-Book of Mormon times in the actual areas of the southern Guatemala highlands where the Nephite and Lamanite settlements are best placed.) But neither major language group seems to have been established on the upper Grijalva, at least not until well into A.D. times. That intermediate zone seems to have been a linguistic frontier. Zarahemla's people had apparently moved into the area from the Gulf Coast through lands occupied by Zoque speakers for centuries. Zarahemla's local followers in Mosiah's day likely spoke a language like Zoquean. Mosiah and his party, coming from the opposite direction, were among the first of a long series of groups who drifted out of Guatemala into this valley over the next thousand years.
So in this context we can better understand Amaleki's statement concerning the people of Zarahemla that "their language had become corrupted" (Omni 1:17). [Brant Gardner, "Book of Mormon Commentary," [http://www.highfiber.com/~nahualli/LDStopics/Omni/] Omni1.htm, pp. 19-20]