The original manuscript is extant here and reads the words at the end of the verse. One wonders if this might not be some sort of error since it seems incomplete. Elsewhere in the text, for the expression “hearken (un)to … word(s)”, the noun word(s) is always made specific, usually by a preceding possessive pronoun or by some kind of postmodification (such as an of-initial prepositional phrase):
“my words” 15 times
“thy words” 1 time
“his words” 4 times
“our words” 1 time
“their words” 4 times
“the words of the Lord” 4 times
“the words of Jesus” 1 time
“the words of the evil one” 1 time
“the words of ” 6 times
“the words of ” 3 times
“the words which the Lord saith” 1 time
“the words of him who gave them unto you” 1 time
“his word” 1 time
“the word of the Lord” 3 times
“the word of God” 1 time
“these words” 1 time
“the words of the book” 1 time
There is only one occurrence of “hearken (un)to … word(s)” without any additional specification for words—namely, here in Alma 35:3.
A number of possible emendations suggest themselves. First of all, Oliver Cowdery initially wrote the final occurrence of the words here in 𝓞 as there words. He immediately corrected the there to the by erasing the re. Since there is a homophone for their, the initial reading in 𝓞, there words, may have been an error for their words. Oliver, after erasing the re, could have neglected to add the ir, thus ending up with the unusual phrase the words. One problem with their words as an emendation is that the reader would tend to interpret the their as referring to the Zoramites rather than to Alma and his missionary companions: “therefore they would not hearken unto their words”. Nonetheless, the text actually has examples of this kind of pronominal usage, as in the following example: “the people hardened their hearts and would not hearken unto their words” (Ether 11:13).
A second possible emendation is that the plural words is an error for word—that is, the original text here read “therefore they would not hearken unto the word”. Here the singular would agree with the preceding occurrence of the word: “they were angry because of the word”. This earlier instance of the singular word is extant in 𝓞 and is supported by the subsequent singular pronoun it (“for it did destroy their craft”). There is clear evidence of mix-ups in the history of the text for word and words, especially with the phrase “to hearken (un)to X”; in fact, two of these mix-ups show the tendency to replace the correct singular word with the plural words (each marked below with an asterisk):
A third possibility is that the bare phrase the words was followed by a postmodifying relative clause that was somehow lost in the early transmission of the text. Further on in Alma 35 the text refers to “the words which had been spoken” (at the end of verse 4) and “the words which had been spoken by Alma and his brethren” (in the middle of verse 6). These examples suggest an emendation in verse 3 like “the words which had been preached unto them”, as suggested by the language earlier in verse 3: “concerning the words which had been preached unto them”. The problem with this emendation is that it seems very doubtful that a whole relative clause would have been lost during the dictation of the text.
But this last possibility suggests that such a postmodification may have been left unexpressed—that is, the use of the words at the end of verse 3 refers to the earlier plural occurrence of the words (that is, “concerning the words which had been preached unto them”). In other words, the text at the end of the verse intentionally avoids the repetition of the postmodifying relative clause “which had been preached unto them”. Another way, of course, to have the final instance of the words in verse 3 refer to the earlier instance of the words (“the words which had been preached unto them”) would be to replace the the with a demonstrative determiner such as these or those: “therefore they would not hearken unto these words” or “therefore they would not hearken unto those words”.
Ultimately, we have a number of possible emendations. Note, however, that no edition has ever attempted to alter the occurrence of the words at the end of verse 3. Probably the best solution is to leave the corrected reading in 𝓞, “therefore they would not hearken unto the words”, where the postmodifying relative clause “which had been preached unto them” is implied.
Summary: Maintain in Alma 35:3 the corrected reading of the original manuscript: “therefore they would not hearken unto the words”; here the words is not postmodified in any way, but the apparent meaning is that the words refers to “the words which had been preached unto them” (stated earlier in the verse).