How did Paul Rivet’s classification of South American languages differ from earlier systems?

The classification of South American Indian languages by the French anthropologist and ethnologist Paul Rivet, published in 1924, differed from earlier systems primarily through its comprehensive scope and evidentiary basis.1

While previous classifications, such as the 1891 work by Daniel Brinton, relied heavily on grammatical criteria and restricted word lists, Rivet’s system was supported by his extensive, detailed prior studies and a wealth of information. His classification, which recognized 77 language families, was based on the similarity of vocabulary items. By incorporating this vast amount of data, Rivet’s work effectively superseded all previous classifications, providing a more robust framework that remained a standard for subsequent researchers, including Cestmír Loukotka, who later expanded upon Rivet's methodology.1