Some of the problems in the comparative study of languages within the phylum are the result of their early extinction and limited documentation.

[3][4] There also seems to be convincing evidence for the Plateau Penutian grouping (originally named Shahapwailutan by J. N. B. Hewitt and John Wesley Powell in 1894) which would consist of Klamath–Modoc, Molala, and the Sahaptian languages (Nez Percé and Sahaptin).

[7] The grouping, like many of Dixon & Kroeber's other phylum proposals, was based mostly on shared typological characteristics and not the standard methods used to determine genetic relationships.

[11][12] Scholars in the mid-twentieth century became concerned that similarities among the proposed Penutian language families may be the result of borrowing that occurred among neighboring peoples, not of a shared proto-language in the distant past.

Mary Haas states the following regarding this borrowing: Even where genetic relationship is clearly indicated ... the evidence of diffusion of traits from neighboring tribes, related or not, is seen on every hand.

The opposite approach was taken following a 1976 conference at Oswego, New York, when Campbell and Mithun dismissed the Penutian phylum as undemonstrated in their resulting classification of North American language families.

[14] Subsequently, Marie-Lucie Tarpent reassessed Tsimshianic, a geographically isolated family in northern British Columbia, and concluded that its affiliation within Penutian is also probable.