This article records tradition as it has been passed down and reported. Its sources are not yet part of the atlas's verified catalogue.
Proposed Groupings gathers three long range hypotheses that would join established language families into larger units at a depth most historical linguists consider unreachable by the comparative method, and none of the three commands the field's acceptance.
Altaic proposes to join Turkic, Mongolic and Tungusic, and in its wider versions Korean and Japonic; most specialists now reject it, on the grounds that its shared vocabulary concentrates in exactly the areas long contact would produce through borrowing, and the core inherited material a family needs has not been found, though the Moscow school continues to defend it. Nostratic proposes a macro-family joining Indo-European to Afro-Asiatic, Uralic, Dravidian and others at a depth of ten thousand years or more; its advocates form a real school with published reconstructions, but most historical linguists hold that the comparative method cannot reach that far back and that the resemblances offered fall within the range of chance. Dene-Caucasian links several otherwise unrelated families, most often Sino-Tibetan, Na-Dene of North America, Basque, the North Caucasian languages and Burushaski, on the strength of proposed shared vocabulary and grammar, and belongs to the same class of contested long range hypothesis as the other two.
The atlas holds all three as proposals under active dispute rather than as demonstrated families, the honest state of a question historical linguistics has not settled.