“Lexical Evidence for the Relative Chronology of Old English Poetry” by Leonard Neidorf:
This article explores the dating implications of rare vocabulary attested in Beowulf, Genesis A, Daniel, Exodus, Maxims I, and Widsið. It argues that these poems preserve an archaic lexical stratum, which consists of words that became obsolete before the composition of ninth-century poetry and prose.
Neidorf presents a compelling argument for an early dating of the aforementioned poems based on the work of Dennis Cronan, while expertly smashing aside the poor, contrarian arguments of Roberta Frank. He analyses a multitude of Old English words, both from Cronan’s earlier work and his own research, and provides a glimpse into early Anglo-Saxon vocabulary that is not present in later, extant manuscripts, let alone the standard dictionaries for the language today.
His conclusion includes a fantastic swipe at those who feel that Beowulf cannot be dated firmly:
The controversy over the dating of Beowulf is a product not of ambiguous linguistic evidence, but of the tendency of literary scholars to ignore linguistic evidence and frame the question of dating in ambiguous terms not conducive to rational debate. (pg. 40)
In dating some of the poems:
The corpus of archaic poetry, encompassing works probably composed at various dates between roughly 675 and 750, consists chiefly of Beowulf, Genesis A, Daniel, Exodus, Guthlac A, and Christ III. (pg. 39)
Consider reading the entire paper, as it discusses these things and more at length.