A psycholinguistic investigation of Old English poetic composition : a thesis presented in partial fulfilment of the requirements for the degree of Master of Arts in English at Massey University, Turitea, Palmerston North, New Zealand
The conventional hypothesis of Old English poetic composition assumed that poets used a limited number of formulae which were tacked together to generate the texts now extant. Old English poetic composition is a combination of natural 'spreading activation' and acquired poetic skill, with one constantly reinforcing the other. This study tests whether the spreading activation theory is correct by investigation into the formulaic properties of selected Old English texts, not including Beowulf, and compares these texts with Beowulf to see if that poem has any special non-formulaic qualities. High rates of collocation may be predictable in the abstract, but need to be established on a quantitative basis. Scholars have widely maintained that Beowulf has special qualities which give rise to the expectation that its language is inherently non-formulaic; that the Beowulf poet was a more original craftsman than the composers of other poems. Some scholars, myself included, have an intuitive feeling that the composition of Beowulf is special and superior. The proximity surveys completed for this study do not prove the hypothesis that Old English poetry was composed by a combination of natural spreading activation plus learned poetic routines. The expected outcome of patterned recurrences has not been demonstrated. The surveys suggest that poets actively suppressed the tendency for a given word to associate on to other alliterating words. This tendency is especially marked in Beowulf as poet allows one word to lead to another over a somewhat longer word-span, which results in the poet being freed to use the same basic collocation more frequently. Beowulf is shown to demonstrate a distinct and impressive lack of formulism.