This paper attempts to review the validity of metrical evidence for secondary stress in OE (Old English). Initiated by Huguenin (1901), the tradition of reconstructing secondary stress on the basis of meter builds on the isomorphy between language and meter. Even though the prosodic reconstruction from meter has proven to be fairly useful, there are certain properties to be carefully considered in order to argue for the existence of secondary stress in OE. It is argued that verse types, not alliteration, cannot be a reliable source for the reconstruction of secondary stress. Due to the differences with regard to fundamental assumptions on OE meter, the predictions on the placement of secondary stress are not consistent. In addition, most OE metrical systems are not free of inherent circularity between verse types and secondary stress. It is also demonstrated that given the distinct nature of stress assignment, secondary stress in compounds should be distinguished from that in noncompound words. Moreover OE secondary stress cannot be claimed to exist without the precise definition between compounds and noncompound words being properly reflected in the reconstruction process.

• Chang Yong Sohn(Seoul National University)