The Grey-Eyed Athena in Parliament: Policy Epithets as Compressed Narratives in the Narrative Policy Framework

The Narrative Policy Framework (NPF) has established itself as a leading approach to the empirical study of policy narratives, providing systematic tools for analysing how actors deploy stories containing characters, plots, settings, and morals to shape policy debates. Policy narratives are typically elaborated as discursive structures in which these elements are explicitly articulated. This paper argues that in some settings—especially when there is repeated engagement with the same policy actors and institutional structures over time—actors frequently deploy compressed rather than elaborated narratives. We introduce the concept of “policy epithets” to theorise this phenomenon. Inspired by oral-formulaic theory from Homeric scholarship, we define policy epithets as formulaic phrases that condense causal argument, moral evaluation, and responsibility attribution into single recognisable units. Like the Homeric epithet “swift-footed Achilles” or the “grey-eyed Athena,” which recalls a complete character without elaboration because the audience already possesses the relevant narrative knowledge, policy epithets function as retrieval devices for shared interpretive frameworks. They are characterised by repetition across instances, dependence on shared meaning within an interpretive community, and formulaic stability in phrasing. We distinguish policy epithets from adjacent concepts. Unlike frames, which structure perception, epithets invoke pre-existing structured understanding. Unlike metaphors, which invite elaboration through cross-domain mapping, epithets foreclose elaboration. Unlike slogans, which aim to persuade uncommitted audiences, epithets presuppose a shared interpretive framework among users and their primary audience. This conceptualisation has methodological implications for NPF research. Current meso-level operationalisations—particularly devil-angel shift analysis—rely on counting explicit character attributions. But when a compressed epithet performs identical argumentative work to an elaborated narrative, counting approaches systematically may undercount the narrative work being done. More automated NLP approaches, likewise, may misinterpret or misclassify tone or subject without the full elaboration that a policy epithet compresses. We explore the concept empirically through analysis of ministers’ use of “austerity by the British government” in Northern Ireland Assembly discourse since the restoration of devolved government. Using a mixed-methods design, we first map the epithet’s emergence, distribution across ministers, and portability across policy domains—establishing its formulaic properties through frequency and stability. We then analyse selected cases representing distinct deployment contexts: routine accountability, domain extension, and an instance that generated media scorn because audiences outside the Assembly’s interpretive community was not able to “unpack” the compressed narrative.