<div>Korean phonology features a cross-linguistically rare tripartite contrast in its stop series between lax, tense, and aspirated segments. Extant evidence suggests this contrast is the result of a fifteenth-century phonological restructuring wherein tense segments, previously an allophone of lax sounds, achieved distinct phonemic status. However, the historical record suggests that almost immediately a pattern of lax segments ‘tensifying’ began, with words featuring lax onset sounds being realized increasingly with tense sounds until the novel pronunciation was universal. While the action of these shifts is sporadic throughout the lexicon, the resulting changes are unidirectional, with the domain of tense segments expanding at the cost of lax sounds. It has been posited in previous research that such sound changes may suggest a rebalancing of functional load across underutilized segments.</div><div>A similar phenomenon in contemporary Korean where speakers exhibit differing pronunciations of onset segments in a number of lexical items is analyzed herein, with the argument that it is best understood as the continuation of these historical processes. Far from an idiosyncratic speaker habit or dialectal quirk, these unexpected tense segments can be interpreted as surface evidence of phonological pressures active since late Middle Korean. The present study explored novel tensified onset pronunciations from a demographic standpoint, aiming to clarify which speaker populations have adopted new variant forms through two experiments. The first featured the elicitation of ‘tensification-prone’ items by native speakers in a production task, while the second used a combination of acceptability judgments of tensified items and attitudinal surveys regarding the use of novel tense pronunciations.</div><div>The results confirm that tensification is active in contemporary Korean, but that a decisive conclusion as to its demographic associations remains elusive. The acceptability judgment experiment suggests that younger speakers and self-affirmed dialect users are more likely to prefer tensified variants, while the production task revealed no significant relationship between these factors and actual pronunciation behavior. Finally, the findings are considered in context of deeper changes in Korean phonology whereby tense and lax segments are increasingly associated with word onset and medial/final position, respectively.</div><div><br></div>