Journal of the Acoustical Society of America, 10/2016, Vol.140(4), pp.3215-3215
We investigated whether listeners process suffixed words differently than prefixed words. We presented primes that combined a clear-speech root plus degraded-speech affix (such as kin-xx, where xx refers to degraded speech), and measured lexical decision RTs to subsequent clear targets ( kin-ship ). Degradation used low-pass filtering (〈 500 Hz), such that affixes were speech-like but incomprehensible. Thus, the prime kin-xx sounds like a complete suffixed word, yet it is compatible with the target kin-ship and should not compete with it for activation. The crucial comparison was between prefixed ( re-group ) versus suffixed ( kin-ship ) targets, which were matched for frequency, familiarity, probability of phonotactic transition across morpheme boundary, and affix type-parsing ratio (Hay & Baayen, 2002). On each trial, participants heard a prime, a 1000 ms ISI, then a target to which they made a speeded decision. Pilot results (n = 5) suggest that comparable speech input activated suffixed roots less strongly than prefixed roots . While xx-group reduced RTs to re-group by a large amount (-236.65 ms), kin-xx reduced RTs to kin-ship by a much smaller amount (-133.20 ms). Pseudo-affixed words ( resort, worship ) did not show a comparable difference, suggesting that the effect arises from morphological constituency.
5th Joint Meeting Of The Acoustical Society Of America And The Acoustical Society Of Japan;