Zhao Xu, Berent Iris
Department of Psychology, Northeastern University, 125 Nightingale Hall, 360 Huntington Avenue, Boston, MA, 02115, USA.
J Psycholinguist Res. 2016 Aug;45(4):795-811. doi: 10.1007/s10936-015-9375-1.
Across languages, certain onset clusters are systematically preferred (e.g., [Formula: see text], "[Formula: see text]" indicates preference), and speakers extend these preferences even to onsets that are unattested in their language. All such demonstrations, however, come from cluster-rich languages, so the observed preferences could reflect not universal linguistic restrictions but lexical analogy. To address this possibility, here, we turn to Mandarin Chinese-a cluster-poor language. We reasoned that, if people are sensitive to the onset hierarchy, then they should repair ill-formed onsets as better-formed ones [Formula: see text]-the worse formed the onset, the more likely its repair, hence, its misidentification. Results were consistent with this hypothesis, and they obtained irrespective of participants' experience with their second language (English). Nonetheless, the effect of syllable structure was strongly modulated by phonetic cues and task demands. These findings suggest that speakers might share broad phonological restrictions, but phonetic factors play a major role in their detection.
在各种语言中,某些开头音丛会被系统地优先选择(例如,[公式:见正文],“[公式:见正文]”表示优先选择),并且说话者会将这些偏好扩展到他们语言中未出现过的开头音丛。然而,所有这些例证都来自于音丛丰富的语言,所以观察到的偏好可能反映的不是普遍的语言限制,而是词汇类推。为了探究这种可能性,我们在此转向汉语普通话——一种音丛较少的语言。我们推断,如果人们对开头音丛层级敏感,那么他们应该会将不合规范的开头音丛修正为更规范的开头音丛[公式:见正文]——开头音丛越不规范,其被修正的可能性就越大,因此,其被误认的可能性也越大。结果与这一假设一致,并且无论参与者的第二语言(英语)经验如何,结果都是如此。尽管如此,音节结构的影响受到语音线索和任务要求的强烈调节。这些发现表明,说话者可能共享广泛的音系限制,但语音因素在其识别中起主要作用。