Gillis Steven, Ravid Dorit
University of Antwerp, Center for Dutch Language and Speech-CNTS, Department of Linguistics, Universiteitsplein 1, B-2610 Wilrijk, Belgium.
J Child Lang. 2006 Aug;33(3):621-59. doi: 10.1017/s0305000906007434.
This study investigates the role of phonological and morphological information in children's developing orthographies in two languages with different linguistic typologies: Hebrew, a Semitic language with a highly synthetic morphology, and Dutch, a Germanic language with a sparse morphology. 192 Israeli and 192 Belgian monolingual schoolchildren in grades 1-6 (aged 6;0-12;0) were administered respective dictation tasks in which homophonous segments were the targets. In each language, these phonologically distinct segments are neutralized phonetically but are nevertheless represented orthographically. In both languages the target segments in the test words differed along two dimensions: (1) their morphological function as part of a stem or root versus as part of an affix; and (2) their morphophonological recoverability. The spelling tests in both languages consisted of four conditions which differed in the number and type of cues for retrieving the correct spelling of homophonous graphemes. The cues were of two types: morphological cues, which offer spellers clues to the correct spelling through consistent orthography/morphology mapping regularities; and morphophonological cues, which offer spellers clues to the correct spelling through the manipulation of orthography/morphophonology conversion procedures. A central finding of this study is the differential treatment of morphological cues by Dutch and Hebrew spelling learners. When faced with neutralized segments with and without morphological function, Hebrew-speaking children find morphology an enormously helpful tool. Dutch-speaking children, in contrast, do not find morphology a good cue provider. The impact of typology on the interface between spoken and written language is invoked as an explanation of the main findings.
希伯来语,一种具有高度综合形态的闪米特语;以及荷兰语,一种形态较为简单的日耳曼语。192名以色列和192名比利时的1 - 6年级单语学童(年龄在6岁0个月至12岁0个月之间)分别接受了听写任务,其中同音片段作为目标。在每种语言中,这些语音上不同的片段在语音上被中和,但在正字法上仍有体现。在这两种语言中,测试单词中的目标片段在两个维度上有所不同:(1)它们作为词干或词根的一部分与作为词缀的一部分的形态功能;(2)它们的形态语音可恢复性。两种语言的拼写测试都由四种条件组成,这些条件在检索同音素正确拼写的线索数量和类型上有所不同。线索有两种类型:形态线索,通过一致的正字法/形态映射规律为拼写者提供正确拼写的线索;以及形态语音线索,通过正字法/形态语音转换程序的操作,为拼写者提供正确拼写的线索。本研究的一个主要发现是荷兰语和希伯来语拼写学习者对形态线索的不同处理方式。当面对具有和不具有形态功能的中和片段时,说希伯来语的儿童发现形态是一个非常有用的工具。相比之下,说荷兰语的儿童并不认为形态是一个好的线索提供者。研究援引语言类型对口语和书面语言接口的影响来解释主要研究结果。