Fast automatic translation and morphological decomposition in Chinese-English bilinguals

In this study, we investigated automatic translation from English to Chinese and subsequent morphological decomposition of translated Chinese compounds. In two lexical decision tasks, Chinese-English bilinguals responded to English target words that were preceded by masked unrelated primes presented for 59 ms. Unbeknownst to participants, the Chinese translations of the words in each critical pair consisted of a fully opaque compound word (i.e., a compound with two constituent morphemes that were semantically unrelated to the compound) and a monomorphemic word that was either the first or the second morpheme of the compound. The data revealed that bilinguals responded faster to English word pairs whose Chinese translations repeated the first morpheme than to English word pairs whose Chinese translations did not repeat the first morpheme, but no effect of hidden second-morpheme repetition was found. This effect of hidden first-morpheme repetition suggests that participants translated English words to Chinese and decomposed the translated compounds into their constituent morphemes. Because the primes were presented for only 59 ms, translation and morphological decomposition must be fast and automatic.     

重复启动作业中词的语义透明度的作用

用词素重复启动的方法考察合成词的心理表征。目标刺激为２４个语义透明度词、２４个不透明词以及４８个非词（都是两个字）。启动刺激有５种：（１）目标词本身,（２）目标词的第一词素,（３）目标词的第二词素,（４）由目标词的两个词素颠倒后形成的非词,（５）包含目标词的一个词素的非词。启动刺激的呈现时间是１００毫秒,随后是目标刺激４００毫秒。被试的任务是尽快准确地按键判断目标刺激是否是一个词。研究发现,（２）、（３）、（５）条件下的重复启动中,透明度表现出显著的效应：透明词的启动量显著大于不透明词。结果支持合成词的表征受语义透明度的影响,透明词和不透明词的词素和整词在心理表征中的关系可能不同：前者是一种兴奋性的连接,后者是一种抑制性的连接。     

Compound fracture: the role of semantic transparency and morphological headedness

This paper explores the role of semantic transparency in the representation and processing of English compounds. We focus on the question of whether semantic transparency is best viewed as a property of the entire multimorphemic string or as a property of constituent morphemes. Accordingly, we investigated the processing of English compound nouns that were categorized in terms of the semantic transparency of each of their constituents. Fully transparent such as bedroom are those in which the meanings of each of the constituents are transparently represented in the meaning of the compound as a whole. These compounds were contrasted with compounds such as strawberry, in which only the second constituent is semantically transparent, jailbird, in which only the first constituent is transparent, and hogwash, in which neither constituent is semantically transparent. We propose that significant insights can be achieved through such analysis of the transparency of individual morphemes. The two experiments that we report present evidence that both semantically transparent compounds and semantically opaque compounds show morphological constituency. The semantic transparency of the morphological head (the second constituent in a morphologically right-headed language such as English) was found to play a significant role in overall lexical decision latencies, in patterns of decomposition, and in the effects of stimulus repetition within the experiment.     

Morphemes in their place: Evidence for position-specific identification of suffixes

Previous research strongly suggests that morphologically complex words are recognized in terms of their constituent morphemes. A question thus arises as to how the recognition system codes for morpheme position within words, given that it needs to distinguish morphological anagrams like overhang and hangover. The present study focused specifically on whether the recognition of suffixes occurs in a position-specific fashion. Experiments 1 and 2 revealed that morphologically complex nonwords (gasful ) are rejected more slowly than orthographic controls (gasfil ) but that the same interference effect is not present when the morphemic constituents are reversed ( fulgas vs. filgas). Experiment 3 went further in demonstrating that reversing the morphemes within words (e.g., nesskind) does not yield morpheme interference effects against orthographic controls (e.g., nusskind). These results strongly suggest that suffix identification is position specific, which imposes important constraints on the further development of models of morphological processing.     

词汇识别中歧义词素语义加工:ERP研究

结合掩蔽启动范式和ERP技术,利用歧义词素的加工,本研究探讨了歧义词素语义的激活。实验采用3×2两因素被试内设计,歧义词素两含义互为启动刺激和目标刺激,两变量分别为启动词类型(歧义词素含义1 vs.含义2 vs.无关词)和目标词类型(歧义词素含义1 vs.含义2)。行为结果发现启动词类型和目标词类型存在交互作用,即只有当共享词素同形同义的情况下,目标词的加工才可能被启动,共享词素同形不同义时并没有产生启动效应。ERP结果进一步肯定了行为结果,在N250和N400两个脑电成分上都发现了启动类型和目标类型交互作用有关的效应,当启动词和目标词共享词素同形同义时,同时引发了N250和N400的变化;当共享词素同形不同义时,时程分析时虽然也发现了N250的变化,但是N250的效应较小,小于同形同义所引发的N250效应,持续时间也较短。实验结果说明,在词汇识别中歧义词素语义可以激活,并对目标词汇加工的早期加工阶段发挥作用。     

汉语并列式合成词的词汇通达

汉语并列式合成词具有三种类型,这三种类型内部不同的词素组合在心理词典中则需要不同的词汇通达时间。实验结果表明,在对汉语并列式合成词的词汇通达过程中,第一成分并不总是起决定作用;同样,并列式合成词中的两个成分在词汇通达过程中并不总是同等重要。实验证明,语义透明度对并列式合成词的词汇通达起重要作用。实验还发现,在心理词典中对并列式合成词的词汇通达需要一个证实原则。     

言语产生中的词长效应

采用图片命名方法考察言语产生中的词长效应,并进而分析语音计划的单位和序列性。选取能同时用单音节(如“羊”)和双音节(“山羊”)命名的图片作刺激材料,发现双音节命名明显慢于单音节命名,这种效应既表现在首音节型(如“花”与“花朵”),也表现在尾音节型(如“羊”与“山羊”)刺激材料上。研究结果支持语音编码从左到右、系列提取和双音节词的词音计划单位大于单音节词的观点。该文还对图片命名中的频率效应进行了事后分析。     

Processing visual words with numbers: Electrophysiological evidence for semantic activation

Perea, Duñabeitia, and Carreiras (Journal of Experimental Psychology: Human Perception and Performance 34:237–241, 2008) found that LEET stimuli, formed by a mixture of digits and letters (e.g., T4BL3 instead of TABLE), produced priming effects similar to those for regular words. This finding led them to conclude that LEET stimuli automatically activate lexical information. In the present study, we examined whether semantic activation occurs for LEET stimuli by using an electrophysiological measure called the N400 effect. The N400 effect, also known as the mismatch negativity, reflects detection of a mismatch between a word and the current semantic context. This N400 effect could occur only if the LEET stimulus had been identified and processed semantically. Participants determined whether a stimulus (word or LEET) was related to a given category (e.g., APPLE or 4PPL3 belongs to the category “fruit,” but TABLE or T4BL3 does not). We found that LEET stimuli produced an N400 effect similar in magnitude to that for regular uppercase words, suggesting that LEET stimuli can access meaning in a manner similar to words presented in consistent uppercase letters.     

Parafoveal processing during reading is reduced across a morphological boundary

A boundary change manipulation was implemented within a monomorphemic word (e.g., fountaom as a preview for fountain), where parallel processing should occur given adequate visual acuity, and within an unspaced compound (bathroan as a preview for bathroom), where some serial processing of the constituents is likely. Consistent with that hypothesis, there was no effect of the preview manipulation on fixation time on the 1st constituent of the compound, whereas there was on the corresponding letters of the monomorphemic word. There was also a larger preview disruption on gaze duration on the whole monomorphemic word than on the compound, suggesting more parallel processing within monomorphemic words.     

A funny thing happened on the way to articulation: N400 attenuation despite behavioral interference in picture naming

We measured Event-Related Potentials (ERPs) and naming times to picture targets preceded by masked words (stimulus onset asynchrony: 80 ms) that shared one of three different types of relationship with the names of the pictures: (1) Identity related, in which the prime was the name of the picture ("socks" - ), (2) Phonemic Onset related, in which the initial segment of the prime was the same as the name of the picture ("log" - ), and (3) Semantically related in which the prime was a co-category exemplar and associated with the name of the picture ("cake" - ). Each type of related picture target was contrasted with an Unrelated picture target, resulting in a 3×2 design that crossed Relationship Type between the word and the target picture (Identity, Phonemic Onset and Semantic) with Relatedness (Related and Unrelated). Modulation of the N400 component to related (versus unrelated) pictures was taken to reflect semantic processing at the interface between the picture's conceptual features and its lemma, while naming times reflected the end product of all stages of processing. Both attenuation of the N400 and shorter naming times were observed to pictures preceded by Identity related (versus Unrelated) words. No ERP effects within 600 ms, but shorter naming times, were observed to pictures preceded by Phonemic Onset related (versus Unrelated) words. An attenuated N400 (electrophysiological semantic priming) but longer naming times (behavioral semantic interference) were observed to pictures preceded by Semantically related (versus Unrelated) words. These dissociations between ERP modulation and naming times suggest that (a) phonemic onset priming occurred late, during encoding of the articulatory response, and (b) semantic behavioral interference was not driven by competition at the lemma level of representation, but rather occurred at a later stage of production.     

Lexicality and stimulus length effects in Italian dyslexics: role of the overadditivity effect.

The effect of lexicality and stimulus length was studied in 32 third- and fourth-grade Italian dyslexics and in 86 age-matched controls. A visual lexical decision task was used. As proposed by Faust et al. (1999), the results were analyzed in terms of raw reaction time (RT) data and using the z-score transformation to control for the presence of overadditivity effects. In terms of RTs, dyslexics showed a larger difference between words and nonwords (lexicality effect) and between short and long stimuli (length effect) than proficient readers. When data were transformed into z scores, only the group by length interaction remained significant while that with lexicality vanished. This pattern indicates that stimulus length has a specific role in Italian dyslexics' reading deficit; in contrast, slowness in responding to nonwords was not specific but was interpreted as one aspect of dyslexics' general inability to deal with alphabetical material (overadditivity effect).     

Friendly drug-dealers and terrifying puppies: Affective primacy can attenuate the N400 effect in emotional discourse contexts

Words that are semantically congruous with their preceding discourse context are easier to process than words that are semantically incongruous with their context. This facilitation of semantic processing is reflected by an attenuation of the N400 event-related potential (ERP). We asked whether this was true of emotional words in emotional contexts where discourse congruity was conferred through emotional valence. ERPs were measured as 24 participants read two-sentence scenarios with critical words that varied by emotion (pleasant, unpleasant, or neutral) and congruity (congruous or incongruous). Semantic predictability, constraint, and plausibility were comparable across the neutral and emotional scenarios. As expected, the N400 was smaller to neutral words that were semantically congruous (vs. incongruous) with their neutral discourse context. No such N400 congruity effect was observed on emotional words following emotional discourse contexts. Rather, the amplitude of the N400 was small to all emotional words (pleasant and unpleasant), regardless of whether their emotional valence was congruous with the valence of their emotional discourse context. However, consistent with previous studies, the emotional words produced a larger late positivity than did the neutral words. These data suggest that comprehenders bypassed deep semantic processing of valence-incongruous emotional words within the N400 time window, moving rapidly on to evaluate the words’ motivational significance.     

Further study of the electrophysiological correlates of lexical decision

Discriminations were required between words, pseudohomophones, and visually matched nonwords. Two tasks were employed, one which could be accomplished on the basis of a visual code (the REAL task, involving discrimination between words and both types of nonword) and another necessitating the use of a phonological code (the REAL/PSEUD task, words and pseudohomophones vs. nonwords). ERPs were recorded from three midline sites and from left and right inferior parietal sites. Two principal results were observed, (i) the peak latency of a late positive component, P637, covaried with RT, with variations in latency of around one half the corresponding RT variations, and (ii) the peak-to-peak amplitude of N100-P187 interacted with stimulus and task, such that it was larger for nonwords in the REAL task and words in the REAL/PSEUD task. No taskor stimulus-dependent asymmetries were observed in any ERP component. The P637 latency data support a model of RT variation based on the interaction of changes in parallel response preparation and stimulus evaluation processes. The observations with respect to N100-P187 suggest that ERPs are sensitive to factors related to the early processing of words and word-like visual material.     

Holistic versus analytic processing: Evidence for a different approach to processing of Chinese at the word and character levels in Chinese children

Among 30 Hong Kong Chinese fourth graders, sensitivities to character and word constructions were examined in judgment tasks at each level. There were three conditions across both tasks: the real condition, consisting of either actual two-character compound Chinese words or real Chinese compound characters; the reversed condition, with either the order of the two morphemes in the word or the order of the semantic and phonetic radicals in the character reversed; and the random condition, with either two morphemes randomly combined to create arbitrary nonwords or a semantic radical and a phonetic radical randomly combined using correct orthographic rules to create noncharacters. On the word level, children performed worst in the reversed condition and best with real words, whereas on the character level, children performed worst in the random condition and best in the reversed condition. Findings imply a holistic tendency toward word-level processing and an analytic approach to character-level processing in these Chinese children.     

Lexical decision and the number of morphemes and affixes

There has been a considerable amount of research looking at the effects of both syllable number and syllable frequency on lexical decision and word naming times. Recently, there has also been an increased interest in morphological variables, but there have been no large scale studies that have examined the role of the number of morphemes in lexical decision for nonwords. This is partly because of the difficulty of identifying morphemes in nonwords. We present a program that identifies the presence of affixes and, therefore, can be used to count the number of morpheme‐like elements in a nonword. We then used the program to measure the importance of affixes/morphemes in predicting lexical decision in nonwords. The results suggested that morphemes have an important role in lexical decision for both words and nonwords.     

言语产生中双词素词的语音编码

采用同音判断和音节监控方法 ,考察言语产生中双词素词语音激活的特点。选用以偏正结构的双词素词为名称的图片和与双词素词中首尾两个词素同音的两组探测字作为实验材料 ,探测字在图片呈现后 5 0毫秒 (实验一 ) ,或图片呈现前 130 0毫秒 (实验二 )出现 ,要求被试判断出现的字是否与图片名称中任何一个词素同音。实验结果与绝大部分言语产生理论的预期相反 ,对图片名称第一词素的反应慢于第二词素。考虑到首尾两个词素对整词意义贡献的差别 ,作者把实验结果解释为词素意义对语音激活的作用 :第二词素的语义重要性决定了对应音节激活的快速性 ,双词素词中词素音位激活的速度和时间性受词义和词素意义激活程度的影响 ,而不完全取决于词素发音的序列性。实验不支持音位编码从左到右、序列进行的观点     

Decomposition: to what extent? The case of Turkish.

It has been proposed that in agglutinative languages, lexical access of morphologically complex words must involve decomposition rather than full listing (Frauenfelder & Schreuder, 1992; Hankamer, 1989). We tested this proposal in Turkish using a simple lexical decision task. Results show that multimorphemic words that consist of frequent affixes are processed as fast as monomorphemic words. This finding suggests that in languages with rich morphology, not all multimorphemic words are accessed in a decomposed form. To the extent that morphemes are in frequent use, they may induce whole-word rather than decompositional lexical access.     

Lexical access to inflected words as measured by lateralized visual lexical decision

In two lateralized visual lexical decision experiments conducted with normal subjects, we studied hemispheric performance in the recognition of case-inflected Finnish nouns. Previous research employing mainly locative cases has indicated that such noun forms undergo morphological decomposition. The present experiments extend this finding to syntactic cases by showing that nouns with partitive or genitive endings take longer to recognize and elicit more errors than otherwise comparable monomorphemic nominative singular nouns. Morpheme-based recognition of all case-inflected forms would be a particularly appropriate solution for mental lexicon in highly inflecting languages like Finnish: it saves storage space and enables fast recognition of inflected forms not encountered before. In real words, morphological structure did not interact with visual field. However, particularly demanding, morphologically decomposable nonwords elicited more errors in the left visual field/right hemisphere than did nondecomposable nonwords. Our results suggest that at least in Finnish, both hemispheres are capable of morpheme-based access, but this mechanism is more accurate in the left than in the right hemisphere. Received: 21 May 1997 / Accepted: 19 January 1998     

Rhyme decisions to spoken words and nonwords

Lexical effects in auditory rhyme-decision performance were examined in three experiments. Experiment 1 showed reliable lexical involvement: rhyme-monitoring responses to words were faster than rhyme-monitoring responses to nonwords; and decisions were faster in response to high-frequency as opposed to low-frequency words. Experiments 2 and 3 tested for lexical influences in the rejection of three types of nonrhyming item: words, nonwords with rhyming lexical neighbors (e.g.,jop after the cuerob), and nonwords with no rhyming lexical neighbor (e.g.,vop afterrob). Words were rejected more rapidly than nonwords, and there were reliable differences in the speed and accuracy of rejection of the two types of nonword. The advantage for words over nonwords was replicated for positive rhyme decisions. However, there were no differences in the speed of acceptance, as rhymes, of the two types of nonword. The implications of these results for interactive and autonomous models of spoken word recognition are discussed. It is concluded that the differences in rejection of nonrhyming nonwords are due to the operation of a guessing strategy.     

Event-related brain potentials to grammatical errors and semantic anomalies

Event-related brain potentials (ERPs) were recorded while subjects silently read several prose passages, presented one word at a time. Semantic anomalies and various grammatical errors had been inserted unpredictably at different serial positions within some of the sentences. The semantically inappropriate words elicited a large N400 component in the ERP, whereas the grammatical errors were associated with smaller and less consistent components that had scalp distributions different from that of the N400. This result adds to the evidence that the N400 wave is more closely related to semantic than to grammatical processing. Additional analyses revealed that different ERP configurations were elicited by open-class (“content”) and closed-class (“function”) words in these prose passages.