Activation of multiple phoneme associates of graphemes in visual word recognition

One general issue in the domain of visual word recognition is to delineate the nature of readers' knowledge of the print-sound mapping. A more specific question is to determine whether multiple grapheme-phoneme associations are available and activated during the phonological transcoding of a letter string. Evidence for the activation of irregular associations during print-to-sound transcoding, independently from lexical influences, was assessed in a letter detection task by examining performance on target-absent pseudowords. We contrasted two types of pseudowords that could be considered homophone with a real word by application of either grapheme-phoneme correspondence rules or of multiple phonemic activation. Performance on both types of homophones was compared to nonhomophone control pseudowords, strictly equivalent in terms of orthographic similarity to the base words. The finding of a homophony disadvantage for the homophones by multiple activation was interpreted as evidence for multiple phonemic activation in the print-to-sound conversion system.     

Levels of representation in word processing

Abstract

Four experiments are described examining the effects of words frequency, orthographic structure, and letter spacing on a range of tasks designed to tap different levels of representation in word processing. In Experiment 1, the task was lexical decision. Effects of both frequency (high-frequency words were recognized faster than low-frequency words) and orthographic regularity (illegal non-words were rejected faster than legal non-words) were found. In Experiment 2 subjects had to detect a rotated letter within the letter strings. Effects of orthographic structure emerged, with a marked disadvantage for illegal non-words with respect to the other types of string. No difference was found among high-frequency words, lowfrequency words and legal non-words. In Experiment 3, subjects had to detect a letter elevated above the horizontal plan with respect to the rest of the string. Effects of both spatial arrangement of letters and number of letters were found (spaced strings were responded to less accurately than non-spaced strings and seven-letter extra-spaced strings were responded to slower than the other strings). Neither lexical nor orthograpic variables affected this task. In Experiment 4 subjects had to detect the presence of a bold segment contained in one of the letters in the strings. Performance was unaffected by both lexical and spatial variables. The pattern of results is discussed with reference to a multi-stage model of word recognition in which lexical and spatial variables affect processing at different stages. At a feature map level, in which features are extracted from the discontinuities of light intensities, processing is independent of both spatial and lexical factors. At a letter-shape map level, in which spatial relationships between features are coded, spacing between letters affects encoding. At a graphemic map level, in which letter identities and their relative positions within strings are coded, orthographic variables have an effect. Lexicality and frequency affect only subsequent stages of processing, when stored lexical information is retrieved (e.g. for lexical decision).     

Priming for new associations in animacy decision: Evidence for context dependency

In four experiments we investigated the context-dependent nature of semantic memory by looking at priming effects in animacy decision for newly formed associations. The first experiment investigated whether the priming effect depended on the nature of the prior relation between the word pairs. The results showed no such effect, replicating earlier findings. Experiments 2, 3, and 4 investigated the role of context overlap between study and test. In Experiment 2 priming for new associations was found only for word pairs that had been presented in the animacy decision task during study. Experiment 3 showed that in order to obtain priming effects for new associations these associations have to be studied in a study task that is aimed at unitized processing of the word pair at a semantic level. Experiment 4 showed that processing the pairs as separate words at an orthographic level cancelled the priming effect. The results are explained by assuming that priming results from the overlap of features that are activated during both study and test.     

The masked onset priming effect in naming: computation of phonology or speech planning?

We investigated factors that modulate the presence of the masked onset priming effect in three naming experiments. In Experiment 1, we showed that the masked onset priming effect is found with regular words, but not with exception words, replicating the finding reported by Forster and Davis (1991). In Experiment 2, we used the conditional naming task in which words are mixed with nonwords and participants are instructed to name the item only if it is a word. The masked onset priming effect was eliminated in this experiment, but the regularity effect remained. In Experiment 3, regular and irregular words were mixed randomly, rather than in separate blocks as in Experiment 1. This reduced the size of the regularity effect, and the masked onset priming effect was again absent. We argue that these results, taken as a whole, are better interpreted within the view that the masked onset priming effect has its origin in the preparation of a speech response, rather than within the original dual-route interpretation proposed by Forster and Davis.     

Tracking the identity of moving words: Stimulus complexity and familiarity affects tracking accuracy

In two experiments, participants tracked the identity and location of moving words. The task bears resemblance to one performed by air traffic controllers who track multiple moving aircraft, where they are identified with relatively complex alphanumeric call signs. In Experiment 1, stimulus familiarity was manipulated by comparing the tracking of familiar words and pseudowords. In Experiment 2, also stimulus complexity was varied by having the participants track short and long words. Stimulus complexity affected tracking: Short words were easier to track than long words. Moreover, familiarity of identity affected tracking of short words (short familiar words were easier to track than short pseudo-words) but not of long words. The results are interpreted within the framework of the model of multiple identity tracking model. Mathematical simulations suggest that observers may not have enough time for full identification of complex identities in a dynamic situation. Practical implications of the results for air traffic control are discussed.     

Priming of new associations in reading time: What is learned?

The priming of new associations has been a controversial topic, with some studies finding significant effects but others failing to replicate these results. Three studies investigated the priming of new associations in a reading time task, presenting lists of word pairs that were read aloud as quickly as possible. In Experiment 1, significant priming of new associations was found after two study presentations, replicating similar results by Moscovitch, Winocur, and MacLachan (1986). In Experiment 2, reading time was facilitated for intact pairs when word positions remained constant relative to when word positions were reversed. This suggested that the associative priming effect was related to specific lower level features of the word pairs rather than to abstract associations. In Experiment 3, the insertion of the wordand between test words eliminated the pairing-specific effect, placing the locus of new association priming at the transition between words within pairs. These findings demonstrate that the knowledge that supports priming of new associations in the reading time task involves perceptual or articulatory information about the transitions between words rather than abstract associative knowledge.     

Different methods of lexical access for words presented in the left and right visual hemifields

Right-handed adults were asked to identify by name bilaterally presented words and pronounceable nonwords. For words in the normal horizontal format, word length (number of letters) affected left visual hemifield (LVF) but not right visual hemifield (RVF) performance in Experiments 1, 2, 3, 5, and 6. This finding was made for words of high and low frequency (Experiment 6) and imageability (Experiment 5). It also held across markedly different levels of overall performance (Experiments 1 and 2), and across different relative positionings of short and long words in the LVF and RVF (Experiment 3). Experiment 4 demonstrated that the variable affecting LVF performance is the number of letters in a word, not its phonological length. For pronounceable nonwords (Experiment 7) and words in unusual formats (Experiment 8), however, length affected both LVF and RVF performance. The characteristics identified for RVF performance in these experiments also hold for the normal reading system. In this (normal) system the absence of length effects for horizontally formatted words is generally taken to reflect the processes involved in lexical access. Length effects in the normal reading system are thought to arise when lexical access for unusually formatted words and for the pronunciation of nonwords requires the short-term storage of information at a graphemic level of analysis. The characteristics of LVF performance indicate that horizontally formatted words presented to the right cerebral hemisphere can only achieve lexical access by a method that requires the short-term storage of graphemic information. This qualitative difference in methods of lexical access applies regardless of whether the right hemisphere is seen as accessing words in the left hemisphere's lexicon or words in a lexicon of its own.     

The effect of phonemic processing on the retention of graphemic representations for words and nonwords

The three experiments reported in this study were each conducted in two phases. The first phase of Experiment 1 involved a same-different comparison task requiring “same” responses for both mixed-case (e.g., MAIN main) and pure-case (e.g., near near) pairs. This was followed by Phase 2, a surprise recognition test in which a graphemic effect on word retention was indicated by the superior recognition accuracy obtained for pure-case compared with mixed-case pairs. The first phases of Experiments 2 and 3 involved pronounceability and imageability judgment tasks, respectively. Graphemic retention was assessed by contrasting recognition accuracy for letter strings presented, during Phase 2, in their original Phase 1 case, with letter strings presented, during Phase 2, in. a graphemically dissimilar new case. The experiments provided evidence that there was minimal retention of the graphemic representations from which the phonemic representations of words are generated and, further, that the locus of this effect is probably postlexical. Nonwords were recognized more accurately than words in all three experiments. The latter result was attributed to differences between nonwords and words in both graphemic retention and semantic distinctiveness.     

Control of stimulus-response translation in dual-task performance

In contradiction to stimulus-response- (S-R-) translation bottleneck models of dual-task control, stimulus processing in a primary task is affected by its compatibility with the response in a secondary, later performed task (Hommel, 1998a)- an indication of parallel S-R translation. Here we show that this backward-compatibility effect is independent of working-memory load, whether this is induced by an extra memory task (Experiment 1) or by increasing the number of S-R alternatives in the primary task (Experiment 2). However, backward effects occur even when the secondary task is no longer carried out (Experiment 3) and they are strongly affected by the inconsistency of previously used S-R mappings (Experiment 4). These findings suggest that S-R translation is (or can be) capacity-independent and automatic even under multiple-task conditions, and that it is mediated by direct S-R associations that emerge after only little practice.     

The relationship between the number of presented stimuli and prospective duration estimates: The effect of concurrent task activity

The attention/distraction models of prospective time estimation predict either a negative relationship or an independence between duration estimates and the number of stimulus events presented during the time period, but not a positive relationship. Two experiments examined this relationship under prospective conditions. Lists of words were presented during a 60-sec time period at either a fast or a slow rate. Subjects either passively viewed the words or actively responded by performing either a graphemic or a semantic classification task on each word. To measure subjective duration, the method of magnitude estimation was used in Experiment 1 and the method of reproduction was used in Experiment 2. Time estimates were independent of the number of presented words in the passive viewing condition, and negatively related in the classification task conditions. Furthermore, the two classification tasks had similar effects on time estimates. These findings are consistent with the attentional models of prospective timing, and they question the robustness of earlier findings of a positive relationship between time estimates of moderately long intervals and the number of presented stimuli under conditions of minimal processing of the interval events.     

The use of abstract graphemic information in lexical access

Three experiments investigated the nature of the information required for the lexical access of visual words. A four-field masking procedure was used, in which the presentation of consecutive prime and target letter strings was preceded and followed by presentations of a pattern mask. This procedure prevented subjects from identifying, and thus intentionally using, prime information. Experiment I extablished the existence of a semantic priming effect on target identification, demonstrating the lexical access of primes under these conditions. It also showed a word repetition effect independent of letter case. Experiment II tested whether this repetition effect was due to the activation of graphemic or phonemic information. The graphemic and phonemic similarity of primes and targets was varied. No evidence for phonemic priming was found, although a graphemic priming effect, independent of the physical similarity of the stimuli, was obtained. Finally Experiment III demonstrated that, irrespective of whether the prime was a word or a nonword, graphemic priming was equally effective. In both Experiments II and III, however, the word repetition effect was stronger than the graphemic priming effect. It is argued that facilitation from graphemic priming was due to the prime activating a target representation coded for abstract (non-visual) graphemic features, such as letter identities. The extra facilitation from same identity priming was attributed to semantic as well as graphemic activation. The implications of these results for models of word recognition are discussed.     

Orthographic complexity and word naming in Italian: Some words are more transparent than others

Italian is a language with a transparent orthography in which printed words can be translated into the correct sequence of phonemes using a limited set of rules. The rules of letter—sound conversion are, however, simpler for some letters than for others: The pronunciations of sequences involving the lettersc andg are determined by complex (i.e., context-sensitive) rules that depend on the letters that follow them. In this article, we report two experiments in which Italian participants read aloud words containing simple or complex letter—sound conversion rules. In Experiment 1, we found that words containing complex rules were read more slowly than words containing simple, noncontextual rules. In Experiment 2, we found that the effect of rule complexity on naming speed held for low-frequency words but not high-frequency words. The results are interpreted in terms of a dual-route model in which rule complexity effects arise from sublexical procedures that are more involved in reading low-frequency words than high-frequency words. The experimental material used in this study may be downloaded from www.psychonomic.org/archive.     

The processing of interlexical homographs in translation recognition and lexical decision: Support for non-selective access to bilingual memory

In three experiments we looked at the processing of interlexical homographs by Dutch-English bilinguals. In Experiment 1 we employed the translation recognition task, a task that forces the participants to activate both language systems simultaneously. In this task the processing of interlexical homographs was inhibited substantially compared to the processing of matched control words, especially when the homograph reading to be selected was the less frequent of the homograph's two readings. In Experiments 2 and 3 we used the lexical decision task: In one condition we asked the participants to categorize letter strings as words or nonwords in Dutch; in a second condition we asked them to do so in English. The makeup of the stimulus set in Experiment 2 permitted the participants to ignore the instructions and to instantiate the task in a language-neutral form-that is, to categorize the letter strings as words in either Dutch or English. Under these circumstances a small, frequency-dependent inhibitory effect for homographs was obtained, but only in condition Dutch. In Experiment 3 the participants were forced in a language-specific processing mode by the inclusion of "nonwords" that were in fact words in the non-target language. Large frequency-dependent inhibitory effects for homographs were now obtained in both language conditions. The combined results are interpreted as support for the view that bilingual lexical access is non-selective.     

Number-of-features effects and semantic processing

Participants list many semantic features for some concrete nouns (e.g., lion) and fewer for others (e.g., lime; McRae, de Sa, & Seidenberg, 1997). Pexman, Lupker, and Hino (2002) reported faster lexical decision and naming responses for high number of features (NOF) words than for low-NOF words. In the present research, we investigated the impact of NOF on semantic processing. We observed NOF effects in a self-paced reading task when prior context was not congruent with the target word (Experiment 1) and in a semantic categorization task (concrete vs. abstract; Experiment 2A). When we narrowed our stimuli to include high- and low-NOF words from a single category (birds), we found substantial NOF effects that were modulated by the specificity of the categorization task (Experiments 3A, 3B, and 3C). We argue that these results provide support for distributed representation of word meaning.     

Redefining the regularity effect

This study reports two experiments that further explore the regularity effect in single-word pronunciation. Experiment 1 shows that regularity effects are found only with irregular words that are “true” exceptions (e.g., PINT, MONK, BROAD). Words that are irregular in terms of grapheme-to-phoneme correspondence rules, but either are regular in terms of a higher order correspondence rule (e.g., PALM, HEALTH) or possess a divergent, although reasonably common, correspondence (e.g., GLOVE, HEAD), produce response times similar to those produced by regular words. These results indicate that the regularity effect is restricted to a smaller set of words than previously has been thought. Experiment 2 examines the regularity effect when subjects are required to delay their responses by 1,500 msec; no difference between exception and regular words is found. This finding indicates that previous demonstrations of the regularity effect cannot be attributed to articulatory differences between exception and regular word samples. Theoretical accounts of the regularity effect are considered briefly.     

Case-Mixing Effects on Spelling Recognition: The Importance of Test Format

In a multiple-choice spelling recognition test, 56 university students were more accurate on more regular than irregular words, and on lower-case than mixed-case words, with the case mixing effect greater for irregular than regular words. In Experiment 2, the same words were presented singly in correct or incorrect spellings and distortion of word shape was achieved by case mixing (32 subjects) or by alternating the size of lower-case letters within a word (32 subjects). The main effects of regularity and distortion were replicated and the effect of distortion was greater for incorrect than correct stimuli, with correctly spelled words suffering a decrement in accuracy of less than 5 percentage points. Case mixing had a greater effect than size mixing on response latencies. In Experiment 3, with comparable test procedures, case mixing interacted with regularity in the subjects analysis for the multiple choice format, but not the single presentation format. This result indicates that comparisons based on visual configuration may be an artifact of multiple-choice tests.     

Dimensions of lexical coding in Chinese and English

Phonology and orthography are closely related in some languages, such as English, and they are nearly unrelated in others, such as Chinese. The effects of these differences were assessed in a study of the roles of phonemic, graphemic, and semantic information on lexical coding and memory for Chinese logographs and English words. Some of the stimuli in the two languages were selected such that the natural confounding between phonemic and graphemic information in English was matched in the set of Chinese words used. An initial scaling study indicated that this attempt to equate degree of phonemic-graphemic confounding was successful. A second experiment used a recognition memory task for English and Chinese words with separate subject groups of native speakers of the two languages. Subjects were to select one of a pair of test words that was phonemically, graphemically, or semantically similar to a word on a previously studied list. Differences in the dimensions of lexical coding in memory were demonstrated in significant Stimulus Type by Decision Type interactions in the recognition data. Chinese-speaking subjects responded most rapidly and accurately in the graphemic recognition task, whereas performance was generally equivalent in all three tasks for the English-speaking subjects. Alphabetic and logographic writing systems apparently activate different coding and memory mechanisms such that logographic characters produce significantly more visual information in memory, whereas alphabetic words result in a more integrated code involving visual, phonological, and semantic information.     

Implicit memory for novel conceptual associations in amnesia

In two experiments, we evaluated the status of implicit memory for novel associations in amnesia. Experiment 1 assessed priming in a category exemplar generation task in which contextual information associated with a target could increase the likelihood of target generation. Control participants, but not amnesic patients, showed associative priming. Amnesics’ impairment was not due to the use of explicit memory by control subjects but reflected a genuine impairment in implicit memory for novel conceptual associations. Experiment 2 assessed priming in a relatedness judgment task, in which associative priming was manifest as longer latencies for old than for recombined pairs of unrelated words. Amnesic patients showed intact associative priming in this task. We discuss differences in the status of implicit memory for novel conceptual associations in amnesia, with reference to the nature of the representation that supports priming in the two tasks and the type of processing that is required at test.     

Integrative and semantic relations equally alleviate age-related associative memory deficits

Two experiments compared effects of integrative and semantic relations between pairs of words on lexical and memory processes in old age. Integrative relations occur when two dissimilar and unassociated words are linked together to form a coherent phrase (e.g., horse-doctor). In Experiment 1, older adults completed a lexical-decision task where prime and target words were related either integratively or semantically. The two types of relation both facilitated responses compared to a baseline condition, demonstrating that priming can occur in older adults with minimal preexisting associations between primes and targets. In Experiment 2, young and older adults completed a cued recall task with integrative, semantic, and unrelated word pairs. Both integrative and semantic pairs showed significantly smaller age differences in associative memory compared to unrelated pairs. Integrative relations facilitated older adults' memory to a similar extent as semantic relations despite having few preexisting associations in memory. Integratability of stimuli is therefore a new factor that reduces associative deficits in older adults, most likely by supporting encoding and retrieval mechanisms.     

Naming cAsE aLtErNaTeD words

A dual-route approach was used as an initial framework to examine the relation between presentation format and lexical processing in a naming task. In Experiments 1 and 3, words were presented in lowercase versus case-alternated format. Presentation format interacted with word frequency and regularity: For irregular words (e.g., pint), case alternation was additive with frequency, whereas for regular words (e.g., mint), case alternation and frequency interacted. Experiment 2 dissociated the locus of case-alternation effects from those of stimulus intensity. Stimulus intensity was additive with frequency and regularity, suggesting that whereas stimulus intensity affects encoding, case alternation affects lexical processing at a postencoding stage in the word recognition system. It was concluded that a dual-route approach provides a suggestive but incomplete account of how case alternation influences lexical processing. As an alternative to a dual-route approach, we show that the present results can be addressed and successfully simulated using an implemented version of Norris's (1994) multilevel model.