No abstract

The last 40 years have witnessed a growing interest in the mechanisms underlying the visual identification of complex words. A large amount of experimental data has been amassed, but although a growing number of studies are proposing explicit theoretical models for their data, no comprehensive theory has gained substantial agreement among scholars in the field. We believe that this is due, at least in part, to the presence of several controversial pieces of evidence in the literature and, consequently, to the lack of a well-defined set of experimental facts that any theory should be able to explain. With this review, we aim to delineate the state of the art in the research on the visual identification of complex words. By reviewing major empirical evidences in a number of different paradigms such as lexical decision, word naming, and masked and unmasked priming, we were able to identify a series of effects that we judge as reliable or that were consistently replicated in different experiments, along with some more controversial data, which we have tried to resolve and explain. We concentrated on behavioral and electrophysiological studies on inflected, derived, and compound words, so as to span over all types of complex words. The outcome of this work is an analytical summary of well-established facts on the most relevant morphological issues, such as regularity, morpheme position coding, family size, semantic transparency, morpheme frequency, suffix allomorphy, and productivity, morphological entropy, and morpho-orthographic parsing. In discussing this set of benchmark effects, we have drawn some methodological considerations on why contrasting evidence might have emerged, and have tried to delineate a target list for the construction of a new all-inclusive model of the visual identification of morphologically complex words.

Numerous studies in psychology, cognitive neuroscience and psycholinguistics have used pictures of objects as stimulus materials. Currently, authors engaged in cross-linguistic work or wishing to run parallel studies at multiple sites where different languages are spoken must rely on rather small sets of black-and-white or colored line drawings. These sets are increasingly experienced as being too limited. Therefore, we constructed a new set of 750 colored pictures of concrete concepts. This set, MultiPic, constitutes a new valuable tool for cognitive scientists investigating language, visual perception, memory and/or attention in monolingual or multilingual populations. Importantly, the MultiPic databank has been normed in six different European languages (British English, Spanish, French, Dutch, Italian and German). All stimuli and norms are freely available at http://www.bcbl.eu/databases/multipic .

It is generally held that noun processing is specifically sub-served by temporal areas, while the neural underpinnings of verb processing are located in the frontal lobe. However, this view is now challenged by a significant body of evidence accumulated over the years. Moreover, the results obtained so far on the neural implementation of noun and verb processing appear to be quite inconsistent. The present review briefly describes and critically re-considers the anatomo-correlative, neuroimaging, MEG, TMS and cortical stimulation studies on nouns and verbs with the aim of assessing the consistency of their results, particularly within technique. The paper also addresses the question as to whether the inconsistency of the data could be due to the variety of the tasks used. However, it emerged that neither the different investigation techniques used nor the different cognitive tasks employed fully explain the variability of the data. In the final section we thus suggest that the main reason for the emergence of inconsistent data in this field is that the cerebral circuits underlying noun and verb processing are not spatially segregated, at least for the spatial resolution currently used in most neuroimaging studies.

A largely overlooked side result in most studies of morphological priming is a consistent main effect of semantic transparency across priming conditions. That is, participants are faster at recognizing stems from transparent sets (e.g., farm) in comparison to stems from opaque sets (e.g., fruit), regardless of the preceding primes. This suggests that semantic transparency may be also consistently associated with some property of the stem word.We propose that this property might be traced back to the consistency, throughout the lexicon, between the orthographic form of a word and its meaning, here named Orthography-Semantics Consistency (OSC), and that an imbalance in OSC scores might explain the "stem transparency" effect. We exploited distributional semantic models to quantitatively characterize OSC, and tested its effect on visual word identification relying on large-scale data taken from the British Lexicon Project (BLP). Results indicated that (a) the "stem transparency" effect is solid and reliable, insofar it holds in BLP lexical decision times (Experiment 1); (b) an imbalance in terms of OSC can account for it (Experiment 2); and (c) more generally, OSC explains variance in a large item sample from BLP, proving to be an effective predictor in visual word access (Experiment 3).Keywords: Orthography-Semantics Consistency, distributional semantic models, megastudies, visual word identification Orthography-Semantics Consistency 3Semantic transparency in free stems: the effect of Orthography-Semantics Consistency in word recognition.Recent research on morphological processing has focused on the role played by semantic transparency in the recognition of derived words. Indeed, the meaning of a derived form can be more or less associated to the meanings of its constituent morphemes: words like nameless, farmer, bakery can be easily understood given the meaning of their roots (transparent words), whereas in words like courteous, fruitful, cryptic root meanings are not fully maintained (opaque words). How early semantic transparency comes to the stage during visual word processing has been the theoretical issue at the center of a long-standing debate (for a review, see . Priming was the main methodological tool adopted in addressing this question, with the assumption that, if the recognition of a root (e.g., farm) is made quicker by the previous presentation of a related derived form (e.g., farmer ) in comparison to a control prime (e.g., speaker ), this would mean that the root is accessed when processing the derived form. Typically, transparent and opaque derived primes are compared for their effectiveness in facilitating the identification of their (pseudo-)roots. In order to rule out any strategic effect related to prime awareness, the prime is often presented very shortly and preceded by an orthographic mask, making it virtually invisible at an explicit level (masked priming; Forster & Davis, 1984). Under these conditions, priming effect is regularly observed for both transparent and opaque prime-targe...

312Previous research on the identification of morphologically complex words like player has established that such words are decomposed into their constituent morphemes (i.e., play er) during recognition. Evidence for decomposition comes largely from the findings that (1) the time taken to recognize a morphologically complex word is partly determined by the frequency of its stem (e.g., Baayen, Dijkstra, & Schreuder, 1997;Bradley, 1980;New, Brysbaert, Segui, Ferrand, & Rastle, 2004) and (2) the recognition of stem targets is speeded by the prior brief presentation of morphologically related words (e.g., Drews & Zwitserlood, 1995;Grainger, Colé, & Segui, 1991; Rastle, Davis, MarslenWilson, & Tyler, 2000) more than would be expected on the basis of pure orthographic or semantic overlap.Another well-described phenomenon used to investigate morpheme recognition is the morpheme interference effect on nonword rejection times. This effect refers to the finding that nonwords comprising existing morphemes (e.g., shootment) are rejected more slowly in lexical decision than are nonwords that do not have a morphological structure (e.g., shootmant). This result was first reported by Taft and Forster (1975), who found that nonwords composed of existing prefixes and bound stems (e.g., dejuvenate) were rejected more slowly than were nonwords composed of the same prefixes but nonexisting stems (e.g., depertoire). Caramazza, Laudanna, and Romani (1988) went on to show that Italian pseudoinflected nonwords comprising existing stems and suffixes (e.g., cant-evi, similar to buyed in English) were rejected more slowly and elicited higher error rates than did (1) nonwords comprising stems plus a nonsuffix endings (e.g., cant-ovi, buyel), (2) nonwords comprising nonstems plus suffix endings (e.g., canz-evi, biyed), and (3) nonwords comprising nonstems plus nonsuffix endings (e.g., canz-ovi, biyel). The usual explanation for this effect is that morphemic representations are activated during the processing of morphologically structured nonwords, thus slowing rejection time (Caramazza et al., 1988). In contrast to some recent models claiming that morphological processing is a postlexical phenomenon (e.g., Giraudo & Grainger, 2001), the morpheme interference effect suggests strongly that morphemic representations are activated prior to the activation of orthographic lexical entries (see also Kazanina, Dukova-Zheleva, Geber, Kharlamov, & Tonciulescu, 2008;Longtin, Segui, & Hallé, 2003;Marslen-Wilson, Bozic, & Randall, 2008;Rastle, Davis, & New, 2004;Taft, 1994).Evidence that morphologically complex words are recognized through a process of decomposition that takes place prior to the activation of orthographic lexical entries raises an important theoretical issue that has largely gone unnoticed in psycholinguistic research. Specifically, how is it that we are able to distinguish between morphologically complex stimuli comprising the same morphemes but in reversed order (e.g., preheat vs. heatpre)? This question relates to a more general issue...

Aphasic patients occasionally manifest a dissociated naming ability between objects and actions: this phenomenon has been interpreted as evidence of a separate organization for nouns and verbs in the mental lexicon. Nevertheless, Bird, Howard and Franklin (2000) suggested that the damage underlying noun-verb dissociation affects the corresponding semantic concepts and not the lexical representation of words; moreover they claimed that many dissociations reported in literature are caused merely by a strong imageability effect (Bird, Howard and Franklin, 2000; 2003). In fact, most authors used a picture naming task to assess patients' naming ability and, due to the fact that this test involves the use of pictures to represent actions and objects, nouns were frequently more imageable than verbs (Luzzatti at al., 2002). In order to overcome this drawback, we devised a new task -Nouns and Verbs Retrieval in a Sentence Context (NVR-SC)-in which nouns and verbs have the same imageability rate. Patients' performance on this task is compared with that obtained by the same patients on a standard picture naming task. Of the sixteen aphasic patients with a selective verb deficit, as revealed by the picture naming task, two continued to show dissociation in the NVR-SC task, while fourteen did not. The data indicate that at least some patients have an imageability-independent lexical deficit for verbs. The functional locus/i of the damage is also considered, with particular reference to the lemma/lexeme dichotomy suggested by Levelt et al., (1999).

Although it is widely accepted that nouns and verbs are functionally independent linguistic entities, it is less clear whether their processing recruits different brain areas. This issue is particularly relevant for those theories of lexical semantics (and, more in general, of cognition) that suggest the embodiment of abstract concepts, i.e., based strongly on perceptual and motoric representations. This paper presents a formal meta-analysis of the neuroimaging evidence on noun and verb processing in order to address this dichotomy more effectively at the anatomical level. We used a hierarchical clustering algorithm that grouped fMRI/PET activation peaks solely on the basis of spatial proximity. Cluster specificity for grammatical class was then tested on the basis of the noun-verb distribution of the activation peaks included in each cluster. Thirty-two clusters were identified: three were associated with nouns across different tasks (in the right inferior temporal gyrus, the left angular gyrus, and the left inferior parietal gyrus); one with verbs across different tasks (in the posterior part of the right middle temporal gyrus); and three showed verb specificity in some tasks and noun specificity in others (in the left and right inferior frontal gyrus and the left insula). These results do not support the popular tenets that verb processing is predominantly based in the left frontal cortex and noun processing relies specifically on temporal regions; nor do they support the idea that verb lexical-semantic representations are heavily based on embodied motoric information. Our findings suggest instead that the cerebral circuits deputed to noun and verb processing lie in close spatial proximity in a wide network including frontal, parietal, and temporal regions. The data also indicate a predominant—but not exclusive—left lateralization of the network.