This paper presents the results of a comprehensive meta-analysis of the relevant imaging literature on word production (82 experiments). In addition to the spatial overlap of activated regions, we also analyzed the available data on the time course of activations. The analysis specified regions and time windows of activation for the core processes of word production: lexical selection, phonological code retrieval, syllabification, and phonetic/articulatory preparation. A comparison of the word production results with studies on auditory word/non-word perception and reading showed that the time course of activations in word production is, on the whole, compatible with the temporal constraints that perception processes impose on the production processes they affect in picture/word interference paradigms. q 2004 Elsevier B.V. All rights reserved.Keywords: Language; Word production; Word perception; Neuroimaging The functional organization of word productionProducing spoken words, whether in isolation or in the context of a larger utterance, involves an extensive neural network. In a recent meta-analysis of 58 neuroimaging studies of word production, Indefrey and Levelt (2000) found that the main components of this network, defined as cerebral regions showing statistically distinguishable activation in word production tasks, are largely left-lateralized. In addition, different regions of activation appeared to be involved with different functional components of the word production process. For instance, the conceptually driven selection of a lexical item, as in picture naming, typically goes with activation in the mid part of the left middle temporal 0022-2860/$ -see front matter q

In the first decade of neurocognitive word production research the predominant approach was brain mapping, i.e., investigating the regional cerebral brain activation patterns correlated with word production tasks, such as picture naming and word generation. Indefrey and Levelt (2004) conducted a comprehensive meta-analysis of word production studies that used this approach and combined the resulting spatial information on neural correlates of component processes of word production with information on the time course of word production provided by behavioral and electromagnetic studies. In recent years, neurocognitive word production research has seen a major change toward a hypothesis-testing approach. This approach is characterized by the design of experimental variables modulating single component processes of word production and testing for predicted effects on spatial or temporal neurocognitive signatures of these components. This change was accompanied by the development of a broader spectrum of measurement and analysis techniques. The article reviews the findings of recent studies using the new approach. The time course assumptions of Indefrey and Levelt (2004) have largely been confirmed requiring only minor adaptations. Adaptations of the brain structure/function relationships proposed by Indefrey and Levelt (2004) include the precise role of subregions of the left inferior frontal gyrus as well as a probable, yet to date unclear role of the inferior parietal cortex in word production.

A hallmark of human language is that we combine lexical building blocks retrieved from memory in endless new ways. This combinatorial aspect of language is referred to as unification. Here we focus on the neurobiological infrastructure for syntactic and semantic unification. Unification is characterized by a high-speed temporal profile including both prediction and integration of retrieved lexical elements. A meta-analysis of numerous neuroimaging studies reveals a clear dorsal/ventral gradient in both left inferior frontal cortex and left posterior temporal cortex, with dorsal foci for syntactic processing and ventral foci for semantic processing. In addition to core areas for unification, further networks need to be recruited to realize language-driven communication to its full extent. One example is the theory of mind network, which allows listeners and readers to infer the intended message (speaker meaning) from the coded meaning of the linguistic utterance. This indicates that sensorimotor simulation cannot handle all of language processing. 347

This article presents the results of a meta-analysis of 30 hemodynamic experiments comparing first language (L1) and second language (L2) processing in a range of tasks. The results suggest that reliably stronger activation during L2 processing is found (a) only for task-specific subgroups of L2 speakers and (b) within some, but not all regions that are also typically activated in native language processing. A tentative interpretation based on the functional roles of frontal and temporal regions is suggested.In recent years, there has been an increasing number of neurocognitive studies investigating language processing in bilingual speakers. Most researchers interested in language are aware of one or another study reporting hemodynamic activation differences between first language (L1) and second language (L2) processing. Given the plethora of experimental details that might lead to signal changes in hemodynamic experiments, however, differences as such might not mean very much as long as they do not overlap across studies with similar paradigms and as long as it is not clear which factors determine the presence or absence

This study investigates whether advanced second language (L2) learners of a nonnull subject language (Dutch) are influenced by their null subject first language (L1) (Turkish) in their offline and online resolution of subject pronouns in L2 discourse. To tease apart potential L1 effects from possible general L2 processing effects, we also tested a group of German L2 learners of Dutch who were predicted to perform like the native Dutch speakers. The two L2 groups differed in their offline interpretations of subject pronouns. The Turkish L2 learners exhibited a L1 influence, because approximately half the time they interpreted Dutch subject pronouns as they would overt pronouns in Turkish, whereas the German L2 learners performed like the Dutch controls, interpreting pronouns as coreferential with the current discourse topic. This L1 effect was not in evidence in eyetracking data, however. Instead, the L2 learners patterned together, showing an online processing disadvantage when two potential antecedents for the pronoun were grammatically available in the discourse. This processing disadvantage was in evidence irrespective of the properties of the learners' L1 or their final interpretation of the pronoun. Therefore, the results of this study indicate both an effect of the L1 on the L2 in offline resolution and a general L2 processing effect in online subject pronoun resolution.

Silent reading and reading aloud of German words and pseudowords were used in a PET study using ( 15 O)butanol to examine the neural correlates of reading and of the phonological conversion of legal letter strings, with or without meaning. The results of 11 healthy, right-handed volunteers in the age range of 25 to 30 years showed activation of the lingual gyri during silent reading in comparison with viewing a xation cross. Comparisons between the reading of words and pseudowords suggest the involvement of the middle temporal gyri in retrieving both the phonological and semantic code for words. The reading of pseudowords activates the left inferior frontal gyrus, including the ventral part of Broca's area, to a larger extent than the reading of words. This suggests that this area might be involved in the sublexical conversion of orthographic input strings into phonological output codes. (Pre)motor areas were found to be activated during both silent reading and reading aloud. On the basis of the obtained activation patterns, it is hypothesized that the articulation of high-frequency syllables requires the retrieval of their concomitant articulatory gestures from the SMA and that the articulation of lowfrequency syllables recruits the left medial premotor cortex.

Mental calculation is a complex cognitive operation that is composed of a set of distinct functional processes. Using functional magnetic resonance imaging (fMRI), we mapped brain activity in healthy subjects performing arithmetical tasks and control tasks evoking a comparable load on visuo-constructive, linguistic, attentional and mnemonic functions. During calculation, as well as non-mathematical tasks, similar cortical networks consisting of bilateral prefrontal, premotor and parietal regions were activated, suggesting that most of these cortical areas do not exclusively represent modules for calculation but support more general cognitive operations that are instrumental but not specific to mental arithmetic. Significant differences between calculation and the non-mathematical tasks were found in parietal sub-regions, where non-arithmetic number or letter substitution tasks preferentially activated the superior parietal lobules whereas calculation predominantly elicited activation of the left dorsal angular gyrus and the medial parietal cortices. We interpret the latter activations to reflect sub-processes of mental calculation that are related to the processing of numerical representations during exact calculation and to arithmetical fact retrieval. Finally, we found that more complex calculation tasks involving the application of calculation rules increased activity in left inferior frontal areas that are known to subserve linguistic and working memory functions. Taken together, these findings help to embed the specific cognitive operation of calculation into a neural framework that provides the required set of instrumental components. This result may further inform the cognitive modeling of calculation and adds to the understanding of neuropsychological deficit patterns in patients.

Spoken language is one of the most compact and structured ways to convey information. The linguistic ability to structure individual words into larger sentence units permits speakers to express a nearly unlimited range of meanings. This ability is rooted in speakers' knowledge of syntax and in the corresponding process of syntactic encoding. Syntactic encoding is highly automatized, operates largely outside of conscious awareness, and overlaps closely in time with several other processes of language production. With the use of positron emission tomography we investigated the cortical activations during spoken language production that are related to the syntactic encoding process. In the paradigm of restrictive scene description, utterances varying in complexity of syntactic encoding were elicited. Results provided evidence that the left Rolandic operculum, caudally adjacent to Broca's area, is involved in both sentence-level and local (phrase-level) syntactic encoding during speaking.