First, words are construed as pairs of phonographic forms with highly schematic concepts which are internally organized as radial categories and function as access sites to encyclopedic knowledge.
Universal Principles in the Repair of Communication Problems
Second, an account of the representational organization of encyclopedic knowledge is provided. According to Fillmore, encyclopedic knowledge is represented in long-term memory in the form of frames , i. Frames provide thus a schematic representation of the elements and entities associated with a particular domain of experience and convey the information required to use and interpret the words employed to talk about it.
The meaning of this word cannot be grasped independently of a prior understanding of the notion of a circle. Langacker argues that domains are typically structured into hierarchies that reflect meronymic relations and provide a basic conceptual ontology for language use. Importantly, individual profiles typically inhere to different domains, and this is one of the factors responsible for the ubiquity of polysemy in natural language. The notion of a frame has become popular in cognitive psychology to model the dynamics of ad hoc categorization e.
Research on the mental lexicon is concerned with a variety of problems for surveys, see, e. From a functional point of view, the mental lexicon is usually understood as a system of lexical entries , each containing the information related to a word mastered by a speaker Rapp A lexical entry for a word w is typically modeled as a complex representation made up of the following components Levelt , :. From this standpoint, a theory of word meaning translates into an account of the information stored in the semantic form of lexical entries. A crucial part of the task consists in determining exactly what kind of information is stored in lexical semantic forms as opposed to, e.
Recall the example we made in Section 3.
Not surprisingly, even in psycholinguistics tracing a neat functional separation between word processing and general-purpose cognition has proven a problematic task. The general consensus among psycholinguists seems to be that lexical representations and conceptual representations are richly interfaced, though functionally distinct e. On the resulting view, lexical activity in humans is the output of the interaction between two functionally neighboring systems, one broadly in charge of the storage and processing of conceptual-encyclopedic knowledge, the other coinciding with the mental lexicon.
The role of lexical entries is essentially to make these two systems communicate with one another through semantic forms see Denes Contrary to the folk notion of a mental lexicon where words are associated to fully specified meanings or senses which are simply retrieved from the lexicon for the purpose of language processing, in these models lexical semantic forms are seen as highly schematic representations whose primary function is to supervise the recruitment of the extra-linguistic information required to interpret word occurrences in language use.
In addition to these approaches, in a number of prominent psychological accounts emerged over the last two decades, the study of word meaning is essentially considered a chapter of theories of the mental realization of concepts see the entry on concepts. Lexical units are seen either as ingredients of conceptual networks or as auditory or visual stimuli providing access to conceptual networks. A flow of neuroscientific results has shown that understanding of certain categories of words correlates with neural activations corresponding to the semantic content of the processed words.
For example, it has been shown that listening to sentences that describe actions performed with the mouth, hand, or leg activates the visuomotor circuits which subserve execution and observation of such actions Tettamanti et al. Such perceptual symbols are not holistic copies of experiences but selections of information isolated by attention.
Related perceptual symbols are integrated into a simulator that produces limitless simulations of a perceptual component, such as red or lift. Simulators are located in long-term memory and play the roles traditionally attributed to concepts: they generate inferences and can be combined recursively to implement productivity. Linguistic symbols i. Though popular among researchers interested in the conceptual underpinnings of semantic competence, the simulationist paradigm faces important challenges.
Three are worth mentioning. First, it appears that imulations do not always capture the intuitive truth conditions of sentences: listeners may enact the same simulation upon exposure to sentences that have different truth conditions e. Moreover, simulations may overconstrain truth conditions. Second, the framework does not sit well with pathological data. For example, no general impairment with auditory-related words is reported in patients with lesions in the auditory association cortex e. Finally, the theory has difficulties accounting for the meaning of abstract words e.
Beginning in the mids, neuropsychological research on cognitive deficits related to brain lesions has produced a considerable amount of findings related to the neural correlates of lexical semantic information and processing. More recently, the development of neuroimaging techniques such as PET, fMRI and ERP has provided further means to adjudicate hypotheses about lexical semantic processes in the brain Vigneau et al. Here we do not intend to provide a complete overview of such results for a survey, see Faust We shall just mention three topics of neurolinguistic research that appear to bear on issues in the study of word meaning: the partition of the lexicon into categories, the representation of common nouns vs.
Two preliminary considerations should be kept in mind. First, a distinction must be drawn between the neural realization of word forms, i. A patient can understand what is the object represented by a picture shown to her and give evidence of her understanding, e. Second, there appears to be wide consensus about the irrelevance to brain processing of any distinction between strictly semantic and factual or encyclopedic information e.
Let us start with the partition of the semantic lexicon into categories. Neuropsychological research indicates that the ability to name objects or to answer simple questions involving such nouns can be selectively lost or preserved: subjects can perform much better in naming living entities than in naming artifacts, or in naming animate living entities than in naming fruits and vegetables Shallice Different patterns of brain activation may correspond to such dissociations between performances: e.
However, the details of this partition have been interpreted in different ways.
Finally, Devlin et al. This is what typically happens in semantic dementia. Empirical evidence has motivated theories of the neural realization of conceptual competence that are meant to account for both modality-specific deficits and pathologies that involve impairment across all modalities. The former may involve a difficulty or impossibility to categorize a visually exhibited object which, however, can be correctly categorized in other modalities e.
While damage to the spokes accounts for modality-specific deficits, damage to the hub and its connections explains the overall impairment of semantic competence. On this model, the hub is supposed to be located in the anterior temporal lobe ATL , since semantic dementia had been found to be associated with degeneration of the anterior ventral and polar regions of both temporal poles Guo et al.
It should be noted that while many researchers converge on a distributed view of semantic representation and on the role of domain-specific parts of the neural network depending on differential patterns of functional connectivity , not everybody agrees on the need to postulate a transmodal hub see, e. Let us now turn to common nouns and proper names.
- R200 PCB.
- Center for Computational Thinking, Carnegie Mellon.
- By the Light of the Moon.
As we have seen, in the philosophy of language of the last decades, proper names of people, landmarks, countries, etc. Neuroscientific research on the processing of proper names and common nouns concurs, to some extent. To begin with, the retrieval of proper names is doubly dissociated from the retrieval of common nouns. Other patients had the complementary deficit. Such findings suggest distinct neural pathways for the retrieval of proper names and common nouns Semenza The study of lesions and neuroimaging research both initially converged in identifying the left temporal pole as playing a crucial role in the retrieval of proper names, from both visual stimuli Damasio et al.
In addition, recent research has found a role for the uncinate fasciculus UF. In patients undergoing surgical removal of UF, retrieval of common nouns was recovered while retrieval of proper names remained impaired Papagno et al. If these results were confirmed, it would turn out that the linguistic category of proper names is not homogeneous in terms of neural processing. Studies have also demonstrated that the retrieval of proper names from memory is typically a more difficult cognitive task than the retrieval of common nouns.
Interestingly, the same difference does not materialize in definition naming, i. Finally, a few words on the distinction between the inferential and the referential component of lexical competence. As we have seen in Section 3.
Beginning with Warrington , many patients had been described that were more or less severely impaired in referential tasks such as naming from vision and other perceptual modalities as well , while their inferential competence was more or less intact. The complementary pattern i. Still, a number of cases have been reported, beginning with a stroke patient of Heilman et al.
In subsequent years, further cases were described. For example, in a study of 61 patients with lesions affecting linguistic abilities, Kemmerer et al. Thus, while double dissociation between inferential processes and naming from vision is well attested, it is not equally clear that it involves referential processes in general. On the other hand, evidence from neuroimaging is, so far, limited and overall inconclusive. Some neuroimaging studies e.
However, the results are not entirely consistent and are liable to different interpretations. For example, the selective activation of the anterior left temporal lobe in inferential performances may well reflect additional syntactic demands involved in definition naming, rather than be due to inferential processing as such see Calzavarini for a discussion.
Basics 1. Historical Background 2. Philosophy of Language 3. Linguistics 4. Cognitive Science 5.
Basics The notions of word and word meaning are problematic to pin down, and this is reflected in the difficulties one encounters in defining the basic terminology of lexical semantics. Historical Background The study of word meaning became a mature academic enterprise in the 19 th century, with the birth of historical-philological semantics Section 2.
Philosophy of Language In this section we shall review some semantic and metasemantic theories in analytic philosophy that bear on how lexical meaning should be conceived and described. For example, consider the following exchange: A: Will Kim be hungry at 11am? B: Kim had breakfast. Linguistics The emergence of modern linguistic theories of word meaning is usually placed at the transition from historical-philological semantics Section 2.
Just as the rules of chess can be stated and analyzed without making reference to the mental properties of chess players, so a theory of word meaning can, and should, proceed simply by examining the formal role played by words within the system of the language. Since the primary explanandum of structuralist semantics is the role played by lexical expressions within structured linguistic systems, structuralist semantics privileges the synchronic description of word meaning.
Diachronic accounts of word meaning are logically posterior to the analysis of the relational properties statically exemplified by words at different stages of the evolution of the language.