Localization of Word Comprehension and Production

As we have seen, patient studies can be used to posit the nature of the functional components of language comprehension and production and their connections. Studies of localization reveal where in the brain those components may lie. Models of brain areas and their functions can be traced back to the late 1800s, beginning with Lichtheim and Wernicke. Wernicke's model assumed that language was represented in the left hemisphere. Findings since that time have indicated that although the left hemisphere is dominant for language in most individuals, the right hemisphere also plays some role. In addition, the right hemisphere is dominant in some individuals. The model incorporated a concept center that received input from sensory word images and provided output to motor word images. The sensory word images were thought to be represented in Wernicke's area, found in the superior (upper), posterior (near the back), left temporal lobe, and the motor images in Broca's area, found in the left frontal lobe near the motor cortex. Damage to Wernicke's area was thought to result in an inability to recognize and understand spoken words, although speech was thought to be fluent. Broca's aphasics were thought to have an impairment in articulating speech (i.e., "essentially mute, except for the repetition of the same few utterances''), but their comprehension was intact. Wernicke's and Broca's areas, though separate, were thought to be connected through a neural fiber tract (i.e., the arcuate fasciculus). This fiber tract was thought to be involved in translating an auditory word representation into a motor word representation. Based on Wernicke's model, one would predict that patients who had intact Broca's and Wernicke's areas but damage to this fiber tract should have good language comprehension and production but have difficulty repeating words and sentences. Patients were reported who showed this pattern (termed conduction aphasics), which seemed to provide a strong confirmation of the model.

There are a number of observations that cause difficulties for the classical model. For example, Broca's aphasics are not equally impaired on all types of words, typically being better able to produce nouns than verbs. Although they have difficulty producing function words, function words occurring in the middle of sentences are omitted less frequently than those that occur in the beginning of sentences. An additional complication for the Wernicke/Lichtheim models is that although Wernicke's aphasics' speech is fluent it shows numerous aberrations, such as mis-ordered phonemes, incorrect words, and nonsense words such as "tarripoi" in the statement spoken by one Wernicke's aphasic, "I can't mention the tarri-poi.'' Moreover, current studies suggest that among individuals classified into any traditional syndrome category (Broca's, Wernicke's, and conduction aphasia), there are wide variations in the nature of their deficits. For example, within conduction aphasics, some patients produce fluent and appropriate speech output but make semantic substitutions or paraphrases in repeating sentences, suggesting that they have difficulty retaining phonological information on the input side. Other conduction aphasics make numerous phonemic errors in their speech output but do not make semantic substitutions in repeating, suggesting that they have difficulty constructing and maintaining phonological representations on the output side. Thus, localization of function on the basis of classical syndromes appears to be a misguided effort. However, careful examination of a series of individual cases who show similar behavioral deficits allows for the determination of the lesion site that is affected in all such individuals. On the basis of these data, one can look to behavior-lesion correspondences to draw conclusions about localization.

Recently, using methods that measure the physiological changes that occur in the brain, it has been possible to determine the brain areas activated in normal individuals while they are performing language tasks. Electrical activity occurring in the brain during various tasks [i.e., event-related potentials (ERPs)], can be measured. These ERPs show negative and positive electrical potentials that are time locked to the onset of the presentation of verbal stimuli. The brain distribution of these potentials can be plotted. The flow of blood into different regions of the brain can be detected through via positron emission tomography (PET) and functional magnetic resonance imaging (fMRI). The use of PET and fMRI for functional localization thus assumes that greater blood flow is observed in brain areas with greater neural activity. Although all these methodologies provide information about when and where activation occurs in the brain, ERPs are known for their good temporal resolution, whereas the strength of fMRI lies in its spatial resolution. The spatial resolution of PET is also substantially better than that from ERPs but worse than that offMRI.

The traditional model of language processing predicts that processing of phonological information of heard words should occur in the posterior left temporal lobe (Wernicke's area) and phonological activation for spoken words should occur in left frontal lobe (Broca's area). Although the brain areas implicated in the traditional model have been confirmed by imaging studies, additional areas have been found to be active. That is, although neuroimaging studies show consistent evidence of Wernicke's area activation during phonological encoding, temporo-parietal and frontal activation have been found as well. Similarly, studies have found activation both in Broca's area and in the lower portion of the posterior temporal lobe (Wernicke's area is the upper portion of the posterior temporal lobe) during language production. One caveat regarding the role of Broca's area is that it has been activated when research participants are asked to make hand and tongue movements, suggesting that activation of this area reflects general motor programming rather than motor programming specific to phonological output.

If visually presented words are converted into their phonological representations, then activation should occur in areas similar to those implicated for auditory word processing. Studies have found activation during reading of visually presented words in left tempor-oparietal cortex, which is also active during auditory word processing. The left posterior middle temporal gyrus, in addition to the occipital-temporal and left inferior temporal/fusiform regions, has also been shown to be active during word reading.

A variety of areas have also been implicated for semantic processing. Two tasks meant to measure semantic processing, categorizing words and generating actions that can be performed with objects, have shown activation in left frontal regions of Broca's and surrounding areas. Other research has suggested that the frontal cortex is involved in retrieving and maintaining semantic information rather than being the locus of representation of semantic knowledge. Recent studies with normal and brain-damaged research participants have suggested the role of temporal areas in semantic processing, including left posterior temporoparietal, inferior temporal, and anterior temporal cortex. Researchers investigating the localization of categorical knowledge with neuroanatomical and neuroimaging studies found a wide variety of brain areas to be involved during tasks that test knowledge of different categories. Even within a category, various brain areas have been activated. For example, the left temporal lobe, the right temporal lobe, and even the frontal and inferior parietal areas have all been reported to be active during processing of animals and living things. Areas responsible for processing nonliving things also appear to be diverse. Despite the varied areas that have been reported to be involved in semantic function, one area that has shown activation across several different studies is the temporal lobe. Work with Alzheimer's patients who have semantic impairment suggests a further category distinction. Alzheimer's patients whose knowledge of the perceptual features of objects is disrupted tend to have temporal region damage, and those whose knowledge of how objects function tend to have frontoparietal impairment.

More consistent findings have been obtained regarding different localization for words of different grammatical classes. Neuropsychological and neuroi-maging results suggest that nouns are represented in the temporal lobe and verbs in the frontal lobe. Some ERP studies have indicated that content words and function words elicit different timing of brain potentials and the involvement of different brain areas. For instance, one study showed early frontal activation from both word types with a similar pattern brain distribution at approximately 200-300 msec after the onset of the word. Approximately 150 msec later, however, differences for the word types were observed, with a different timing of the potential waveforms for the two word types and greater left hemisphere activation for the function words than for the content words. However, because this study measured processing of these word types in a sentence context, the different patterns may be due to the different roles that these words play in sentence processing rather than to different localization of lexical representations. For example, function words may lead to more predictions about the syntactic structure of upcoming words in a sentence.

Recent behavioral and neuroimaging data suggest that the right hemisphere may play more of a role in language processing than was previously assumed. Behavioral studies on this issue have used a priming paradigm in which word recognition (as measured by time to pronounce a word or to make a word-nonword decision) is facilitated by the prior presentation of a related word. To study hemispheric contributions, written words are presented to the left or right visual field, which engages processing in the contralateral cerebral hemisphere first. Priming studies have shown that priming occurs for only the most highly related words in the left hemisphere but for a broader range of related words in the right hemisphere. Also, for ambiguous words such as "bank" that have a dominant meaning (money) and a subordinate meaning (river), priming results indicate that immediately following presentation of an ambiguous word both meanings are activated in both hemispheres. However, within a short time, only the dominant meaning is available in the left hemisphere, whereas both meanings are still available in the right hemisphere. Taken together, the evidence suggests that the meanings of words are more coarsely coded in the right hemisphere. Neuroimaging studies of word processing have also often found activation in the right hemisphere in areas corresponding to the traditional language areas on the left, although the activation is often less than that obtained in the left hemisphere.

Unraveling Alzheimers Disease

Unraveling Alzheimers Disease

I leave absolutely nothing out! Everything that I learned about Alzheimer’s I share with you. This is the most comprehensive report on Alzheimer’s you will ever read. No stone is left unturned in this comprehensive report.

Get My Free Ebook


Post a comment