Language and other processing systems
المؤلف:
Paul Warren
المصدر:
Introducing Psycholinguistics
الجزء والصفحة:
P228
2025-11-13
49
Language and other processing systems
We have considered above the issue of modularity within language processing, i.e. whether there are separate modules for different aspects or levels of language production or comprehension. We have also looked at the number and nature of the modules that are utilised in the production and comprehension of spoken and written words, as well as the connections between them. In this section we consider another aspect of the modularity debate, namely the extent to which there is a separate module for language processing, contrasted with other cognitive skills or systems.
On a non-modular view, language, memory, reasoning, etc. can all be seen as examples of a general and powerful capacity for forming concepts and manipulating information. On a modular view, however, it is argued that there are numerous very different kinds of mental abilities, each operating in its own way, and language is one of several such modules.
Support for a modular view includes arguments along the lines of speech is special’. That is, there are differences e.g. between language processing and other sensory processing, so that for example speech processing and music processing involve different brain hemispherical specialisations though with some overlap, e.g. for speech prosody and music. In addition, the processing of the visual input for sign language involves different neurological and psychological processes from those involved in processing spatial relations though again there is some overlap. Further support for the modular view would result if there was clear evidence of separation of processes involved in language production or comprehension from central cognitive processes (i.e. those involved in memory, reasoning, etc.). The strongest sort of dissociations of this type would be double dissociations, in this case evidence not only for language deficits with preserved cognitive abilities but also for cognitive deficits with preserved language abilities.
Specific language impairment SLI has often been cited as a case of dissociation of linguistic and other abilities (Bishop Snowling, 2004; Cohen, 2002; Leonard, 1998). In SLI there is early and often lasting impairment in language capabilities. This is coupled with normal performance
on nonverbal tests of general intelligence. There is generally also no evidence of speech impediment or hearing loss that might explain problems at a more peripheral level. SLI speech tends to be slow and deliberate, and typically exhibits problems with grammatical forms. For instance, SLI speech does not consistently use appropriate grammatical endings, with SLI sufferers producing sentences such as the girl is jump over the rope or I can see three cat. Because there is some evidence that SLI runs in families and so may have a genetic component, it has been suggested that it may involve hereditary problems with some of the neural wiring responsible for language functions.
A contrasting set of symptoms is exhibited by so-called chatterbox syn drome also known as Williams syndrome. Sufferers typically speak fluently and grammatically, and also score well on specifically linguistic tasks such as grammaticality judgement. In contrast to this evidence for intact linguistic skills, they show poor performance in tests of logical reasoning and in questions concerning general knowledge (Pinker, 1994).
SLI and chatterbox syndrome appear on the surface to give us a double dissociation of linguistic and other cognitive skills, which would support the existence of a separate language module from general intelligence, etc. That is, SLI has impairment of language skills with preservation of general cognitive function, while chatterbox syndrome has preservation of language skills but poor general cognitive function. However, it has been argued that SLI may be an unwitting result of difficulties in other areas which are in turn not specifically linguistic in nature. For example, SLI sufferers show an impaired ability to track where a sound is coming from. This is revealed in tasks where they listen to a sound such as a tone or click and have to point to where they think the sound is coming from. One of the early signs of children’s perceptual development for language is their ability to locate the source of speech sounds. This is important for phonological development, since attention can be better directed to the input if its location can be tracked. The fact that SLI sufferers are less well able to track a moving signal may have led to impaired phonological development. In turn this may have had a detrimental effect on the course of other linguistic development. In the case of chatterbox syndrome, there are many sufferers who fail to learn to read or write, i.e. there are at least some linguistic skills admittedly secondary skills, in terms of the development both of the species and of the individual that are not preserved.
Visual world paradigms
More direct evidence against modular separation of linguistic and other cognitive skills includes research results that show how general knowledge is used with great immediacy and efficiency in constraining language processing. Consider for instance the bridging inferences discussed in Chapter 12 with reference to examples such as (13.10), where it is argued that the ease with which we understand the two sentences together depends on our ability to use our understanding of the world and of the fact that car is implicit in the role of vehichle used to drive to London first sentence.
Other evidence comes from studies of how we use non-linguistic visual information from the environment during language processing, using what is known as a visual world paradigm. Typically, this involves presenting a scene the visual world’ to participants on a computer screen and monitoring their eye movements as they listen to some pre-recorded instructions. The point of interest is usually where in the visual world they direct their gaze as they hear and interpret the information in the instruction. So, for instance Eberhard et al. (1995) used different arrangements of objects and recorded participants’ eye movements as they listened to spoken commands to move one of the objects. The sentences used in the command utterances investigated the parsing strategy of Minimal Attachment see Chapter 10. This strategy predicts that the initially preferred interpretation of the instruction in (13.11) at the point that put the apple on the towel has been heard will be that the apple should be moved onto the towel. This is because attachment of the prepositional phrase (pp) on the towel as an argument of the verb i.e. telling the participant the goal of the action is syntactically simpler than attachment of the same PP as a modifier of the noun phrase the apple (where it would be telling the participant something about where the apple already is).
The sentence in (13.11) can be contrasted with the sentence in (13.12), where the phrase a s makes it unambiguously clear that the PP modifies the apple.
Eberhard e al used the pair of sentences (13.11) and (13.12), along with others with different object names. Each sentence in a pair was listened to in combination with two different configurations of four objects. In both configurations three of the quadrants of a worktable contained an apple on a towel, a towel with nothing on it, and a box. In one configuration – the single referent configuration in Figure (13.4) – the fourth quad rant had an object not related to any of the others, e.g. a pencil. In the other configuration – the double referent configuration in Figure (13.5) – the fourth object was an apple that was not on a towel. In this case, the phrase apple on the towel with a non -Minimal Attachment interpretation (where the PP on the towel modifies the noun apple rather than the verb put) is made contextually relevant by the presence of the other apple, since the phrase distinguishes the two apples one is on a towel, the other is not. In the single referent configuration, participants’ eye gaze as they listened to the ambiguous instruction in (13.11) moved first to the apple (on hearing apple), then to the uncovered towel (on hearing towel), and then at the end of the sentence back to the apple and then to the box. This pattern of eye movements, including a fixation on the uncovered towel, indicates that participants were initially garden-pathed into considering an action of moving the apple onto the uncovered towel. This contrasted with eye movements when hearing the unambiguous sentence in (13.12), again with the single referent condition in Figure (13.4), where participants’ gaze moved only to the apple and then to the box. The phrase that’s on in (13.12) made the interpretation clear, much as we saw for examples of explicit marking of syntax in Chapter 10.
In the double referent condition, as participants heard the apple on the towel in (13.11), they typically looked at both apples, but rarely at the uncovered towel. They were no longer misled by the Minimal Attachment preference into considering the towel as a goal. This indicates that the visual information in this double referent condition gives relevance to the use of the PP as a modifier of the noun phrase, and eliminates the garden path.
This example, along with other studies that have measured eye movements during spoken language processing, shows that the linkage between non-linguistic and linguistic information is an important factor in the interpretation of utterances, and can even overrule strong syntactic preferences shown in a range of other studies.
الاكثر قراءة في Linguistics fields
اخر الاخبار
اخبار العتبة العباسية المقدسة
