The relationship of production and comprehension
المؤلف:
Paul Warren
المصدر:
Introducing Psycholinguistics
الجزء والصفحة:
P223
2025-11-12
69
The relationship of production and comprehension
Issues in language production and in language comprehension have largely been presented separately in the earlier and later chapters of this book respectively. At points along the way, however, we have seen that production and comprehension share some common properties, and that there is alignment of the needs of production with those of comprehension. This is not at all surprising, given that most language users are both producers and comprehenders. As an example, we saw in Chapter 2 that during fluent speech production speakers tend to line up their pauses with the boundaries between phrases and clauses. As a consequence, these pauses assist the listener in structuring the input into phrases and clauses. In Chapter 6 we saw that speakers frequently accompany their speech with iconic gestures that represent some salient aspect of what is being talked about, or with batonic gestures that emphasise parts of their message. Again, while these can be considered as features of speech pro duction and as such may facilitate aspects of the production process, they can clearly also serve processes of comprehension.
A more general question, though, concerns the extent to which the same representations and processes are used in these two aspects of language processing. Although the question raises issues in a number of areas, we will focus here on the study of the mental lexicon, and consider the level at which the representations of words as used in production and comprehension start to differ. Two relatively simplistic views are illustrated in Figure (13.1). Note that the figure includes a phonological lexicon or lexicons, i.e. it deals with spoken rather than written language comprehension and production. We will return to discussion of issues to do with written input and output in later sections.
On the left of Figure (13.1) we see that the semantic lexicon the set of lemmas is linked to a single set of forms in a phonological lexicon. This set of forms, which might take the shape of phoneme strings, is used to drive the articulatory system and is at the same time the set of representations onto which the auditory analysis of the input is mapped. On the right of the figure we have another view, with separate form representations for output and input. This second architecture might incidentally allow different kinds of sub-lexical representations for production and for perception/comprehension, such as syllable structures in word production and phoneme strings in word recognition.
Researchers have looked to both experimental and clinical sources for evidence concerning whether production and comprehension share a common lexicon or have separate systems, at least in terms of the form based aspects of words. The existence of separate input and output path ways is supported by the finding that when participants have to listen for a target word in a list of spoken words it causes little interference with their ability to read words aloud at the same time (Shallice, McLeod & Lewis, 1985). This is supported also by research into the neurophysiological aspects of language processing – brain imaging studies have shown that different brain areas are activated when listening to words from those activated during speaking, either when reading aloud or when repeating words. On the other hand, there does appear to be some overlap between the production and comprehension of spoken words (Indefrey & Levelt, 2004). For instance, the production of a word in picture naming is facilitated if a phonologically related word is heard before the naming response is initiated (Levelt et al., 1999).
Clinical evidence supporting a separation of the pathways for word production and recognition typically includes some sort of dissociation between aspects of production and recognition or comprehension. For example, patients with anomia find it difficult to give the names for objects, indicating that they have difficulty in finding words for production. et they can nevertheless usually recognise and understand the same words in others’ speech. They are also able to speak fluently apart from their difficulty in producing object names, and so they clearly do not have a basic problem with control of their speech articulators or with generating a plan for articulation based on word knowledge. This suggests a model along the lines of the right-hand part of Figure (13.1), with separate input and output lexicons. The anomic patients have damage to the output route from the semantic lexicon through to the phonological output lexicon.
Additionally, there are patients who show word deafness. In this condition, patients can read and write and can speak quite normally, but are unable to understand words spoken to them, although their hearing sys team is otherwise unimpaired. They are also unable to repeat back words spoken to them. This combination suggests that the input route through the phonological input lexicon is damaged, so that the patients cannot use auditory input to access words for comprehension or for triggering the production of words during repetition.
Further patterns of problems that patients have with repetition and other tasks require us to modify the illustration on the right-hand side of Figure (13.1), along the lines of Figure (13.2). Imagine that a participant is asked to repeat back a word that they hear. The illustration in Figure (13.2) shows three routes for doing this. The sublexical route amounts to an echoing back or shadowing of what the participant hears. This route is required for repeating nonwords, which by definition do not have a lexical representation and so would not require access to the lexical system. The lexical route is used for repeating words, but without comprehension of those words. The semantic route indicates repetition that includes understanding. Damage to one or more of these routes will lead to different symptoms. Support for the existence of the three routes would arise if the right combinations of symptoms are found across different patients. As the next paragraph shows, this appears to be the case.
First, there are patients who are particularly poor at repeating non words, but who can understand and repeat real words. The fact that they understand these words shows that the input route through to the semantic lexicon is intact. The poor performance at repeating nonwords suggests impairment of the sublexical route. That is, these patients have to access words from their mental lexicon in order to find a pronunciation for them, which does not help them with nonwords. By contrast, there are other patients who can repeat both real words and nonwords but have no understanding of the words they can repeat. These patients would appear to have an impaired semantic route. Note though that neither of these two types of patient, nor the combination of what they are capable of achieving, sup ports a separate lexical route. All we have with these patients is support for a separation of semantic and sublexical routes. Crucially though for the model shown in Figure (13.2), there are also patients who cannot repeat nonwords and can repeat words, but with little understanding of the words they repeat. Such patients have an intact lexical route, but are unable to make good use of the other two routes. (For further discussion of these cases, see Harley 2008, Chapter 15).
These examples of patients with different types of impairment have not been discussed in detail here, since that would be beyond the scope of this book. They have been summarised to illustrate the complexities of the relationships between different types of processing, in this case of linguistic elements at one level, i.e. the lexical level. But they also show how considering a combination of patients with different capabilities can reveal more about the probable architecture of an intact language system than would be possible by looking at one patient – or unimpaired language user – alone.
الاكثر قراءة في Linguistics fields
اخر الاخبار
اخبار العتبة العباسية المقدسة
