Editor theories
المؤلف:
Paul Warren
المصدر:
Introducing Psycholinguistics
الجزء والصفحة:
P79
2025-11-03
43
Editor theories
The finding that speakers monitor their own speech and correct it as they go shows that an editing process takes place during speech production. The precise nature of the editor has been an issue of some debate. Is it, for instance, a mechanism that is specifically geared to self-monitoring and to the editing of the speaker’s own output This would suggest that self-monitoring is a special part of the speech production process, perhaps involving the speaker in a comparison of what they said with what they intended. On the other hand, is self-monitoring simply a particular implementation of a more general monitoring device that is also used when we are listening to other speakers in this case, self-monitoring may operate without continual reference to what the speaker intended, but might use more general criteria, such as does the output I have just heard make sense’?
Neuroimaging studies (Indefrey, 2007) indicate that we actually use similar brain areas in two types of self-monitoring, internal and external, as well as in listening to others. The speed with which errors are detected and corrected also supports the existence of internal as well as external monitoring. The presence of covert repairs reinforces this suggestion, since in covert repairs the errorful word is not even pronounced. Further evidence comes from experimental studies of speech errors using the SLIP technique (see p. 73). As well as spoonerism outcomes, the SLIP technique often results in partial spoonerisms, such as shide shame for hide shame, as well interrupted and corrected spoonerisms such as sh- hide shame for the same target. Most of the repaired spoonerisms in these tasks are very rapid, with an interruption very early in the error, arguably much earlier than would be compatible with monitoring of and responding to the actual spoken output.
Is this interruption and repair in response to lexical or phonetic monitoring of the speaker’s own production That is, is the repair a response to the realisation that the output is not the word the speaker intended or indeed not a word at all, or does it follow from a check of the actual sounds produced against the target sounds One piece of evidence that supports a lexical argument is that there is a greater likelihood that an experimentally induced error will be interrupted if it would have result ed in a nonword. This has been taken as evidence that there is a quick and dirty’ general check that the output consists of real words rather than nonsense. This is compatible too with the general tendency noted in Chapter 4 for spontaneous sound errors to result in real words more often than they result in nonsense words, and with the finding noted earlier in the current chapter that induced spoonerisms are more likely to be completed if they result in real words (short hurt) than if they result in a nonsense sequence (sheel helf). The nonsense word sequences are simply edited out early and efficiently by a general monitor for real word status, but the real word sequences pass this general check and are allowed to proceed.
On the other hand, in cases where an error correction is made, then it is equally fast regardless of whether the error would have produced real words or nonwords. This suggests that from an early stage the monitoring system also makes a phonetic comparison of the speaker’s articulation plan with the intended output. So internal monitoring has access to details of the phonetic plan, which it assesses without reference to the lexical status of the output.
الاكثر قراءة في Linguistics fields
اخر الاخبار
اخبار العتبة العباسية المقدسة
