Speech and Brain

Braitenberg and Schuz (1998) stated that the average number of synaptic connections per neuron in the human cortex is approximately 10,000. This knowledge implies a huge potential for referential and recursive relationships between mental representations of elements like words and rules of syntax. Consider how the study of speech errors such as anticipation and perseveration are an important aspect of research on speech production.

What is the most significant characteristic these types of speech errors reveal about the speaker's processing of intent into spoken words? Why? Language is a characteristic of humans: every normal human being speaks; animals do not (Joyshree, 2014). That is, language is limited to humans only. While other animals interact with each other through a rigid collection of symbols, no group of animals possesses human language's combinatorial conventions, wherein symbols permute into an indefinite number of combinations, all of which have determinate meanings (Joyshree, 2014).

Sequencing faults in elicited and natural speech have been used for long, for informing phonological encoding and comprehending the process through which sequential ordering in speech is attained (Wook and Melissa, 2012). In Fromkin's1971 classic paper, the actuality of phonological components like syllables, features, and phonemes based on erroneous displacements or transpositions of these components in natural language is discussed.

For instance, Fromkin maintained that syllables denoted representational units, due to the fact that when phonemes, clusters or features, were shifted or reordered from a particular target, the syllable position of the target was generally preserved. From the time of Fromkin's paper in 1971, several researchers have employed identical sequencing errors, on not just understanding linguistic components' psychological reality, but also for comprehending serial ordering methods in speech and speech plan architecture (as cited in Wook and Melissa, 2012).

Psycholinguistic speech production models have integrated numerous insights from Shattuck-Hufnagel (1979, 1983 as cited in Wook and Melissa, 2012); however, the emphasis has transferred from speech plan structures and components to elements' storage and subsequent serial retrieval. The above models hypothesize that those linguistic elements' distributed depiction and gradient initiation in the course of speech planning may cause errors in speech.

Aside from giving an explanation for speech error creation and distribution in terms of psychological plausibility, network models underline speech production's dynamic nature: the models explain impacts of the past as well as future upon the present. With reference to dynamics, experts treat speech planning in the form of an incremental procedure, occurring simultaneously at multiple organizational levels. Despite the focus on mentally plausible factors, network models' elemental nodes are essentially syllabic and phonemic units, with more advanced nodes being characteristically semantically- and syntactically-tagged lexemes.

This architecture results in a provision of strong explanations of word-productions, with little understanding of connected words, in the models. Additionally, linguistic focus has shifted from word-level to phrase-level patterns of sound (Wook and Melissa, 2012). Au, Harper and Croot, in 2010 (as cited in Wook and Melissa, 2012), studied how speech errors are impacted by prosodic prominence and phrase position. It was discovered that, usually, errors transpired less frequently in well-known words and were particularly rare in well-known words occurring at the phrase-initial point.

This finding was also in line with the 2002 interpretation that more advanced prosodic structures are identified before encoding of word forms in phonetic and phonological detail (contra Levelt 1989, and Levelt et al. 1999; as cited in Wook and Melissa, 2012). Croot and coworkers, as well as Shattuck-Hufnagel and Keating, referred to a scan-copier prototype for discussing sequential ordering in multiple-word prosodic frameworks (as cited in Wook and Melissa, 2012). Both maintained that elements have to be tagged beforehand for prominence.

Tagging reduces the frequency of errors linked to prosodically prominent terms as few terms may be tagged so in any particular speech (Wook and Melissa, 2012). If we assume that positional influence on errors, reported by Croot and coworkers, is prosodic in nature, we may expect to come across the same influence at Intonational Unit (IU) limits in the utterance-medial location.

If components intended for the initial spot are tagged uniquely, while others aren't, the protection offered by tagging must not cover other locations; further, errors must be uniformly high over each non-initial spot in the unit. If, however, a positional impact on errors corresponds to triggering of an IU-sized phonological encoding domain, with errors minimized at high activation levels, we can hypothesize an asymmetric error distribution over the span of an intonational unit, consistent with initial high activation, followed by activation decay (Wook and Melissa, 2012).

References Croot K., Claudia A., Amy H. (2010).Prosodic structure and tongue twister errors. In Wook K.C. And Melissa A.R. (2012). The distribution of speech errors in.