Top-down processing in speech perception

This is a paper I wrote for my introductory Psychology class.
I would like to note that I was one of three working on this paper. Other contributors: Ka Jacky Chan and Laura Robertson.

Top-down processing is the process by which our own beliefs and/or expectations guide our perception of a stimulus. Our group has three journal articles that provide strong evidence that top-down processing contributes abundantly to speech perception.

            Cole, Jakimik, and Cooper (1980) did a study consisting of four experiments. The first three were very similar, and demonstrated that prior context affects the perception of word boundaries (Cole et al., 1980).  Participants were told one of two stories that were the context for the line “looking down from the pier they saw the carko on the ferry” (Cole et al., 1980). The first story was about a man sent to the docks to pick up a shipment of goods, and the second story was about two parents that were watching their son drive a jeep onto a ferry. The context of the story affected the perception of the mispronounced word, ‘carko’ (Cole et al., 1980). The former story caused the person to perceive the word ‘cargo’, whereas the latter story caused the person to perceive the words ‘car go’ (Cole et al., 1980). The speed of perception also differed between the two stories; when the context of the story caused the person to expect a single word, such as in the former story, perception was faster (Cole et al., 1980). The latter story consists of the perception of two words, and in this case the person must continue to listen to the sentence in order to put the mispronounced word into context (Cole et al., 1980). This evidence suggests that when we put into context the speech that we are interpreting, our brain makes assumptions of what it will encounter next; in essence this is top-down processing at work (Cole et al., 1980).
            Borsky, Tuller, and Shapiro (1998) demonstrated that prior context plays a role in the categorization of words. In this study lists of words were combined to form sentences that were the context for a target word (Borsky et al., 1998). The target word consisted of two “boundaries”, the word ‘goat’ and the word ‘coat’, as well as many manipulations in between (Borsky et al., 1998). A sentence would be read to the participant, and one of the goat/coat manipulations would be inserted as the target word. The word goat or coat was then shown to the participant on a computer screen, and they were asked to press a ‘yes’ button if it matched the word they heard (congruent trial) or a ‘no’ button if it was different than the word they heard (incongruent trial) (Borsky et al., 1998). During congruent trials, words that were the least manipulated (closest to ‘goat’ or ‘coat’ boundary) had the fastest reaction time and most accurate perception (Borsky et al., 1998). Generally, the more manipulated the target word was the longer the reaction time (Borsky et al., 1998). Our group feels this evidence suggests top-down processing keeps stimuli in context. If a stimulus occurs that does not fit the context of previous stimuli, top-down processing will work to alter the way we perceive the out-of-context stimulus.
            Innovations in neuroimaging technology have allowed us to more fully understand cognitive processes. In the study conducted by Zekveld, Helsenfeld, Festen and Schoonhoven (2006), functional magnetic resonance imaging (fMRI) was used to measure the brain activity involved in speech comprehension. While participants were presented different speech stimuli consisting of various degrees of intelligibility (a measure of how well the speech can be understood), fMRI images were captured to study the corresponding active areas of the brain (Zekveld et al., 2006). The study found that when participants are presented with unintelligible speech stimuli, one of the two sections of Broca’s area known as Brodmann area 44 (BA44) becomes highly activated (Zekveld et al., 2006). This finding suggests that Broca’s area, especially BA44, is the area of the brain that is most likely responsible for top-down processing in speech comprehension (Zekveld et al., 2006), and there are copious amounts of other research demonstrating top-down processing in action.

Works Cited
Ronald A. Cole, Jola Jakimik, and William E. Cooper (1980). Segmenting speech into words. The Journal of the Acoustical Society of America 67, (4), 1323-1332.

Borsky, S., Tuller, B., & Shaprio, L.P. (1998). ‘‘How to milk a coat:” The effects of semantic and acoustic information on phoneme categorization. The Journal of the Acoustical Society of America 103, (5), 2670-2676.

Zekveld, A.A., Heslenfeld, D.J., Festen, J.M., & Schoonhoven, R. (2006). Top-down and
bottom-up processes in speech comprehension. NeuroImage, 32, 1826 – 1836.