I don’t know about you, but when I read books, I have a personal narrator who reads the book out to me in my head. The narrator’s voice varies, depending on the book’s characteristics. When it comes to reading school-related things, the voice reading out the text belongs to the teacher of that subject, since that teacher discussed those things out loud in class. It’s when you read a text message or chat with a friend and say “Oh my gosh I can really imagine/ hear you saying that!” You know what I mean more or less, right?
People attend to, learn, and respond to linguistic and nonlinguistic properties of speech when they encounter spoken language. In other words, we don’t just pay attention to the structure (grammar, words, sentences) and content a person says, but also how they say it (their voices and manner of speaking). We pay attention to ‘talker-specific properties’ that make that person’s speaking style unique. Listeners have been found to identify the speaker’s dialect, status, health, sex, and emotional state. Such sensitivity to nonlinguistic properties of speech drove researcher to examine how talker-specific information influences the representation and processing of spoken language. Research has suggested that such properties of speech are retained in memory and influence perceptual processing of spoken language in different levels (phonological, lexical, and sentential, if you were curious).
Such research goes against the ‘traditional’ approaches to linguistic representation which depend on linguistic (‘technical’) properties of speech. Such a view entails context-free representations that are independent of nonlinguistic properties such as the speaker’s voice. The approach against this traditional view proposes that representations of spoken language include nonlinguistic properties (surface characteristics). Such a view believes that these properties aren’t necessarily independent of linguistic content. Instead, they constitute an integral component of the speech and language perceptual process. This approach has called for a renewed focus on the importance of nonlinguistic info and their roles in processing and representing speech. Research has been increasingly examining the surface characteristics listeners attend to and retain in spoken language.
The researchers in this study examined whether readers would engage in a form of auditory imagery for speech both when reading aloud and when reading silently. The researchers got 99 participants of introductory psych classes, all of whom received credit for participating. (Sounds familiar?) It was ensured that the students had normal hearing. The students were then assigned to see one of two films: either two women talking for two minutes, or two men talking for two minutes. For each film, one of the speakers was slow, and the other was fast. All speakers were actors, and their speaking rates were really measured to ensure that one was fast and the other was slow (5.01 and 4.66 words per second for the females, 6.18 and 3.80 words per second for the females). The participants then answered a few questions about the conversation they watched to see if they were listening. The participants then had to read two texts, one “written” by one of the speakers, the other “written” by the other speaker. The texts themselves were manipulated based on level of difficulty, one easy and one hard. The participants were also randomly assigned to either read the second passage out loud or silently. They were told to read the passages once. Duration of reading was measured for the different set-ups. Three experiments were conducted to ensure more confidence in results.
Findings in this study demonstrate that readers can engage in auditory imagery for talkers’ voices. Across experiments, reading times, both silent and aloud, were significantly slower when readers were told that the passage was written by the speaker with a slow speaking rate and significantly faster when the passage was written by the fast speaker. Change of reading speed was based on a brief two-minute exposure to a conversation between talkers. However, the extent to which readers engaged in auditory imagery depended on text difficulty. Readers were more likely to access talker-specific representations during silent reading for difficult than for easy texts. Still, readers who scored high in auditory imagery for reading were more likely to access talker-specific representations for easy texts. These effects suggest that the linguistic representations accessed during reading preserve or include surface characteristics of spoken language. Auditory imagery appears to be talker-specific and to mirror the encoding of talker-specific information during speech perception and spoken word recognition.
I like to read, and what I liked about this study is that it took a look at how the way we read, whether out loud or silently, can be influenced by our perception of the author, specifically the author’s speaking voice. It’s interesting to know that not only is auditory imagery, that “inner voice”, something experienced by everyone, but also that it actually affects how fast or slow we read things. It’s also fascinating how well our memory works. After being exposed to a speaker’s voice for such a short time, we can quickly and easily incorporate it to that person’s writings, even if we’re not told to, and even if we’re not completely sure that that person was really the author of the text. Speaking rate, an aspect of a person’s speaking voice, actually influences our processing of written text. It’s so interesting how so many things come together: listening to sentences and understanding their meanings, reading that text whether silently or verbally and incorporating the author’s voice based on how we remember them speak, etc. It’s vision, hearing, memory, structure, content, and perception/ experience all together.
I give props to this study for different things they considered. They considered not just fast and slow speaking, but they even made sure gender did not get in the way of auditory imagery. They also considered manner of reading, which is either silently or out loud. They also conducted three experiments so that they could come up with general findings from the three. And they also made sure to lessen individual differences in hearing. What I didn’t like about this study though was that I felt like I needed more hard evidence. Although the researchers controlled and considered different variables, I don’t think timing reading is enough. Maybe it’s because I’m currently exposed to our 135 book where there’s always neuroimaging and other methods in supporting findings. The article actually mentioned research from neuroimaging studies of auditory imagery from before and how findings then can mean that imagery in this study can be re-experiencing sounds of speech or reenacting an articulatory event when the actual stimulus is not present (in other words, experiencing something similar to hearing or simulating the talker reading the text). Although support for their studies make sense, it would have been better if they made use of neuroimaging studies aside from measuring time instead of just connecting the two.
What studies can be done in the future? A bigger sample size to see how ‘true’ the results are, a comparison between males and females to see if there are differences with how they process written text, cross-cultural comparisons, and even a look into written text in the form of online chatting and text messaging. It would also be interesting to compare with levels of hearing. This study ensured that the individuals were comparable with hearing. But would people who cannot hear well be as affected by surface properties as much as others? How about people who can’t hear at all? Or in such case, would speed of a signer affect them just as speed of talkers affects normal hearing people?
I wonder why we can’t help but hear those narrators in our heads. Why is it so important for us to process things based on surface characteristics? I wonder if, somehow, in the future, technology will be that advanced for us to find out. The mere fact that people are researching on such a topic now already shows how advanced a society we are, considering there are so many other “pressing” studies to research on. But hey, we can take home something from this study. Now that we know that how fast we read can be affected by the “author” of the written text, maybe we can try (if it’s possible) to tell ourselves that the authors of our books and readings are fast speakers. That way, we can get a lot more readings done, instead of ideally planning out our time and then realizing an hour or two later that a) we read slower than we thought or hoped and b) why oh why were we so idealistic about the time and our reading skills?
Source:
Alexander, J., & Nygaard, L. (2008). Reading voices and hearing text: Talker-specific auditory imagery in reading. Journal of Experimental Psychology: Human Perception and Performance, 34(2), 446-459.
No comments:
Post a Comment