noc19-hs14 Lecture 27 - Theories of Reading

Summary

This video discusses two prominent models of eye movement in reading: the E-Z Reader model and the SWIFT model. The E-Z Reader follows a sequential process of reading, where attention is focused on one word at a time, with saccades controlling movement. In contrast, the SWIFT model proposes a parallel attention process, where multiple words are processed simultaneously within the perceptual span. The video compares these models in terms of attention allocation, visual processing, and the influence of parafoveal words, offering insights into how cognitive and linguistic factors affect reading efficiency and comprehension.

Takeaways

• 😀 The E-Z Reader model focuses on sequential processing in reading, where eye movements are closely tied to word fixations and lexical access.
• 😀 The E-Z Reader model struggles to explain higher-level language processing, such as understanding metaphors or abstract connections.
• 😀 The Swift model offers a parallel approach to reading, processing up to four words at a time, including the fixated word and adjacent words in the perceptual span.
• 😀 In the Swift model, attention is distributed across the perceptual span with the fixated word receiving the most attention, followed by a gradual decrease in attention for words on the left and right.
• 😀 The Swift model proposes that linguistic properties of words within the perceptual span influence the attention and fixation times, with more salient words receiving higher attention.
• 😀 An activation field is computed in the Swift model, where words closer to the fixation point (fovea) have higher salience, while words farther away have lower salience.
• 😀 The Swift model suggests that processing multiple words at once can lead to longer reading times when the preview of a parafoveal word is manipulated.
• 😀 The **Parafoveal-on-Foveal Effect** refers to how words in peripheral vision (parafoveal words) can influence fixation times on the currently fixated word (foveal word).
• 😀 Research shows that a delay in parafoveal preview can increase reading times, supporting the idea that peripheral word processing impacts reading efficiency.
• 😀 While parallel models like Swift emphasize simultaneous processing of multiple words, serial models like E-Z Reader predict more sequential processing of words, one after another.

Q & A

What is the primary difference between the E-Z Reader and SWIFT models of eye movement control in reading?

The primary difference is that the E-Z Reader model is a **sequential attention model**, where words are processed one at a time, while the SWIFT model is a **parallel attention model**, where multiple words are processed simultaneously.

How does the E-Z Reader model explain eye movement during reading?

The E-Z Reader model explains eye movement as a process where attention is focused on one word at a time, starting from the fixated word and moving to the next word sequentially, with each word being processed before moving on.

What are the limitations of the E-Z Reader model?

The E-Z Reader model struggles to explain complex aspects of reading, such as the processing of metaphors and abstract connections, as well as how higher-level language processing affects eye movement.

What is the role of 'perceptual span' in the SWIFT model?

In the SWIFT model, the perceptual span refers to the area of text around the fixated word that is processed simultaneously. It typically includes the fixated word, one word to the left, and two words to the right.

How does the SWIFT model describe attention allocation?

The SWIFT model describes attention allocation as being based on a **gradient of attention**, where the fixated word receives the most attention, and the attention decreases as the distance from the fixated word increases.

What is the concept of the 'activation field' in the SWIFT model?

The activation field in the SWIFT model refers to the distribution of attention across the entire perceptual span, with the most salient words (both visually and linguistically) receiving the highest attention, influencing the timing of eye movements.

How does the SWIFT model explain the impact of Parafoveal words on reading times?

The SWIFT model suggests that words in the Parafoveal area (the peripheral words) can influence the fixation time on the central, fixated word, a phenomenon known as the Parafoveal-on-Foveal effect, where the previewing of peripheral words can impact eye movement timing.

What is a significant empirical finding that supports the SWIFT model?

A significant finding supporting the SWIFT model is that when the preview of Parafoveal words is denied or delayed, it increases reading times, as shown in studies where previewing the words ahead of time benefited the readers' speed.

What is the role of linguistic properties in the SWIFT model?

Linguistic properties, such as whether a word is a content word or a function word, affect the **salience** of words within the perceptual span, influencing how much attention is given to different words during reading.

What are the limitations of the SWIFT model?

While the SWIFT model accounts for parallel processing, it has limitations in explaining all aspects of reading, particularly when it comes to **mislocated fixations**, which may influence reading times in ways not entirely predicted by the model.