🚨New Paper Alert🚨 - Revising Noun Predictions

Fresh evidence on how English measure phrases help us revise our predictions, studied using visual-world eye-tracking.

Celebrating my first publication

I’m excited to share that my very first research paper as a PhD student has been published online (and will soon be in print) in the journal Language, Cognition and Neuroscience ! This project was the first one I worked on at the start of my PhD, and I owe a huge thanks to my incredible supervisor, Wing-Yee, for her constant support and guidance. I couldn’t have done it without her.

The paper’s title is a bit technical: Revising Noun Predictions Based on English Measure Phrases: Evidence from Visual-World Eye-Tracking. But don’t worry - I’ll explain it in simpler terms! You can click here to read the entire paper.

What’s the paper about?

In everyday conversations, our brains are always trying to guess what’s coming next. For example, if I say I went to Starbucks and bought…, you probably expect the next word to be coffee because that’s something you’d typically buy at Starbucks. But what if I then say … a piece of ? Your brain would quickly revise its guess from coffee to something like cake instead.

Our study looks at how fast the brain can change these predictions when new information comes in. To do this, we used a method called visual-world eye-tracking. This technique measures where people are looking on a screen while they listen to sentences. By seeing which objects they look at, we can tell what they’re predicting.

For example, in our experiment, participants saw pictures of different objects on a screen: a cup of coffee, a piece of cake, and two unrelated things like a chair and a bird. We tracked where they were looking as they heard the sentence, I went to Starbucks and bought… At this point, most people looked at the picture of coffee, showing they expected coffee to be mentioned. But when the sentence continued with …a piece of…, they quickly shifted their gaze to the cake, revising their original guess. This shows how fast people can adjust their predictions when the new information doesn’t match what they first expected.

What’s unique about this study?

My supervisor and one of her previous students did a similar study with Mandarin Chinese. In Chinese, there’s something called classifiers that come between a number and a noun (like saying 一朵花 <one CLduo flower> or 两本书 <two CLben books>, where CL stands for classifier). They found that Mandarin speakers could use these classifiers to quickly adjust their predictions.

English doesn’t have classifers but has a similar structure called measure phrases - phrases like a piece of, a glass of etc. Unlike classifiers in Chinese, measure phrases in English are optional. For instance, you can say three apples in English, but in Chinese, you’d have to say 三颗苹果 (three CLke apples).

Our study compared how quickly people revise their predictions in English and Chinese when they hear these structures. We found no significant difference in how fast listeners in both languages make these revisions. This suggests that it’s not about whether the structure is required, but how informative it is for helping listeners adjust their expectations.

Why is this important?

Being able to adjust predictions based on what we hear helps us understand language quickly and accurately. Our study provides new evidence showing just how fast and flexible this process can be!

If you are interested in further discussion with me, please find my contact information at the bottom of the About page.

Below is the Chinese version of this article.

下面是这篇短文的中文版。

庆祝我的第一篇论文发表

我作为博士生的第一篇研究论文已经在线发表了（即将印刷出版），发表在《语言、认知与神经科学》期刊上！这是我在博士生涯初期参与的第一个项目，感谢我超棒的导师 Wing-Yee 在整个项目中的持续支持和指导，对我非常有帮助。

我的论文题目是《基于英语度量短语的名词预测更新：来自视觉世界眼动追踪的证据》。接下来，我会用简单的语言来解释这篇文章的内容。点击这里可以阅读整篇论文。

论文内容是什么？

在日常对话中，我们的大脑会不断猜测接下来会出现的词。比如，当我说 “ 我去星巴克买了…… ”，你可能会预测下一个词是 “ 咖啡 ”，因为我们经常会在星巴克买咖啡。但如果我接着说 “ …… 一块 …… ”，你的大脑会迅速把预测从 “ 咖啡 ” 更新为 “ 蛋糕 ” 之类的东西。

我们的研究探讨了当新的信息出现时，大脑能够多快地改变这些预测。我们使用了一种叫做 “ 视觉世界眼动追踪 ” 的方法。这种技术通过追踪人们在听句子时，看向屏幕上的哪个物体，来判断他们的预测。

在我们的实验中，参与者看到屏幕上展示了几张不同的物品图片：一杯咖啡、一块蛋糕，还有两个无关的物体，比如一把椅子和一只鸟。当他们听到 “ 我去星巴克买了…… ” 时，大多数人会看向咖啡的图片，表明他们预期接下来会听到 “ 咖啡 ”。然而，当继续听到 “ …… 一块 …… ” 时，他们迅速转移视线，看向蛋糕，更新了他们最初的预测。这表明当新的信息不符合预期时，人们可以很快调整自己的预测。

这项研究的独特之处是什么？

我的导师和她之前的一位学生曾经用汉语做过类似的研究。在汉语中，存在一种叫做 “ 量词 ” 的结构，它出现在数字和名词之间（例如 “ 一朵花 ” 或 “ 两本书 ”，其中 “ 朵 ” 和 “ 本 ” 是量词）。他们发现，汉语母语者可以通过量词快速更新他们的预测。

虽然英语中没有量词，但有类似的结构叫做 “ measure phrases ” （我在前文将它翻译为 “ 度量短语 ”），比如 “ a piece of ”、“ a glass of ” 等等。与汉语不同，英语中的这类短语不是必须使用的。比如，在英语中可以说 “ three apples ”, 也可以说 “ three bags of apples ”，但在汉语中必须说 “ 三颗苹果 ” (three CLke apples) 和 “ 三袋苹果 ”（three CLdai apples）。

我们的研究比较了，英语和汉语母语者在听到这些结构时，更新预测的速度。我们发现他们在更新预测的速度上没有显著差异。这表明，影响预期更新的关键因素，不是某种语言结构是否是必须的，而是它能给我们多大的信息量来调整预期。

为什么这很重要？

根据新的信息，迅速调整预期，有助于我们更快速、更准确地理解语言。我们的研究再次显示了预期更新这一过程的快速性和灵活性！

如果您有兴趣与我进一步讨论，可以在 About 页面底部找到我的联系方式。