PTE口语改革后DI & RL新模板

前因

2024年10月31日，PTE官网突然更新了一篇文章，称PTE学术考试即将迎来升级，并且不会影响现有的备考方式和计划，只是将打分变得更加科学和系统——判卷将引入专家人工审查。

视频里的话术听起来还挺好笑的，因为他们说「真人专家一直都是打分系统的一部分」，但这多少有些欲盖弥彰了。我们都心知肚明，PTE备考就是一个与机器斗智斗勇的考试。

几个要点：

1. 只影响口语的DI和RL
2. 只影响Academic和UKVI考试，不影响Core
3. 不影响出分时间
4. 人工审查的专家看不到原题，只会听音频，只审查内容评分，流利度和发音仍然由机器负责判定
5. 改动2024年11月4日起生效

之前在上课的时候，我就一直在和学生说，不知道培生是故意的还是不小心，明明知道大量学生通过背模板通过考试，但是却迟迟不对考试做改变。现在的AI技术理应足够发达到检测学生的内容是否能很好地回答题目。

当然或许是出于成本和口碑的考量，没有一步走到位也情有可原，但改革迟早是要来的。现在引入人工审查是第一步，先上两个题型，之后其他套模板重灾区如SST和WE我估计也快了，PTE将不再是那个「宇宙第二简单的语言考试」了。

模板

DI

在我之前写的PTE听力口语快速备课指南里有提到，原本市面上DI的模板分成大致可分成3个流派：

1. 老实人派——模板为长句子，需要学生自己组织语言描述图片
2. 废话+流水账派——先说一大堆废话，然后把画面上的关键词按顺序念一遍
3. 没有感情派——用and also about连接一起，看见什么说什么

在很长一段时间里，我都让学生练习的没有感情派，因为这能很好地提高学生整体流利度，并且命中得分关键词的概率相对来说会高一些，背诵记忆模板的负担也小了很多。

但是这次改革之后，官方DI Test Tips点名批评了这种机械的描述（describe the image in a mechanical way），因此无论是and also about还是from the picture we can see...，这次改革之后都必不能再用了。

另外一种被点名批评的是，使用背诵过的句子和套句（use memorised responses or templated sentences），所有企图通过不直接描述图片内容增加流利度的方法，都将会被判定为0分，因此流派2也被彻底封杀。

那究竟怎么做，才能既避免机械地说话，又能够让学生继续使用模板呢？

我觉得模板套句的长度很关键。假如你让大街上随便一个路人来描述一张图片，大概率每个人都会用The picture is about...或者The picture talks about来开头，然后用X shows Y来描述细节。这不是模板，这是非常正常的语言描述。

我们仍然需要用一定的套路来衔接内容，但是再也不能像以前那样抛弃大脑，拥抱模板了。这更像是雅思口语的框架，大框架是这样，剩下的细节自己填入，主要目标仍然是保证流利度，而不是关键词。

这并不是世界末日，我觉得它仍然是可以操作的。

在网上搜寻一番之后，我觉得目前Skills PTE Academic的模板比较契合我的理念，博主也是很多国内学生口中津津乐道的“印度小哥”。

The image provided is a [图片类型] that gives information about [主题].

(如果有坐标轴) While X axis talks about __ , the Y axis __ talks about ____ .

1. It is evident that ____ (最大值).
2. Conversely, ____ (最小值).
3. Another interesting point is ____ (随便一个趋势或观察)

Overall, this image illustrates [主要信息或整体趋势]

这个模板相较之前多了很多连接词，While... Conversely... Another interesting point...让整个答案听起来更像人类会说的话，从句后面所接的内容用最简单的句型正常描述即可。

在实际操作中，我可能会把It is evident that改成It is clearly seen that...，因为大部分国内学生对这个词都很陌生；我还会把conversely改成on the other hand，那些基础只有40、50分的学生绝对记不住这歌词；最后的illustrate可能也要改成show。

下面是随便几个题型的示范——

折线图：

The image provided is a line graph that gives the information about most cellphone owners in the U.S. have smartphones.

It is clearly seen that over 60% of the U.S. adults have any cellphone in 2005, on the other hand, over 85% of the people have a smart phone after 2020.

Another interesting point is that people start to have smartphone since 2010.

Overall, the image shows the rise of the smartphone.

柱状图：

The image provided is a bar graph that gives the information about the number of U.S. Households keeping pets.

It is clearly seen that the category with the highest number is dogs, which is 63.4million, on the other hand, the one with the lowest number is saltwater fish, which is 1.6million.

Another interesting point is that the number of horses is the same as that of saltwater fish.

Overall, the image shows how many kets U.S. Households are keeping.

地图题：

The image provided is a map that gives the information about distribution of harlequin ladybird.

It is clearly seen that there were many birds in 2004, which is shown in the red colour, on the other hand, in 2005, there weren't many left, which is shown in the black colour.

Another interesting point is that the birds were all at the southwest of England.

Overall, the image shows the decrease of the number of the ladybird.

流程图：

The image provided is flow chart that gives the information about supply chain management.

It is clearly seen that first, you need to have raw materials and change them into components, on the other hand, you give the components to manufacturer and retailer.

Another interesting point is that consumer is the last one in the supply chain.

Overall, the image shows the process of supply chain management.

总体来说，在之后的课上，像雅思那样的遣词造句练习还是很有必要的。真的只需要把0%的大脑开机到10%就可以了。

DI仍然不看逻辑，不看语法，只看关键词和流利度，并且仍然以流利度优先，如何用最简单地废话说出最流畅的答案依旧是重点。

RL

按照之前解读的思路理解，RL模板的设计也遵循了「将长套句改成人类会说的短句」原则，下面展示的模板仍然来自Skills PTE Academic。

The lecture discusses __.
In the beginning, the speaker explained about __.
Later, the speaker said __.
Here, the speaker also mentioned __.
In conclusion, the lecture effectively summarised __.

在实战中，我会和学生说，discuss、explained about、said、mentioned的顺序不重要，你只需要用这几个词引出后面的内容即可。

同样的，我可能会再提供几个连接词，让学生学着用不同的串词引导内容，而不是机械地套用later和here。

至于总结的内容，我们的策略稍微有一些改变，因为我们不再能够使用and also about串联关键词了。在听听力的时候，我们需要尽量写下10个以上的关键词，然后用最简单的句子将内容串联起来。

下面是示范：

#286 Venus

There is a picture, sort of artist's impression， before the space age of what Venus might be like on its surface and so this was looking at the planet Venus, it was science fiction and science fact all the way up to 56 before the start of the space age but it wasn't completely disproved, this idea of a really sort of lush environment on Venus until 1967, which is when the first measurements in detail were done at Venus.

So Mariner four and Mariner five confirmed the feeling from an earlier space mission that in fact the surface of Venus was not like this at all, but extremely hot and, and also that the clouds were made of sulphuric acid so there wasn't a nice water cycle like is going on in this picture and so, that it had to wait for these in situ measurements by space craft to actually do that and so Venus turned out not to be quite as Earth like as we thought and I'll sort of tell you about some of the latest results from Venus Express, which, which they actually there are some Earth like features, but to a large extent, it's not like the Earth.

假设对于这篇文章，我们只能听到以下关键词：picture, Venus, science, space, environment, hot, nice, Earth。

在每句模板后面，我们需要在尽可能短的时间内用1-2个单词造句。

答案：

The lecture discusses a picture of Earth.

In the beginning, the speaker explained about the environment of Venus is very hot.

Later, the speaker said the environment of the Earth is very nice.

Here, the speaker also mentioned the picture is very important for the space science .

In conclusion, the lecture effectively summarised the science of the Earth and Venus.