实战专项：段末限定与结论收窄

唯一新困难：段末限定

段落前部可能给出强趋势，最后一句再限定适用范围。做题时不能读到支持自己预判的证据就停止；段末的 still, only, at least, for now 常决定正确选项强度。

自拟短文

Automatic translation has improved rapidly for routine instructions. In controlled tests, users completed simple tasks almost as accurately with machine-translated text as with professional translation. The results are encouraging, but they apply only to short, predictable documents; legal and medical decisions were not tested.

题目：What is the most appropriate conclusion?

A. Machine translation is ready for every high-stakes document.
B. It performed promisingly on limited routine tasks.
C. Professional translation is now unnecessary.
D. Legal translation was proved inaccurate.

先趋势后护栏

前两句给积极趋势，末句加三道护栏：短、可预测、未测高风险领域。B 保留 promising 和 limited routine tasks。

A、C 删除护栏并扩大应用；D 把“未测试”偷换成“已证明不准”。

未测试的正确处理

证据空白意味着结论未知，不支持正面也不支持负面断言。段末限制经常用来排除看似符合前半段的强选项。

迁移练习

为什么 almost as accurately 不能改成 more accurately？

**答案：**前者表示接近但未说明超过，后者改变比较方向。

独立动作验收

先只根据前两句写预判，再读末句修改预判，明确写出三处被收窄的边界。