Abstract This article focuses on two types of constructional effects of indirect evidential marking in Harakmbut (isolate, Peru). Both types originate in a clash of interpretation: the use of indirect evidential marking indicates a shift of perspective away from the speaker (as if they did not witness the event, thus disclaiming epistemic authority), while the events referred to are in principle directly accessible to them. As the signalled shift is not fully realized in interpretation, the effects will be characterized as showing perspective persistence. The first type involves constructions with a first person agent, and indirect evidential marking is found to produce the interpretation that the speaker performed the action referred to unintentionally, finding out about the outcome of this action only later. Other types of non-volitional events – without pragmatic inference on the part of the speaker – are found not to carry indirect evidential marking; they use different linguistic means to signal non-volitionality. The second type involves constructions with impersonal predicates referring to the cycle of the sun, and the use of indirect evidential marking yields emphasis on the completion of the event referred to. It is proposed that both types of effects can be explained in terms of endpoint emphasis (cf. DeLancey 1985).

中文翻译：

---

Harakmbut 中间接证据标记的结构效应

摘要 本文重点讨论 Harakmbut（秘鲁，孤立）间接证据标记的两种结构效应。两种类型都源于解释的冲突：间接证据标记的使用表明从说话者的角度转移（好像他们没有目睹事件，因此否认认知权威），而所指的事件原则上是可直接访问的给他们。由于信号转换在解释中没有完全实现，效果将被表征为显示透视持久性。第一种类型涉及第一人称代理的结构，发现间接证据标记产生解释，即说话者无意中执行了提到的动作，后来才知道该动作的结果。其他类型的非自愿事件——没有说话者的语用推理——被发现不带有间接证据标记；他们使用不同的语言手段来表示非意志性。第二种类型涉及带有非人称谓词的结构，指的是太阳的周期，间接证据标记的使用强调所指事件的完成。有人建议，这两种类型的影响都可以用终点强调来解释（参见 DeLancey 1985）。间接证据标记的使用强调所指事件的完成。有人建议，这两种类型的影响都可以用终点强调来解释（参见 DeLancey 1985）。间接证据标记的使用强调所指事件的完成。有人建议，这两种类型的影响都可以用终点强调来解释（参见 DeLancey 1985）。