Comparative psychologists study cognition by characterizing the behavior of individual species and explicitly comparing behavior across species. We use the extensive comparative literature on transitive inference (TI) as a case study to evaluate four central methodological questions that continue to be debated in the field of comparative psychology: 1) Are contextual variables sufficient to explain species differences in cognition? 2) Can cognitive performance be accounted for by associative processes alone? 3) Can we determine the cognitive mechanisms by which animals solve tasks? and 4) What is the role of ecologically driven hypotheses in comparative psychology? Although contextual variables and associative processes undeniably influence choice behavior in TI tasks, neither is sufficient to explain all performance. Instead, multiple distinct cognitive mechanisms, including associative processes, logical inference, and spatial representations, can and do result in successful TI performance. TI is not a unitary task solved using a single mechanism; multiple processes are recruited, with their degree of involvement dependent on context, species, and evolutionary pressures. This suggests that rather than asking whether animals possess a certain cognitive ability, research should focus on differences in when and how species employ tools from what is often a reasonably similar cognitive toolbox. We join others who have proposed that a main goal of comparative psychology should be to determine how animals solve cognitive tasks, through minimizing and studying the influence of contextual variables, evaluating the contributions of associative processes, clearly characterizing and testing alternative cognitive mechanisms, and using strong evolutionary hypotheses to guide predictions. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

中文翻译：

---

认知是否因物种而异，我们如何知道？传递推理研究的经验教训。

比较心理学家通过表征单个物种的行为并明确比较跨物种的行为来研究认知。我们使用关于传递推理 (TI) 的大量比较文献作为案例研究来评估比较心理学领域继续争论的四个核心方法论问题：1) 上下文变量是否足以解释物种在认知方面的差异？2) 认知表现能否仅由联想过程来解释？3）我们能否确定动物解决任务的认知机制？4）生态驱动的假设在比较心理学中的作用是什么？尽管上下文变量和联想过程不可否认地影响 TI 任务中的选择行为，但这都不足以解释所有的表现。反而，多种不同的认知机制，包括联想过程、逻辑推理和空间表示，可以而且确实导致成功的 TI 表现。TI 不是使用单一机制解决的单一任务；招募了多个过程，其参与程度取决于环境、物种和进化压力。这表明，与其询问动物是否具有某种认知能力，不如研究物种在何时以及如何使用通常相当相似的认知工具箱中的工具的差异。我们与其他人一起提出比较心理学的主要目标应该是确定动物如何解决认知任务，通过最小化和研究上下文变量的影响，评估联想过程的贡献，清楚地表征和测试替代认知机制，并使用强大的进化假设来指导预测。（PsycInfo 数据库记录 (c) 2021 APA，保留所有权利）。