University of Toronto, Mississauga, Canada; University of Toronto, Scarborough, Canada; University of Toronto, Canada.
University of Toronto, Mississauga, Canada; University of Toronto, Canada.
Cognition. 2023 Sep;238:105489. doi: 10.1016/j.cognition.2023.105489. Epub 2023 May 8.
Some visual stimuli are consistently better remembered than others across individuals, due to variations in memorability (the stimulus-intrinsic property that determines ease of encoding into visual long-term memory (VLTM)). However, it remains unclear what cognitive processes give rise to this mnemonic benefit. One possibility is that this benefit is imbued within the capacity-limited bottleneck of VLTM encoding, namely visual working memory (VWM). More precisely, memorable stimuli may be preferentially encoded into VLTM because fewer cognitive resources are required to store them in VWM (efficiency hypothesis). Alternatively, memorable stimuli may be more competitive in obtaining cognitive resources than forgettable stimuli, leading to more successful storage in VWM (competitiveness hypothesis). Additionally, the memorability benefit might emerge post-VWM, specifically, if memorable stimuli are less prone to be forgotten (i.e., are "stickier") than forgettable stimuli after they pass through the encoding bottleneck (stickiness hypothesis). To test this, we conducted two experiments to examine how memorability benefits emerge by manipulating the stimulus memorability, set size, and degree of competition among stimuli as participants encoded them in the context of a working memory task. Subsequently, their memory for the encoded stimuli was tested in a VLTM task. In the VWM task, performance was better for memorable stimuli compared to forgettable stimuli, supporting the efficiency hypothesis. In addition, we found that when in direct competition, memorable stimuli were also better at attracting limited VWM resources than forgettable stimuli, supporting the competitiveness hypothesis. However, only the efficiency advantage translated to a performance benefit in VLTM. Lastly, we found that memorable stimuli were less likely to be forgotten after they passed through the encoding bottleneck imposed by VWM, supporting the "stickiness" hypothesis. Thus, our results demonstrate that the memorability benefit develops across multiple cognitive processes.
一些视觉刺激在个体之间始终比其他刺激更容易被记住,这是由于可记忆性的变化(刺激内在特性决定了将其编码到视觉长期记忆 (VLTM) 中的容易程度)。然而,目前尚不清楚是什么认知过程产生了这种记忆优势。一种可能性是,这种优势存在于 VLTM 编码的容量有限瓶颈中,即视觉工作记忆 (VWM)。更准确地说,令人难忘的刺激可能更容易被编码到 VLTM 中,因为它们在 VWM 中存储所需的认知资源较少(效率假说)。或者,令人难忘的刺激可能比容易忘记的刺激更有竞争力,以获得更多的认知资源,从而更成功地存储在 VWM 中(竞争力假说)。此外,如果记忆刺激在通过编码瓶颈后比容易忘记的刺激更不容易忘记(即更“粘性”),那么记忆优势可能会在 VWM 之后出现(粘性假说)。为了检验这一点,我们进行了两项实验,通过在工作记忆任务的背景下操纵刺激的可记忆性、集大小和刺激之间的竞争程度来检查记忆优势是如何出现的。随后,在 VLTM 任务中测试了他们对编码刺激的记忆。在 VWM 任务中,与容易忘记的刺激相比,令人难忘的刺激的表现更好,支持效率假说。此外,我们发现,当直接竞争时,与容易忘记的刺激相比,令人难忘的刺激也更能吸引有限的 VWM 资源,支持竞争力假说。然而,只有效率优势转化为 VLTM 中的表现优势。最后,我们发现,令人难忘的刺激在通过 VWM 施加的编码瓶颈后更不容易被遗忘,支持“粘性”假说。因此,我们的结果表明,记忆优势是在多个认知过程中发展起来的。