Mechanisms Responsible for the Animacy Effect in Memory: Examining the Role of Animacy on Episodic and Working Memory

Abstract

Words representing animate entities are recalled and recognised more accurately than inanimate words, which has been termed the animacy effect. Understanding the processes responsible for generating the animacy effect is important for a complete understanding of how memory encodes, stores and retrieves words representing animate and inanimate entities. This research project consisted of six experiments, each of which examined the animacy effect in different paradigms. Experiment 1 examined the animacy effect in free recall to determine whether the animacy effect was impacted by the serial position of the words in the study list. Animate words were recalled more accurately than inanimate words and importantly, the magnitude of the animacy effect did not differ significantly between words appearing at the beginning (initial component), middle and at the end of lists (final component). These results provided evidence that the animacy effect is independent of where in a list the words appear and that processes associated with serial position do not explain the animacy effect. Experiment 2 examined whether the build-up of proactive interference (PI) differed between the broad animate and inanimate semantic categories using the Brown-Peterson paradigm. There was a significant build-up of PI over trials 1 to 4 for both animate and inanimate words. The overall magnitude of the build-up of PI did not differ significantly between animate and inanimate words but animate words were more resistant to the build-up of PI between trials 2 and 3 than inanimate words. Animacy effects were present for the first three trials but were eliminated on trial 4, as the build-up of PI peaked. There was a significant release from PI after switching from animate and inanimate words to the opposite category of animacy, as well as when switching to words representing colours. Release from PI was stronger when switching to animate words than switching to inanimate words. The results from Experiment 2 suggest that the animacy effect in free recall is not explained by differences in the build-up of PI between animate and inanimate words. Experiments 1 and 2 examined long-term memory (LTM), while Experiments 4 to 6 examined working memory (WM). No known research had examined the animacy effect in WM and such an examination would provide information on whether the animacy effect extends to WM. Experiments 3 and 4 found that serial order recall was more accurate for animate words than inanimate words at set sizes 4, 5 and 6 but not at set size 3, providing evidence that animacy effects extend to WM. Experiments 3 and 4 also examined different WM maintenance processes. Animacy effects were present when participants undertook directed attention, articulatory rehearsal or were given no instruction on how to remember the words (Experiment 3), as well as under articulatory suppression to prevent rehearsal (Experiment 4). This suggests that animacy effects are present for multiple WM maintenance process. Experiments 5 and 6 examined item-context binding in WM to understand how animacy may enhance memory for contextual details. In Experiment 5, the n-back task (2-back and 3-back) was used to examine the creation and updating of item-temporal order bindings for animate and inanimate words. Responding was more efficient for animate words than inanimate words, indicating animacy effects extend to temporal order binding. Experiment 6 examined item-location binding using the Local Recognition Task. Responding was more efficient for animate words than inanimate words at set size 5 but not set sizes 3 and 4, suggesting that animacy enhances item-location binding when WM demands are high. The results of Experiments 5 and 6 provide evidence that the animacy effect in WM extends to contextual details. The animacy effect in WM may be the result of animate words placing less demand on WM capacity than inanimate words, particularly when task demands are high. The findings of this project are consistent with functional evolutionary theory, as animacy effects in both LTM and WM would likely have increased the probability of ancestral survival. While the demographically homogenous samples limit the generalisability of the findings to younger adults, the findings suggest that animacy is an important determinant of recall in both LTM and WM.