Vocabulary acquisition is a lifelong journey. As we navigate through life, we continuously encounter new words and, crucially, new meanings for words we already know. This expansion of Familiar Meaning is especially pertinent in our rapidly evolving world, driven by technological advancements and shifting social landscapes.
In today’s digital age, adults are frequently challenged to grasp novel interpretations of familiar terms. Consider the word “troll,” once relegated to fairy tales, now commonly denotes someone who intentionally provokes arguments online (Hulme et al., 2019; Rodd, 2020). Similarly, entering a new hobby or professional field often requires learning specialized meanings for familiar words. Gamers, for instance, understand “farming” not in its agricultural sense, but as repetitive in-game actions to accumulate resources (Eligio & Kaschak, 2021; Rodd et al., 2012, 2016). Even geographical relocation can introduce dialectal variations, where a familiar word like “piece” might take on a new meaning, such as “sandwich” in Scots dialect.
The brain’s ability to adapt to these evolving familiar meanings is remarkable. However, integrating new interpretations with existing knowledge of words presents a cognitive challenge. Our mental lexicon, a vast repository of word knowledge, must accommodate these additions without disrupting our grasp of established meanings. The question then arises: how do we efficiently learn and consolidate these new familiar meanings? This article delves into the fascinating role of sleep in this process, exploring how overnight consolidation may facilitate the learning of new meanings for familiar words in adults.
The Complementary Learning Systems Theory and Vocabulary Acquisition
The Complementary Learning Systems (CLS) theory (Davis & Gaskell, 2009) offers a compelling framework for understanding how new vocabulary, including new familiar meanings, integrates with our pre-existing linguistic knowledge. This theory, grounded in the CLS model of memory (McClelland et al., 1995) and studies of spoken word learning, proposes a two-stage process for vocabulary acquisition.
Initially, new word information, including novel meanings for familiar words, is encoded in the hippocampus, a brain region crucial for episodic memory. However, for these new meanings to become firmly embedded in our semantic memory, residing in the neocortex, an active period of offline consolidation is required, particularly during sleep. This sleep-related consolidation is theorized to be the mechanism by which new vocabulary becomes a stable part of our mental lexicon, enabling faster recognition and seamless integration with existing word knowledge (Tamminen & Gaskell, 2013; Leach & Samuel, 2007).
This CLS model aligns with broader systems consolidation theories that emphasize sleep’s active role in memory strengthening (Dudai, 1996, 2004; Marr, 1970; McClelland et al., 1995). It shares similarities with the active systems consolidation (ASC) model (Born & Wilhelm, 2012; Diekelmann & Born, 2010), although distinctions exist in the specific mechanisms involved in transferring memories from the hippocampus to the neocortex.
In essence, the CLS theory posits that sleep actively safeguards newly learned vocabulary, including new familiar meanings, by reinforcing and stabilizing memory traces, protecting them from interference. This “active account” predicts that learning new meanings just before sleep is particularly beneficial, as these items are readily available in episodic memory for efficient consolidation. Conversely, an alternative “passive account” suggests sleep’s benefit is simply due to reduced interference during sleep, protecting new meanings from the constant stream of linguistic input encountered during wakefulness (Jenkins & Dallenbach, 1924). While the debate on the precise nature of sleep’s memory benefits continues (Ellenbogen, Payne, et al., 2006), growing evidence supports an active role, with studies demonstrating that manipulating brain oscillations and targeted memory reactivation can enhance word learning (Ngo et al., 2013; Cairney et al., 2014).
Diagram illustrating the experimental design of Experiment 1, comparing a sleep group learning new familiar word meanings before sleep and a wake group learning before a period of wakefulness to investigate the impact of sleep on vocabulary learning.
Evidence from Word Form Learning
Much of the evidence supporting the CLS theory originates from studies focusing on spoken word form learning. For example, Gaskell & Dumay (2003) demonstrated that newly learned word forms that sounded similar to existing words were not immediately integrated into the mental lexicon. Participants learned non-words like cathedruke (similar to cathedral). Testing immediately after learning showed no interference with the recognition of the familiar word cathedral. However, when tested after a delay that included sleep, participants showed inhibited access to cathedral, suggesting that the new word form cathedruke had become integrated and was now competing for lexical access. This implied that offline consolidation, especially during sleep, is crucial for integrating new word forms into our mental lexicon (Tamminen & Gaskell, 2013). Subsequent studies have replicated these findings, further solidifying the link between sleep and lexical integration of new word forms (Davis et al., 2009; Dumay & Gaskell, 2007; Tamminen et al., 2010; Tamminen & Gaskell, 2008). Several of these studies have specifically highlighted the importance of sleep, distinguishing it from mere passage of time (Dumay & Gaskell, 2007) and linking specific sleep components like sleep spindles to lexical integration (Tamminen et al., 2010).
However, counter-evidence suggests that lexicalization can occur without sleep. Kapnoula et al. (2015) showed that new word forms could compete with existing words immediately after training, without sleep consolidation. Similarly, Lindsay and Gaskell (2013) found immediate integration of novel word forms with spaced learning and exposure to competitor words. These studies indicate that under certain conditions, immediate integration of new word forms is possible, challenging the necessity of sleep-dependent consolidation.
Learning Word Meanings: Beyond Word Forms
Studies focusing solely on word forms may lack ecological validity, as vocabulary learning in real life is rarely isolated from meaning. Most vocabulary acquisition is incidental, occurring naturally through conversations, reading, and media (Nagy et al., 1985, 1987). Research by Henderson et al. (2015) using naturalistic story reading found that new words only integrated with existing lexical knowledge after offline consolidation, highlighting the importance of considering semantic information. Indeed, incorporating semantic information into vocabulary learning appears to engage complementary memory systems differently than word form learning alone (Takashima et al., 2014).
Studies combining new word forms with semantic information, using methods like pairing non-words with invented meanings (Clay et al., 2007), attaching meaningful affixes (Tamminen et al., 2012), or using low-frequency words (van der Ven et al., 2015), have largely supported the CLS theory. These studies suggest that semantic integration, like word form integration, benefits from overnight consolidation, implying that processing word meaning also requires time for integration into semantic memory (van der Ven et al., 2015).
However, similar to word form learning, evidence exists for semantic integration without sleep-dependent consolidation, particularly in language production (Oppenheim, 2015). Encoding methods also play a crucial role. Coutanche and Thompson-Schill (2014) demonstrated that “fast-mapping” (inferring meaning through elimination) led to immediate lexical integration, while explicit encoding required consolidation. These findings suggest that when learning conditions promote online connections between new and existing knowledge, offline consolidation may not be a prerequisite for integration (Fang et al., 2017; Himmer et al., 2017), as seen in fast mapping (Coutanche & Thompson-Schill, 2014; Himmer et al., 2017), spaced learning (Lindsay & Gaskell, 2009), and test-enhanced learning (Antony et al., 2017).
The Specific Case of Learning New Familiar Meanings
Crucially, previous research often conflates learning new word forms and new meanings simultaneously. Learning a new meaning for a familiar word with existing semantic associations is a distinct process. Only a few studies have directly investigated this specific scenario, focusing on how assigning a novel invented meaning to an unambiguous word impacts our understanding (Fang et al., 2017; Fang & Perfetti, 2017; Hulme et al., 2019; Hulme & Rodd, 2021; Maciejewski et al., 2020; Rodd et al., 2012). While these studies explored the immediate and delayed memory of these new meanings, they did not specifically examine the role of overnight consolidation and disentangle sleep’s effect from the passage of time.
To address this gap, the following experiments directly investigated the impact of sleep on learning new meanings for familiar words acquired incidentally through story reading.
Experiment 1: Sleep vs. Wakefulness in Learning New Familiar Meanings
Experiment 1 aimed to determine if overnight consolidation benefits the learning of new meanings for familiar words. The study employed a between-groups design, training participants on new meanings incidentally through story reading, mirroring natural vocabulary acquisition (Hulme et al., 2019; Hulme & Rodd, 2021). Participants were trained either in the evening or morning, followed by a 12-hour delay period of either sleep or wakefulness, and finally a test session.
Explicit knowledge of the new meanings was assessed using two measures: a cued recall test, requiring participants to generate the new meaning when presented with the word, and a multiple-choice meaning-to-word matching test. Both productive (cued recall) and receptive (multiple-choice) measures were used to ensure comprehensive assessment, capturing potential learning differences regardless of overall performance level. Additionally, an implicit measure, a semantic relatedness judgement task, measured reaction times (RT). This task assessed semantic integration by examining if learning a new meaning for a familiar word (e.g., “foam”) interfered with accessing its original meaning when judging its relatedness to a probe word (e.g., foam-soap vs. foam-belt) (Fang et al., 2017; Fang & Perfetti, 2017; Maciejewski et al., 2020; Rodd et al., 2012).
Based on prior word form learning studies (Dumay et al., 2005; Henderson et al., 2015; Takashima et al., 2014; Tamminen et al., 2010), it was hypothesized that participants who slept between training and testing would exhibit better explicit memory for the new word meanings. Furthermore, it was predicted that the sleep group would be slower in the semantic relatedness task for trained words, indicating increased competition between the new and old meanings, suggesting successful lexical integration.
Experiment 1: Method Highlights
Eighty-four participants were recruited and divided into a sleep group and a wake group. Stimuli consisted of 16 unambiguous words assigned new, unrelated meanings, incorporated into four short stories. Participants read two stories containing eight target words. Explicit memory was tested using cued recall and multiple-choice tests. Implicit memory was assessed using a semantic relatedness judgement task. Sleepiness was measured using the Stanford Sleepiness Scale (SSS).
Experiment 1: Key Findings
The sleep group demonstrated significantly better explicit memory for new word meanings in both cued recall and multiple-choice tests compared to the wake group. However, the semantic relatedness judgement task showed no significant differences between groups or training conditions.
Graph displaying the mean percentage of correct responses in the cued recall test for Experiment 1, showing improved recall for the sleep group compared to the wake group in learning new familiar word meanings.
Graph illustrating the mean percentage of correct responses in the multiple-choice test for Experiment 1, indicating higher accuracy for the sleep group over the wake group in matching new meanings to familiar words.
Experiment 1: Discussion
Experiment 1 provided evidence that a 12-hour delay period including sleep enhances the learning of new meanings for familiar words compared to a wakeful delay. The explicit memory benefits aligned with previous word form learning studies (Dumay & Gaskell, 2007; Tamminen et al., 2010). However, the absence of effects in the semantic relatedness task raised questions about whether the new meanings were truly integrated into the lexicon or if the task lacked sensitivity. Participant dropout, although potentially a confound, was considered, and time-of-day effects were deemed unlikely due to non-significant SSS differences between groups. The study suggested a sleep benefit for explicit memory of new familiar meanings, but further investigation was needed to clarify the underlying mechanisms and lexical integration.
Experiment 2: Disentangling Active Consolidation from Passive Protection
Experiment 2 aimed to further investigate the sleep benefit observed in Experiment 1, utilizing a “12:12” design to differentiate between active sleep-related consolidation and passive protection from interference. Participants were divided into morning-test (AM-test) and evening-test (PM-test) groups. Both groups underwent two training sessions, 12 hours apart, learning half of the target words in each session. Testing occurred 12 hours after the second training session for both groups, allowing for comparisons of 12-hour and 24-hour retention intervals and the role of sleep timing.
It was hypothesized that if active consolidation drove the sleep benefit, the AM-test group would perform better in the 12-hour delay condition (sleep immediately after learning), while the PM-test group would excel in the 24-hour delay condition (sleep opportunity after initial wake period). Conversely, the passive protection account predicted superior performance for the AM-test group in the 12-hour delay condition only, due to minimized daytime linguistic interference. Semantic relatedness task predictions varied depending on the strength of learning and the specific account.
Experiment 2: Method Highlights
Eighty-four new participants were recruited for Experiment 2, again split into AM-test and PM-test groups. Materials and procedures were similar to Experiment 1, with minor adjustments to the semantic relatedness task timing to potentially enhance sensitivity. The design included 12-hour and 24-hour delay conditions, manipulated within-participants.
Experiment 2: Key Findings
Cued recall results showed a significant interaction between group and training condition. Replicating Experiment 1, the AM-test group outperformed the PM-test group in the 12-hour delay condition. However, no significant group difference was found in the 24-hour delay condition. Neither the multiple-choice test nor the semantic relatedness task showed significant effects of training or group.
Graph displaying the mean percentage of correct cued recall responses in Experiment 2, highlighting the interaction between test group and delay, with the AM-test group showing better recall at 12-hour delay for learning new familiar word meanings.
Experiment 2: Discussion
Experiment 2 replicated the 12-hour sleep benefit for explicit memory of new familiar meanings observed in Experiment 1, specifically in the cued recall task. However, the crucial finding was the absence of a sleep benefit in the 24-hour delay condition. This pattern contradicts the active consolidation account’s prediction of a PM-test group advantage at 24 hours. The results are more consistent with the passive protection hypothesis, suggesting the 12-hour sleep benefit may be due to reduced linguistic interference during sleep rather than active consolidation. Time-of-day effects and differential attrition were again ruled out as primary explanations. The lack of effects in the semantic relatedness and multiple-choice tasks mirrored Experiment 1’s findings, suggesting potential limitations in task sensitivity or learning intensity.
General Discussion: Implications for Understanding Familiar Meaning Acquisition
These experiments provide compelling evidence that sleep enhances explicit memory for newly learned meanings of familiar words after a 12-hour period. This 12-hour sleep benefit was consistently observed in cued recall across both experiments. However, the absence of a sustained sleep benefit at 24 hours in Experiment 2 challenges the notion that active overnight consolidation is the primary driver.
The findings lean towards a “passive protection” account, where sleep’s primary role is to shield new familiar meanings from interference encountered during wakefulness. This suggests the 12-hour sleep benefit might be transient, diminishing once individuals experience a full day of linguistic input. While this study highlights the immediate advantage of learning new familiar meanings before sleep, it questions the long-term real-world benefits if these gains are quickly eroded by subsequent wakefulness.
It is crucial to acknowledge that these results do not definitively rule out active consolidation. The absence of a 24-hour sleep benefit could be due to various factors, including interference between training sessions or the possibility that passive protection and active consolidation contribute additively to the observed 12-hour effect.
The consistent null findings in the semantic relatedness task across both experiments warrant further consideration. Possible explanations include insufficient learning intensity due to the incidental learning paradigm and the relatively low number of new meanings learned per session, potentially hindering the detection of subtle competition effects in lexical access. Future research could explore more intensive learning paradigms or enhance the perceived relevance of the new meanings to boost learning and task sensitivity.
Future Directions and Conclusion
Future studies should incorporate polysomnographic recordings to directly investigate the relationship between specific sleep stages and the consolidation of new familiar meanings. Combining neurophysiological measures with behavioral tasks will provide a more nuanced understanding of sleep’s role in vocabulary learning.
In conclusion, this research demonstrates a clear 12-hour sleep benefit for explicit memory of new meanings for familiar words, likely attributable to passive protection from interference. While the active consolidation account cannot be entirely dismissed, the current findings underscore the importance of considering passive protection as a significant factor in sleep-related vocabulary learning, particularly when acquiring new familiar meanings. Understanding these mechanisms is crucial for developing effective strategies to enhance vocabulary acquisition throughout life, especially in our ever-evolving linguistic landscape where adapting to new familiar meanings is increasingly essential.
Data Accessibility
All materials, data, and analysis scripts are publicly available on the Open Science Framework (OSF): https://osf.io/m3pj6. Experiment 2 was preregistered on OSF: https://osf.io/uvgp4.
References
[References] bibliography.bib
