The role of explicit and implicit learning processes in concept discovery

Analysis of individual learning curves and concurrent verbal protocols from three experiments concerning discovery of a non-salient verbal concept and a pictorial analogue (Chinese ideograph) of the concept show that a substantial transition phase occurs in which discrimination of exemplars from non-exemplars of the concept is above chance but not yet asymptotic. Under most conditions the ability to verbalize knowledge of the concept occurred almost simultaneously with the onset of the transition phase. However, the addition of noise in the form of false feedback (Experiment 3) created a temporary dissociation between task performance and verbalizable knowledge. Additional results suggest that individual hypothesis revision/rejection strategies affect the length of the transition phase of learning, whereas the size of the domain of hypotheses being sampled affects the number of trial blocks before the transition phase begins. The effect of feedback error on the relation between early rates of hypothesis generation and subsequent transition phase length also suggests that a strategy of quick rejection of falsified hypotheses becomes less adaptive in noisy task environments (e.g. when there are many exceptions to a rule or the concept is probabilistic). Finally, failure to find effects of variables known to affect implicit learning suggests that implicit learning processes do not play a large role in the discovery of this type of concept.     

Distinguishing implicit and explicit learning

I agree with Dienes and Berry’s (1997) and Neal and Hesketh’s (1997) call for investigations of the qualitative differences between implicit and explicit learning and note that such investigations must be guided by a workable definition of what is implicit and by theories that predict what the qualitative differences might be. Following Schacter, Bowers, and Booker’s (1989) retrieval intentionality criterion, I propose using anencoding intentionality criterion to distinguish implicit from explicit learning; we can reasonably infer that implicit learning has occurred when a variable known to influence explicit learning has no effect in a comparable implicit learning condition. I then suggest that implicit learning depends on noncognitive, nonhierarchical associations, whereas explicit learning depends on cognitive, hierarchical associations, and briefly describe an experiment that confirms a qualitative difference between implicit and explicit learning predicted by this hypothesis.     

The emergence of explicit knowledge from implicit learning

Substantial evidence has highlighted the ability of observers to incidentally extract statistical contingencies present in visual environments. This study examined whether the knowledge extracted regarding statistical contingencies is unconscious initially, even when it becomes fully accessible to conscious awareness after extensive training. Using a “typical” contextual cuing procedure adapted to real-world scenes, we first observed that, after extensive training in searching for a target within repeated scenes, knowledge about regularities was associated with conscious awareness (Experiment 1). However, both subjective and objective measures of consciousness revealed that in the early phase of training, learning of regular structures first takes place at an unconscious level (Experiments 2 and 3). These results are discussed in the light of the causal relationships between learning and consciousness.     

Effects of learning contexts on implicit and explicit learning

Two parallel tasks involving rule learning were identified in Experiment 1A and were used to assess implicit and explicit learning. In both tasks, subjects had to input numbers in order to reach the target values of outputs. The relationship between inputs and outputs was either simple (in the simple task) or complex (in the complex task), and the way in which target values were presented could be in the form of either numbers (in the simple task) or lines (in the complex task). Experiment 1B examined the validity of the explicit measure in the complex task. Experiments 2–4 investigated the interaction between learning contexts and the simple/complex learning tasks. Verbalization and instructions to search for the rules facilitated the simple-task learning and hurt or have no effect on the complex-task learning. In the observational-learning condition, no learning occurred for the simple task, and the complex task learning was impaired. These results suggest that the complex task and simple task involve two distinct learning systems. Other implications are also discussed.     

Development of implicit and explicit category learning

We present two studies that examined developmental differences in the implicit and explicit acquisition of category knowledge. College-attending adults consistently outperformed school-age children on two separate information-integration paradigms due to children’s more frequent use of an explicit rule-based strategy. Accuracy rates were also higher for adults on a unidimensional rule-based task due to children’s more frequent use of the irrelevant dimension to guide their behavior. Results across these two studies suggest that the ability to learn categorization structures may be dependent on a child’s ability to inhibit output from the explicit system.     

Psychometric intelligence dissociates implicit and explicit learning

The hypothesis that performance on implicit learning tasks is unrelated to psychometric intelligence was examined in a sample of 605 German pupils. Performance in artificial grammar learning, process control, and serial learning did not correlate with various measures of intelligence when participants were given standard implicit instructions. Under an explicit rule discovery instruction, however, a significant relationship between performance on the learning tasks and intelligence appeared. This finding provides support for Reber's hypothesis that implicit learning, in contrast to explicit learning, is independent of intelligence, and confirms thereby the distinction between the 2 modes of learning. However, because there were virtually no correlations among the 3 learning tasks, the assumption of a unitary ability of implicit learning was not supported.     

Adult age differences in implicit and explicit memory: time course and encoding effects

Young and older adults were tested at three delays on word-stem completion or cued recall following semantic or structural word judgments. Identical three-letter stems were present at retrieval for both implicit (completion) and explicit (cued recall) tasks; only the intention to recall list words differed. The young adults outperformed the older adults on both implicit and explicit tasks at all test delays. Under some conditions, the older but not the young adults performed more poorly on cued recall than on stem completion, suggesting a possible failure to use implicitly available information to support explicit remembering. These results suggest that some forms of implicit memory decline with normal aging.     

Differential time course of implicit and explicit cueing by colour and orientation in visual search

Using a cued conjunction search task, Anderson, Heinke, and Humphreys (2010) demonstrated larger effects from cueing target colour than from cueing target orientation. In this study, we separated the implicit (nonexpectation-dependent) and explicit (expectation-dependent) effects of orientation and colour visual cues. In Experiment 1, we replicated the original findings for short cue durations (100–200 ms), demonstrating that cues matching the physical property of the target on 80% of trials exert a rapid effect on search. These early cueing effects on reaction times were supported by evidence of guidance from cues on early eye movements. Experiment 2 introduced a feature to the cue that randomly matched the colour or orientation of the target. When cue orientation was predictive, there were strong implicit effects based on whether the colour of the cue and target matched. When cue colour was predictive, there were only weak effects from the cue's orientation. Implicit effects from cue colour remained when orientation-predictive cues were used and colour was unlikely to predict the target (Experiment 3). The data suggest that strong effects of colour cueing result from a combination of implicit and explicit processes, whereas effects from orientation cues are largely limited to the explicit guidance of visual search.     

Contextual interference in implicit and explicit motor learning

The purposes of this study were to examine the contextual interference effect in implicit learning and to compare implicit and explicit learning. Thirty-two participants performed a pursuit-tracking task for 60 acquisition and ten retention trials. The middle segment of target pathways had only two patterns whereas other segments had random patterns. A combination of awareness and practice order for the middle segment created four acquisition conditions: implicit-blocked, implicit-serial, explicit-blocked and explicit-serial. A questionnaire and a recognition test revealed that implicit groups were unaware of the repetition. Results showed no contextual interference effect in either implicit or explicit learning. Acquisition and retention performances were better for the middle segment than the other segments regardless of awareness. No difference between the implicit and explicit groups was found, suggesting that the implicit learning condition led to learning equivalent to the explicit learning condition.     

The combination of explicit and implicit learning processes in task control

Summary Two experiments look at the combination of explicit and implicit learning processes on a single task. Subjects are required to control the rate of sugar output in a small sugar production factory while maintaining cordial relations with the union. Experiment 1 investigates whether subjects can learn to control this task, which relies on their knowing about both salient and nonsalient relationships. It looks at how performance on the task relates to explicit verbalisable knowledge (as assessed by written questionnaire). Experiment 2 considers alternative ways of tapping the more implicit aspects of subjects' knowledge.     

The impact of response mode on implicit and explicit sequence learning

Evidence regarding the influence of response mode on sequence learning in serial reaction time (SRT) tasks has been mixed so far. In the present study, sequence learning was investigated under two different response conditions: manual (button presses) versus verbal (pronunciation of digits). Additionally, participants were divided post hoc into subgroups differing in their degree of explicit knowledge about the sequence. Results showed an interaction between response mode and type of learning (implicit vs. explicit), with explicit learning functioning more effectively under verbal than under manual conditions, whereas implicit learning was unaffected by the variation of the response mode. Implications concerning different underlying learning mechanisms (R-R learning vs. R-S learning) are discussed. Specifically, we suggest that the high response-effect distinctiveness of the verbal responses facilitated R-R learning.     

Qualitative differences between implicit and explicit sequence learning

Four experiments investigate the differences between implicit and explicit sequence learning concerning their resilience to structural and superficial task changes. A superficial change that embedded the SRT task in the context of a selection task, while maintaining the sequence, did selectively hinder the expression of implicit learning. In contrast, a manipulation that maintained the task surface, but decreased the sequence validity, affected the expression of learning specifically when it was explicit. These results are discussed in the context of a dynamic framework (Cleeremans & Jiménez, 2002), which assumes that implicit knowledge is specially affected by contextual factors and that, as knowledge becomes explicit, it allows for the development of relevant metaknowledge that modulates the expression of explicit knowledge.     

Comparison of implicit and explicit learning processes in a probabilistic task

This experiment compared the performance with explicit (rule-application and rule-discovery) and implicit (nonrule-instructed) learning approaches on the performance of a probabilistic video game task requiring fine motor control. The task required visual tracking of a small ball of light and "catching" it by means of joystick manipulation. A general pattern of improvement with practice occurred for all conditions. All conditions showed use of predictive relations among stimulus events. However, task performance of the rule-application and rule-discovery conditions were inferior to the nonrule-instructed implicit condition, particularly during the early phases of rule acquisition and application. This pattern strongly suggests substantial performance costs associated with attempting to discover or apply probabilistic rules. Decrements are likely due to increased cognitive demands associated with attempting to remember and strategically apply provided probability rules or attempting to discover and apply potentially important and useful probability information from a complex visual display.     

The projection of implicit and explicit goals

In 3 studies, the authors analyzed whether projection occurs for both conscious and nonconscious goals. In Experiment 1, participants who were predisposed to hold a learning goal over a performance goal rated others as possessing more of a learning goal. In Experiment 2, participants who were either implicitly primed with or explicitly assigned to have the goal to compete perceived others as striving for competitive goals more than control participants. In Experiment 3, the authors demonstrated that it was the actual goal to compete rather than the trait construct of competitiveness that was projected. The control of automatic goal projection effects is discussed, and interpersonal consequences of goal projection are delineated.     

Role of prior knowledge in implicit and explicit learning of artificial grammars

Artificial grammar learning (AGL) performance reflects both implicit and explicit processes and has typically been modeled without incorporating any influence from general world knowledge. Our research provides a systematic investigation of the implicit vs. explicit nature of general knowledge and its interaction with knowledge types investigated by past AGL research (i.e., rule- and similarity-based knowledge). In an AGL experiment, a general knowledge manipulation involved expectations being either congruent or incongruent with training stimulus structure. Inconsistent observations paradoxically led to an advantage in structural knowledge and in the use of general world knowledge in both explicit (conscious) and implicit (unconscious) cases (as assessed by subjective measures). The above findings were obtained under conditions of reduced processing time and impaired executive resources. Key findings from our work are that implicit AGL can clearly be affected by general knowledge, and implicit learning can be enhanced by the violation of expectations.     

The interaction of the explicit and the implicit in skill learning: a dual-process approach

This article explicates the interaction between implicit and explicit processes in skill learning, in contrast to the tendency of researchers to study each type in isolation. It highlights various effects of the interaction on learning (including synergy effects). The authors argue for an integrated model of skill learning that takes into account both implicit and explicit processes. Moreover, they argue for a bottom-up approach (first learning implicit knowledge and then explicit knowledge) in the integrated model. A variety of qualitative data can be accounted for by the approach. A computational model, CLARION, is then used to simulate a range of quantitative data. The results demonstrate the plausibility of the model, which provides a new perspective on skill learning.     

Effects of sleep loss, time of day, and extended mental work on implicit and explicit learning of sequences

Tacit knowledge is part of many professional skills and can be studied experimentally with implicit-learning paradigms. The authors explored the effects of 2 different stressors, loss of sleep and mental fatigue, on implicit learning in a serial-response time (RT) task. In the 1st experiment, 1 night of sleep deprivation was shown to impair implicit but not explicit sequence learning. In the 2nd experiment, no impairment of both types of sequence learning was found after 1.5 hr of mental work. Serial-RT performance, in contrast, suffered from both stressors. These findings suggest that sleep deprivation induces specific risks for automatic, skill-based behavior that are not present in consciously controlled performance.     

Eliciting false memories on implicit and explicit memory tests after incidental learning

Abstract: In a false memory experiment, lists of semantic associates (e.g., newspaper, letter, book, etc.) were presented to three groups of participants to induce false memories for critical nonpresented (CN) words (e.g., read) in an incidental learning task. The control group simply estimated the frequency rate in everyday Japanese discourse of each word on a list. The imagery instruction group received an additional instruction to imagine a thematically related converging word from the target words on a list. Participants in the imagery plus writing group received the same instructions as those in the imagery instruction group, but were also required to write down the word they imagined for each list. The results from the implicit and explicit memory tests given after the incidental learning episode showed that the level of priming for CN words was equivalent to that for actually presented target words for all three groups on the implicit test, whereas explicit memory results showed that participants explicitly recognized more target words than CN words. The implications for implicit associative response and fuzzy‐trace theories of false memory, as well as implicit priming, are discussed.     

Providing explicit information disrupts implicit motor learning after basal ganglia stroke

Despite their purported neuroanatomic and functional isolation, empirical evidence suggests that sometimes conscious explicit processes can influence implicit motor skill learning. Our goal was to determine if the provision of explicit information affected implicit motor-sequence learning after damage to the basal ganglia. Individuals with stroke affecting the basal ganglia (BG) and healthy controls (HC) practiced a continuous implicit motor-sequencing task; half were provided with explicit information (EI) and half were not (No-EI). The focus of brain damage for both BG groups was in the putamen. All of the EI participants were at least explicitly aware of the repeating sequence. Across three days of practice, explicit information had a differential effect on the groups. Explicit information disrupted acquisition performance in participants with basal ganglia stroke but not healthy controls. By retention (day 4), a dissociation was apparent--explicit information hindered implicit learning in participants with basal ganglia lesions but aided healthy controls. It appears that after basal ganglia stroke explicit information is less helpful in the development of the motor plan than is discovering a motor solution using the implicit system alone. This may be due to the increased demand placed on working memory by explicit information. Thus, basal ganglia integrity may be a crucial factor in determining the efficacy of explicit information for implicit motor-sequence learning.