learning is best construed as the capacity to learn intricate information in an
incidental manner, in the absence of being aware of what has been learned (Sun,
2008). It is when the learning proceeds in an unconscious manner and when the
knowledge we have acquired is difficult to express. The definitions of implicit
and explicit learning usually focus on the absence or presence of conscious
operations as an integral differentiating factor. Whittlesea and Dorken have
described implicit learning as just everyday learning without one being aware
of the effects of that learning. Implicit learning thus juxtaposes strongly
with explicit learning which is usually hypothesis driven.
learn something without being aware that this knowledge is being acquired or
processed? This question has been under examination for about thirty years now
and has had some significant developments (Frensch and Runger, 2003). A renewed
attentiveness towards learning and memory has attracted widespread attention to
a continuous debate. The debate focuses on the part consciousness plays in
cognition. This essay aims to review a number old and recent studies conducted
to better understand the concept of implicit learning. The essay will explore
what implicit learning is, how it differs from explicit learning and the role
it plays in shaping our thoughts and behaviors.
It is quite
clear that there is a strong relationship between knowledge and awareness.
Usually, a learning situation involves acquiring a novel piece of information which
can manifest itself at a later stage through an individual’s behavior (Shanks,
2017). In most cases, an individual is aware of this new or changed behavior,
and it can be expressed in words. Sometimes however, the knowledge we have
acquired is held implicitly and is not available to our conscious. Such
implicit knowledge cannot be assessed by simply asking the participant what he
or she has learned, it must be assessed indirectly via a task or some other
means (Shanks, 2017)
our daily life seems full of instances where we know more than we can explain,
it has proven particularly challenging to provide a suitable definition for
work has described implicit learning as a means by which theoretical knowledge can
be assimilated in an incidental manner through exposure to relevant situations.
The learning takes place unconsciously and automatically. Our cognitive system
thereby gets bestowed with a ‘smart’ unconscious (Loftus and Klinger, 1992). Predictably,
such a radical description of the phenomenon created substantial disagreement.
Several studies have been carried out leading to the emergence of some new
implicit learning usually includes stimuli that exist beyond the threshold of
consciousness (“Supraliminal”, n.d.) and tasks which need attentiveness to the operational
associations “between stimulus items as opposed to
the specific stimuli” (Cheesman & Merikle, 1984).
Most studies have adopted the reasoning that to express implicit learning, it
is enough to validate that performance on a task surpasses the subject’s awareness
of the learned information. The
assessment situation typically includes three components: 1) exposure to an
intricate setting under learning circumstances that are incidental in nature;
2) a gage of how well the participants can deliberate the information they have
just acquired through their performance on a similar or slightly altered task;
and 3) a gage of how aware the participants are of the knowledge they have
learned (Cleeremans, Destrebecqz & Boyer, 1998). Three studies that have
followed this experimental design have investigated “artificial grammar learning, sequence learning, and dynamic system
Reber carried out an experiment to investigate the role implicit learning
played in an artificial grammar learning (AGL) (Reber, 1967). The first phase was
the memorization phase. In this phase, the participants were asked to memorise
a set of letter strings created by a “finite-state
grammar” (Reber, 1967). In the next phase,
the participants were told that the strings were representative of the rubrics of
grammar and were asked to categorize a new set of strings as grammatically
correct or incorrect. The participants performed better than one would have
expected even though they failed to explain the rules of grammar in verbal
reports (Reber, 1993) (Reber, 1989). The disconnection between the
participants’ performance and verbal report lead Reber to describing the
learning as implicit. However, perhaps his method of measuring awareness could
have been better. He could have asked the participants more specific questions
to gage their understanding.
sequential learning task, the subjects were asked to respond to every component
of sequentially and visually arranged events in a “choice reaction task” (Niseen & Bullemer, 1987). On
each trial, the subjects would see stimuli emerge at one out of many locations
on a screen and were told to press the corresponding key as quickly and
correctly as possible. The subjects did not know that the orders of the
continuous stimuli followed a pattern that repeated itself or were directed by
a set of rubrics that described permitted transitions among consecutive stimuli
(Cleeremans, 1993). The experiment was testing whether the reaction times
differed between structured stimuli and unstructured stimuli. The results
showed that subjects produced faster reaction times when exposed to structured
stimuli even though they could not express their knowledge of the patterns
verbally. This suggests that subjects responded quicker when they had some
knowledge of the pattern, implying that some learning took place implicitly.
Dynamic System Control tasks, the subjects need to learn how to regulate the
computer simulation of a collaborative system like a coffee production factory
(Berry & Broadbent, 1984). The subjects are told some semantics regarding
the output variables such as the quantity of coffee output the factory manufactures.
Their goal was to attain a certain level of production by controlling the
inputs such as labor. The state of the system is “computed by way of an equation” that relates the inputs and outputs. The results of this experimental
task showed that even though the subjects could not display a verbal
understanding of the rules of the system, they achieved a high level of control
of the system.
the most influential implicit learning studies is the Iowa Gambling Task. It is
a psychological task that has been used to study decision making in many
clinical and developmental samples thought to stimulate real-life decision
making (Toplak et al., 2010). It was introduced by Antoine Bechara, Antonio Damasio,
Hanna Damasio and Steven Anderson. The participants in the task were presented
with four decks of cards each of which offered different payoffs. The goal for
the participants was to win as much money as they could by choosing the decks
that would reward them the most money. Bechara and his colleagues found that
after 40-50 trials, the participants seemed to be able to identify the good
decks and bad decks (Bechara, Damasio, Damasion & Anderson, 1994). Bechara
and his colleagues assessed the participants’ awareness of the task structure
and their choice behaviour. After the first 20 trials and then after every 10
additional trials, the participants were asked to describe what they knew about
the task and how they were making their choices (Shanks, 2017). Most of the
participants managed to get to a ‘conceptual’ period in which they had learned
enough to confidently label the good and bad decks and in this period.
So far, all
the researchers have assumed that if the knowledge is conscious, the
participants will be able to verbally report it (Shanks, 2017). However, Newell
and Shanks have pointed out a flaw in this assumption (2014). How can we be
certain that the questions asked were strong enough to press the participant
for the information the researchers were looking for?
to such concerns about Bechara’s method of assessing the participants’ awareness,
Konstantinidis and Shanks (2014) employed a more cautious approach to assessing
the same. Instead of just recording the participants’ responses to open-ended
questions about their thoughts and feelings towards the task, Konstantinidis
and Shanks asked the participants to rate each deck on a mathematical scale, rationalize
their ratings, report what they thought the average winnings or losses would be
if 10 cards were selected from each deck, and to state which deck they would
choose if they could only select from one deck for the remainder of the game
(Shanks and Konstantinidis, 2014). The participants’ answers to such specific
questions gave Shanks and Konstantinidis a clear indication of how aware the
participants were of the test structure and how much they had managed to
implicitly learn before making their choices. Their revised tests of verbal
reports gave them more reliable results than measuring the participants’
awareness based on their overt behavior.
studies paved the way for a discussion suggesting that verification of implicit
knowledge might not appear from studies relying merely on verbal reports.
Bayley and his colleagues presented a remarkable illustration. Their
participants were controls and 2 heavy amnesiacs, EP and GP (Bayley, Frascino
& Squire, 2005). The participants were shown 8 simultaneous object discriminations in which
a pair of objects was shown (e.g., A/B) and the participant had to choose the
‘correct’ one based on feedback provided to them. The pairs were presented
repeatedly in a random order. The controls managed to grasp this task pretty
quickly whereas the participants suffering from amnesia took more time to learn
and had very little recollection of the task. During the rest, 16 objects were
shown and the participants had to sort them into correct and incorrect objects.
The controls managed this without any problem while the amnesiacs struggled
with it. Thus, we can gather that the controls managed to pick up the task
pretty quickly and succeed at it whereas the participants suffering from
amnesia took a lot more time to learn the task and seemed to have very little
recollection of it. The learning demonstrated by the participants suffering
from amnesia was inflexible and unconscious whereas that in the controls was
declarative and flexible.
experiments and studies among countless others have made it clear to us that
implicit learning does take place in our everyday lives and plays an important
role in governing our thoughts and behaviours. Findings from the artificial
grammar learning task show that implicit learning helps subjects demonstrate a
performance that is better than what chance would predict when asked to make
judgements in grammar. The knowledge implicitly learned through this task can
also manifest itself in other slightly similar or more varied tasks without us
being aware of it. Likewise, participants’ performance in the sequential
learning task suggest the attainment of information on the rubrics used to create
the stimulus material (Lewicki, Czyzewska & Hoffman, 1987). These findings
show that while we might know what form our newly acquired knowledge takes, we
are certain that it is representative of the structure of the stimuli and the
relationships between them. Thus, we can conclude from these two examples and
all other implicit learning situations, that we acquire a novel piece of
knowledge in a form we might not know. We might not even be able to describe
what we have learned and how we learned it. However, this newly acquired piece
of knowledge will manifest itself at a later stage through our behaviour.
many researchers have realised some strong associations between implicit learning
and language acquisition and many of them have investigated this association
empirically. Saffran et al showed us how exposure to auditory material in an
incidental manner that is artificial in nature was adequate to facilitate
children and adults to divide the uninterrupted sequence of sounds they had
listened to into the artificial words that it comprised (Saffran, 1997). The
connection is not so surprising once one realises that implicit leaning and
language acquisition involve learning conditions of complicated information
that are incidental and unconscious in nature (Berry & Dienes, 1993).
or implicit learning aids us in acquiring certain skills, habits, emotional
responses etc. Language acquisition is just one example of that. Another
example is our inter-personal skills. In most cases we don’t explicitly learn
how to interact with people. We just pick up these skills through continued
exposure to relevant learning situations. Explicit learning on the other hand
aids in our learning of semantics like facts and events. It involves us making
an effort to seek out the information we are exposed to.
majority of research on the topic has paved the way for an agreement on quite a
few features that differentiate implicit from explicit learning. Implicit
learning is a passive process through which we acquire knowledge about the structure
of a complex stimulus environment by a natural and simple process (Bruneau,
n.d.). It is characterized by the absence of consciously accessible knowledge.
In contrast to this, explicit learning is an dynamic process. It involves
people consciously seeking out the assembly of a piece of information they are
Berry have summarized this consensus into three crucial differentiating points.
First, Implicit learning shows specificity of transfer. This means that
knowledge acquired implicitly tends to be reasonably inaccessible, inflexible
and bound to the surface characters of the material (Dienes & Berry, 1997).
Second, the learning situations in which we acquire implicit knowledge tend to
be incidental rather than intentional. Lastly, Implicit knowledge tends to stay
strong even over time.
Most of the
research leads us to conclude that implicit learning does not assume the presence
of some enigmatic processes of flaccid, automatic and unconscious attainment of
conceptual information. It is simply a result on continuous processing and the
awareness accompanies the learning. As Axel Cleeremans, Arnaud Destrebecqs and
Maud Boyer propose, implicit learning is best interpreted as a complicated form
of priming that takes place in continuously learning neural systems, and that
the distributional knowledge so acquired can be causally efficacious in the
absence of awareness that this knowledge was acquired and that it is currently
influencing processing, that is, in the absence of metaknowledge.