encoding

Context & the conditionalization of knowledge

Context is absolutely critical to successful communication. Think of the common experience of being a stranger at a family gathering or a meeting of close friends. Even familiar words and phrases may take on a different or additional meaning, among people who have a shared history. Many jokes and comments will be completely unintelligible, though you all speak the same language.

American anthropologist Edward Hall makes a useful distinction between ‘High context’ and ‘Low context’ communications. Your family gathering would be an example of a high context situation. In this setting, much of the meaning is carried in the speakers, their relationships, their knowledge of each other. In a low context situation, on the other hand, most of the meaning is carried in the actual words.

Part of the problem with email, as we all recognize, is that the context is so lacking, and the burden lies so heavily on the words themselves.

The importance of context for comprehension has, of course, profound implications for learning and memory.

I was reminded of this just the other day. I’m a fan of a TV program called NCIS. I only discovered it, however, at the beginning of the third season. After I’d watched it for some weeks, I purchased the DVDs of the earlier seasons. Most recently, I bought the DVD of season 3, which I had, of course, seen on TV. Watching the first episode of that season, which was the first episode of NCIS I ever saw, I was surprised to hear a line which I had no memory of, that was freighted with significance and led me to a much deeper understanding of the relationship between two of the characters — but which had meant absolutely nothing to me when I originally saw it, ignorant as I was of any of the characters and the back story.

The revelation meant nothing to me as a novice to the program, and so I didn’t remember it, but it meant everything to me as (dare I say it?) an expert.

Context is such a slippery word; so hard to define and pin down. But I think it’s fair to say that the difference between the novice and the expert rests on this concept. When an expert is confronted with a piece of information from her area of expertise, she knows what it means and where it belongs — even if the information is new to her. Because of this, she can acquire new information much more easily than a novice. But this advantage applies only in the expert’s area of expertise.

To take another example from the frivolous world of popular culture, a British study of fans of the long-running radio soap opera The Archers were given one of two imaginary scripts to read. One story was representative of the normal events in The Archers (a visit to a livestock market); the other was atypical (a visit to a boat show). These experts were able to remember many more details of the typical, market story than a group of subjects who knew little about the soap opera, but were no better at remembering details for the atypical story. Most importantly, this occurred even though the two stories shared many parallel features and most of the questions (and answers) used to assess their memory were the same. This indicates the specificity of expert knowledge.

Part of the advantage experts have is thought to rest on the ‘conditionalization’ of knowledge. That is, experts’ knowledge includes a specification of the contexts in which it is relevant.

It is surprising to many, this idea that it is not necessarily a lack of knowledge that is the problem — that people often have relevant knowledge and don’t apply it. In reading, for example, readers often don’t make inferences that they are perfectly capable of making, on the knowledge they have, unless the inferences are absolutely demanded to make sense of the text.

Another example comes from the making of analogies. I discuss this in my workbook on taking notes. Here’s a brief extract:

------------------------------------------

Rutherford’s comparison of the atom to the solar system gave us a means to understand the atom. The story goes that Newton ‘discovered’ gravity when an apple fell on his head — because of the comparison he made, realizing that the motion of an apple falling from a tree was in some sense like the motion of the planets. These are comparisons called analogies, and analogy has been shown to be a powerful tool for learning.

But the problem with analogies is that we have trouble coming up with them.

Generally, when we make analogies, we use an example we know well to help us understand something we don’t understand very well. This means that we need to retrieve from memory an appropriate example. But this is clearly a difficult task; people frequently fail to make appropriate connections — even, surprisingly, when an appropriate connection has recently come their way. In a study where people were given a problem to solve after reading a story in which an analogous problem was solved, 80% didn’t think of using the story to solve the problem until the analogy was pointed out to them.

It’s thought that retrieving an appropriate analogy is so difficult because of the way we file information in memory. Certainly similarity is an important attribute in our filed memories, but it’s not the same sort of similarity that governs analogies. The similarity that helps us retrieve memories is a surface similarity — a similarity of features and context. But analogies run on a deeper similarity — a similarity of structure, of relations between objects. This will only be encoded if you have multiple examples (at least more than one) and make an explicit effort to note such relations.

----------------------------------------------

The conditionalization of knowledge is of course related to the problem of transfer. Transfer refers to the ability to extend (transfer) learning from one situation to another (read more about it here) . Transfer is frequently used as a measure of successful learning. It’s all very well to know that 399-(399*0.1) = 359.1, but how far can you be said to understand it — how much use is it — if you can’t work out how much a $3.99 item will cost you if you have a 10% discount? (In fact, the asymmetry generally works the other way: many people are skilled at working out such purchase calculations, but fall apart when the problem is transferred to a purely numerical problem).

Transfer is affected by the context in which the information was originally acquired — obviously transfer is particularly problematic if you learn the material in a single context — and this is partly where the experts achieve their conditionalization: because, spending so much time with their subject they are more likely to come across the same information in a variety of contexts. But the more important source is probably the level of abstraction at which experts can operate (see my article on transfer for examples of how transfer is facilitated if the information is framed at a higher level of abstraction).

In those with existing expertise, an abstract framework is already in place. When an expert is confronted by new information, they automatically try and fit it into their existing framework. Whether it is consistent or inconsistent with what is already known doesn’t really matter — either way it will be more memorable than information that makes no deep or important connections to familiar material.

Let’s return to this idea of high and low context. Hall was talking about communications, in the context of different cultures (interestingly, he found cultures varied in the degree to which they were context-bound), but the basic concept is a useful one in other contexts. It is helpful to consider, when approaching a topic, either as student or teacher, the degree to which understanding requires implicit knowledge. A high context topic might be thought of as one that assumes a lot of prior knowledge, that assumes a knowledge of deeper structure, that is difficult to explain in words alone. A low context topic might be thought of as one that can be clearly and simply expressed, that can largely stand alone. Learning the basics of a language — how to conjugate a verb; some simple words and phrases — might be thought of as a low context topic, although clearly mastery of a language requires the complex and diverse building up of experiences that signifies a high context topic (and also clearly, some languages will be more ‘high context’ than others).

There is nothing particularly profound about this distinction, but an awareness of the ‘contextual degree’ of a topic or situation, is helpful for students, teachers, and anyone involved in trying to communicate with another human being (or indeed, computer!). It’s also helpful to be aware that high context situations require much more expertise than low context ones.

This article first appeared as "Context, communication & learning" in the Memory Key Newsletter for April 2007

References: 

Reeve, D.K. & Aggleton, J.P. 1998. On the specificity of expert knowledge about a soap opera: an everyday story of farming folk. Applied Cognitive Psychology, 12 (1), 35-42.

How your brain chunks ‘moments’ into ‘events’

image from study

We talk about memory for ‘events’, but how does the brain decide what an event is? How does it decide what is part of an event and what isn’t? A new study suggests that our brain uses categories it creates based on temporal relationships between people, objects, and actions — i.e., items that tend to—or tend not to—pop up near one another at specific times.

Reactivate if you want to remember

action replay box

We know sleep helps consolidate memories. Now a new study sheds light on how your sleeping brain decides what’s worth keeping. The study found that when the information that makes up a memory has a high value—associated with, for example, making more money—the memory is more likely to be rehearsed and consolidated during sleep.

It’s not the noise in the brain; it’s the noise in the input

noise on TV screen

A new study has found that errors in perceptual decisions occurred only when there was confused sensory input, not because of any ‘noise’ or randomness in the cognitive processing. The finding, if replicated across broader contexts, will change some of our fundamental assumptions about how the brain works.

Cognitive decline in old age related to poorer sleep

A new study confirms the role slow-wave sleep plays in consolidating memories, and reveals that one reason for older adults’ memory problems may be the quality of their sleep.

Recent research has suggested that sleep problems might be a risk factor in developing Alzheimer’s, and in mild cognitive impairment. A new study adds to this gathering evidence by connecting reduced slow-wave sleep in older adults to brain atrophy and poorer learning.

The study involved 18 healthy young adults (mostly in their 20s) and 15 healthy older adults (mostly in their 70s). Participants learned 120 word- nonsense word pairs and were tested for recognition before going to bed. Their brain activity was recorded while they slept. Brain activity was also measured in the morning, when they were tested again on the word pairs.

As has been found previously, older adults showed markedly less slow-wave activity (both over the whole brain and specifically in the prefrontal cortex) than the younger adults. Again, as in previous studies, the biggest difference between young and older adults in terms of gray matter volume was found in the medial prefrontal cortex (mPFC). Moreover, significant differences were also found in the insula and posterior cingulate cortex. These regions, like the mPFC, have also been associated with the generation of slow waves.

When mPFC volume was taken into account, age no longer significantly predicted the extent of the decline in slow-wave activity — in other words, the decline in slow-wave activity appears to be due to the brain atrophy in the medial prefrontal cortex. Atrophy in other regions of the brain (precuneus, hippocampus, temporal lobe) was not associated with the decline in slow-wave activity when age was considered.

Older adults did significantly worse on the delayed recognition test than young adults. Performance on the immediate test did not predict performance on the delayed test. Moreover, the highest performers on the immediate test among the older adults performed at the same level as the lowest young adult performers — nevertheless, these older adults did worse the following day.

Slow-wave activity during sleep was significantly associated with performance on the next day’s test. Moreover, when slow-wave activity was taken into account, neither age nor mPFC atrophy significantly predicted test performance.

In other words, age relates to shrinkage of the prefrontal cortex, this shrinkage relates to a decline in slow-wave activity during sleep, and this decline in slow-wave sleep relates to poorer cognitive performance.

The findings confirm the importance of slow-wave brainwaves for memory consolidation.

All of this suggests that poorer sleep quality contributes significantly to age-related cognitive decline, and that efforts should be made to improve quality of sleep rather than just assuming lighter, more disturbed sleep is ‘natural’ in old age!

Why learning gets harder as we get older

A mouse study shows that weakening unwanted or out-of-date connections is as important as making new connections, and that neurological changes as we age reduces our ability to weaken old connections.

A new study adds more support to the idea that the increasing difficulty in learning new information and skills that most of us experience as we age is not down to any difficulty in acquiring new information, but rests on the interference from all the old information.

Memory is about strengthening some connections and weakening others. A vital player in this process of synaptic plasticity is the NMDA receptor in the hippocampus. This glutamate receptor has two subunits (NR2A and NR2B), whose ratio changes as the brain develops. Children have higher ratios of NR2B, which lengthens the time neurons talk to each other, enabling them to make stronger connections, thus optimizing learning. After puberty, the ratio shifts, so there is more NR2A.

Of course, there are many other changes in the aging brain, so it’s been difficult to disentangle the effects of this changing ratio from other changes. This new study genetically modified mice to have more NR2A and less NR2B (reflecting the ratio typical of older humans), thus avoiding the other confounds.

To the researchers’ surprise, the mice were found to be still good at making strong connections (‘long-term potentiation’ - LTP), but instead had an impaired ability to weaken existing connections (‘long-term depression’ - LTD). This produces too much noise (bear in mind that each neuron averages 3,000 potential points of contact (i.e., synapses), and you will see the importance of turning down the noise!).

Interestingly, LTD responses were only abolished within a particular frequency range (3-5 Hz), and didn’t affect 1Hz-induced LTD (or 100Hz-induced LTP). Moreover, while the mice showed impaired long-term learning, their short-term memory was unaffected. The researchers suggest that these particular LTD responses are critical for ‘post-learning information sculpting’, which they suggest is a step (hitherto unknown) in the consolidation process. This step, they postulate, involves modifying the new information to fit in with existing networks of knowledge.

Previous work by these researchers has found that mice genetically modified to have an excess of NR2B became ‘super-learners’. Until now, the emphasis in learning and memory has always been on long-term potentiation, and the role (if any) of long-term depression has been much less clear. These results point to the importance of both these processes in sculpting learning and memory.

The findings also seem to fit in with the idea that a major cause of age-related cognitive decline is the failure to inhibit unwanted information, and confirm the importance of keeping your mind actively engaged and learning, because this ratio is also affected by experience.

Meditation can produce enduring changes in emotional processing

A new study provides more evidence that meditation changes the brain, and different types of meditation produce different effects.

More evidence that even an 8-week meditation training program can have measurable effects on the brain comes from an imaging study. Moreover, the type of meditation makes a difference to how the brain changes.

The study involved 36 participants from three different 8-week courses: mindful meditation, compassion meditation, and health education (control group). The courses involved only two hours class time each week, with meditation students encouraged to meditate for an average 20 minutes a day outside class. There was a great deal of individual variability in the total amount of meditation done by the end of the course (210-1491 minutes for the mindful attention training course; 190-905 minutes for the compassion training course).

Participants’ brains were scanned three weeks before the courses began, and three weeks after the end. During each brain scan, the volunteers viewed 108 images of people in situations that were either emotionally positive, negative or neutral.

In the mindful attention group, the second brain scan showed a decrease in activation in the right amygdala in response to all images, supporting the idea that meditation can improve emotional stability and response to stress. In the compassion meditation group, right amygdala activity also decreased in response to positive or neutral images, but, among those who reported practicing compassion meditation most frequently, right amygdala activity tended to increase in response to negative images. No significant changes were seen in the control group or in the left amygdala of any participant.

The findings support the idea that meditation can be effective in improving emotional control, and that compassion meditation can indeed increase compassionate feelings. Increased amygdala activation was also correlated with decreased depression scores in the compassion meditation group, which suggests that having more compassion towards others may also be beneficial for oneself.

The findings also support the idea that the changes brought about by meditation endure beyond the meditative state, and that the changes can start to occur quite quickly.

These findings are all consistent with other recent research.

One point is worth emphasizing, in the light of the difficulty in developing a training program that improves working memory rather than simply improving the task being practiced. These findings suggest that, unlike most cognitive training programs, meditation training might produce learning that is process-specific rather than stimulus- or task-specific, giving it perhaps a wider generality than most cognitive training.

Dopamine decline underlies episodic memory decline in old age

Findings supporting dopamine’s role in long-term episodic memory point to a decline in dopamine levels as part of the reason for cognitive decline in old age, and perhaps in Alzheimer’s.

The neurotransmitter dopamine is found throughout the brain and has been implicated in a number of cognitive processes, including memory. It is well-known, of course, that Parkinson's disease is characterized by low levels of dopamine, and is treated by raising dopamine levels.

A new study of older adults has now demonstrated the effect of dopamine on episodic memory. In the study, participants (aged 65-75) were shown black and white photos of indoor scenes and landscapes. The subsequent recognition test presented them with these photos mixed in with new ones, and required them to note which photos they had seen before. Half of the participants were first given Levodopa (‘L-dopa’), and half a placebo.

Recognition tests were given two and six hours after being shown the photos. There was no difference between the groups at the two-hour test, but at the six-hour test, those given L-dopa recognized up to 20% more photos than controls.

The failure to find a difference at the two-hour test was expected, if dopamine’s role is to help strengthen the memory code for long-term storage, which occurs after 4-6 hours.

Individual differences indicated that the ratio between the amount of Levodopa taken and body weight is key for an optimally effective dose.

The findings therefore suggest that at least part of the reason for the decline in episodic memory typically seen in older adults is caused by declining levels of dopamine.

Given that episodic memory is one of the first and greatest types of memory hit by Alzheimer’s, this finding also has implications for Alzheimer’s treatment.

Caffeine improves recognition of positive words

Another recent study also demonstrates, rather more obliquely, the benefits of dopamine. In this study, 200 mg of caffeine (equivalent to 2-3 cups of coffee), taken 30 minutes earlier by healthy young adults, was found to improve recognition of positive words, but had no effect on the processing of emotionally neutral or negative words. Positive words are consistently processed faster and more accurately than negative and neutral words.

Because caffeine is linked to an increase in dopamine transmission (an indirect effect, stemming from caffeine’s inhibitory effect on adenosine receptors), the researchers suggest that this effect of caffeine on positive words demonstrates that the processing advantage enjoyed by positive words is driven by the involvement of the dopaminergic system.

The value of intensive practice

Changing your brain - which is what happens when you learn, and when yoe encode new experiences, or have new thoughts - is crucial for keeping your brain 'young'.

Learning a new language may be especially beneficial for keeping your brain flexible and thus fighting age-related cognitive decline.

Intense periods of learning may be especially beneficial.

Let’s talk about the cognitive benefits of learning and using another language.

In a recent news report, I talked about the finding that intensive learning of a very novel language significantly grew several brain regions, of which two were positively associated with language proficiency. These regions were the right hippocampus and the left superior temporal gyrus. Growth of the first of these probably reflects the learning of a great many new words, and the second may reflect heavy use of the phonological loop (a part of working memory).

Growing the brain with a new language

A new study adds to the growing evidence for the cognitive benefits of learning a new language, and hints at why some people might be better at this than others.

A small Swedish brain imaging study adds to the evidence for the cognitive benefits of learning a new language by investigating the brain changes in students undergoing a highly intensive language course.

The study involved an unusual group: conscripts in the Swedish Armed Forces Interpreter Academy. These young people, selected for their talent for languages, undergo an intensive course to allow them to learn a completely novel language (Egyptian Arabic, Russian or Dari) fluently within ten months. This requires them to acquire new vocabulary at a rate of 300-500 words every week.

Brain scans were taken of 14 right-handed volunteers from this group (6 women; 8 men), and 17 controls that were matched for age, years of education, intelligence, and emotional stability. The controls were medical and cognitive science students. The scans were taken before the start of the course/semester, and three months later.

The brain scans revealed that the language students showed significantly greater changes in several specific regions. These regions included three areas in the left hemisphere: the dorsal middle frontal gyrus, the inferior frontal gyrus, and the superior temporal gyrus. These regions all grew significantly. There was also some, more selective and smaller, growth in the middle frontal gyrus and inferior frontal gyrus in the right hemisphere. The hippocampus also grew significantly more for the interpreters compared to the controls, and this effect was greater in the right hippocampus.

Among the interpreters, language proficiency was related to increases in the right hippocampus and left superior temporal gyrus. Increases in the left middle frontal gyrus were related to teacher ratings of effort — those who put in the greatest effort (regardless of result) showed the greatest increase in this area.

In other words, both learning, and the effort put into learning, had different effects on brain development.

The main point, however, is that language learning in particular is having this effect. Bear in mind that the medical and cognitive science students are also presumably putting in similar levels of effort into their studies, and yet no such significant brain growth was observed.

Of course, there is no denying that the level of intensity with which the interpreters are acquiring a new language is extremely unusual, and it cannot be ruled out that it is this intensity, rather than the particular subject matter, that is crucial for this brain growth.

Neither can it be ruled out that the differences between the groups are rooted in the individuals selected for the interpreter group. The young people chosen for the intensive training at the interpreter academy were chosen on the basis of their talent for languages. Although brain scans showed no differences between the groups at baseline, we cannot rule out the possibility that such intensive training only benefited them because they possessed this potential for growth.

A final caveat is that the soldiers all underwent basic military training before beginning the course — three months of intense physical exercise. Physical exercise is, of course, usually very beneficial for the brain.

Nevertheless, we must give due weight to the fact that the brain scans of the two groups were comparable at baseline, and the changes discussed occurred specifically during this three-month learning period. Moreover, there is growing evidence that learning a new language is indeed ‘special’, if only because it involves such a complex network of processes and brain regions.

Given that people vary in their ‘talent’ for foreign language learning, and that learning a new language does tend to become harder as we get older, it is worth noting the link between growth of the hippocampus and superior temporal gyrus and language proficiency. The STG is involved in acoustic-phonetic processes, while the hippocampus is presumably vital for the encoding of new words into long-term memory.

Interestingly, previous research with children has suggested that the ability to learn new words is greatly affected by working memory span — specifically, by how much information they can hold in that part of working memory called phonological short-term memory. While this is less important for adults learning another language, it remains important for one particular category of new words: words that have no ready association to known words. Given the languages being studied by these Swedish interpreters, it seems likely that much if not all of their new vocabulary would fall into this category.

I wonder if the link with STG is more significant in this study, because the languages are so different from the students’ native language? I also wonder if, and to what extent, you might be able to improve your phonological short-term memory with this sort of intensive practice.

In this regard, it’s worth noting that a previous study found that language proficiency correlated with growth in the left inferior frontal gyrus in a group of English-speaking exchange students learning German in Switzerland. Is this difference because the training was less intensive? because the students had prior knowledge of German? because German and English are closely related in vocabulary? (I’m picking the last.)

The researchers point out that hippocampal plasticity might also be a critical factor in determining an individual’s facility for learning a new language. Such plasticity does, of course, tend to erode with age — but this can be largely counteracted if you keep your hippocampus limber (as it were).

All these are interesting speculations, but the main point is clear: the findings add to the growing evidence that bilingualism and foreign language learning have particular benefits for the brain, and for protecting against cognitive decline.

  • 1