Welcome to Memory Mondays, where I read a textbook on memory and talk about what I learned. If you like your cognitive psychology neatly summarized, with a healthy dose of unnecessary commentary and excessive amounts of semi-colons, this is the series for you!
Continuing from where I left off last time in Ch. 7, this week I will be discussing the concept of concepts, how concepts are stored in the brain, and where schema fit into all of this.
Concepts were traditionally considered to be abstract and stable representations. They are abstract because they are detached from sensory and motor processes, and stable because a given individual uses the same representation of the concept in different moments, and different people have similar representations.
The above definition has been challenged by Barsalou, who considers context to be king for the representations of concepts. He argues that the representation of a concept in a given moment depends on the situation and the individual’s goals. He demonstrates this by using an example of a bicycle. The idea of a bicycle that is activated depends on the situation. The tires will be activated if you have a flat, the basket if you’re trying to figure out if your groceries will fit, and the overall size of the bike if you’re winding your way through traffic (the latter two may be my examples).
There is some evidence to back up this idea that context influences concepts. When we think of the characteristics of an object, we sometimes visualize it, activating our perceptual systems, and this influences which characteristics we think of. For example, when participants in a study were asked to list characteristics of a watermelon, they gave properties like ‘rind’ and ‘green’, but provided properties like ‘pip’ and ‘red’ when asked about a half watermelon. We also think of characteristics related to a situation where the object might be found, rather than just the object itself. For example, when asked for the properties of ‘lawn’, participants thought of ‘picnic’ and ‘you can play on it’.
An experiment design similar to the Stroop effect shows how we activate gestural information related to an object. Participants learned to make movements associated with certain colours (e.g. poking for the colour red, making an open grasp for the colour blue). Then, they would respond to words of objects with the movement, based on the colour the word was written in. Sometimes the movement related to both the object and the colour (e.g. blue, open grasp, and pliers). Sometimes, it didn’t (e.g. red, poke, pliers). When it didn’t match with the object, participants took longer to respond, showing that gestural knowledge for the object was activated even when it wasn’t needed.
It makes sense that perceptual and motor systems are activated for concepts, because we use concepts in everyday life through our perceptual and motor systems. However, concepts are not entirely context-dependent. Both the traditional conception of concepts and Barsalou’s challenge are correct. Concepts have a stable, abstract core, and features that are dependent on the situation.
For teachers, I think this demonstrates the importance of concrete examples when teaching abstract concepts. We want students to have the abstract core of the concepts, to make the knowledge flexible and easier to apply in different situations; considering that these concepts are also inseparable from situational information, they can be made more accessible by including context. The abstract core can be developed by looking at variable examples.
Concepts in the Brain
Looking back at the previous posts on models for how concepts are organized in the brain, there were two main problems with the spreading activation model. First, the assumption that a concept has a single representation, which was tackled by Barsalou in his exploration of the influence of context. Second, the assumption that a concept is represented at a single node and held at a single location, which will be addressed by looking at how concepts are stored in the brain.
For a given concept, different kinds of information are stored in different areas of the brain. For example, visual information about a fan is in the visual part of the brain, while the auditory information of the whirring sound it makes is in the auditory area. Which makes sense.
However, as we saw in the previous section, concepts are not just perceptual and motor information. They have an abstract, stable core that allows us to identify the concept and detect similarities, even when surface features vary greatly (such as the difficult-to-define game).
The hub-and-spoke model shows how concepts can be stored in the brain, both the sensory and motor information and the abstract core. The spokes are modality-specific, which means they are stored based on the type of information, such as visual, verbal, or motor, in the area of the brain where that modality is processed. The hub is modality-independent, and helps us to unify the sensory and motor information about the concept. It is probably located in the anterior temporal lobes.
There is an interesting study involving transcranial magnetic stimulation (TMS) that shows abstract concept information is stored in this lobe. Remember that TMS uses a magnetic field to temporarily impair part of the brain; it is like studying the effects of a brain injury, without the brain injury. By studying how our functions and actions are impaired as a result, we better understand the role of that part of the brain.
In this study, one of two areas of the brain received TMS: the anterior temporal lobes, where the abstract aspect of concepts was theorized to be found, or the inferior parietal lobule, which is thought to control actions we make for objects (which would be a spoke in the hub-and-spoke model). Participants named objects from three categories: living things, manipulable objects, and non-manipulable human-made things. When the inferior parietal lobule was impaired, naming was slowed down for manipulable objects, but when the anterior temporal lobes were, naming was slowed down for all three categories. These results are predicted by the hub-and-spoke model.
Why is this important? Well, recall that neuroscience provides support for cognitive psychology theories. Where the neuroscience and cognitive psychology agree, we can be confident we are on the right track with our ideas. Mapping concepts onto their locations in the brain provides support for both their concrete and abstract nature.
Schema, Scripts, and Frames
Much of the knowledge we have about the world is not stored in individual, separate concepts, but in large structures or chunks of information called schema.
There are different types of schemas. Scripts deal with events, such as a restaurant meal. Frames are about some aspect of the world (e.g. vehicle), with fixed features (e.g. wheels and fuel) and space for variable information (e.g. type, paint colour).
Concepts and schemas are distinct types of knowledge, and this is supported by research on brain-damaged patients. Patients with semantic dementia, showing impaired knowledge of concepts, are still able to execute actions in accordance with scripts, such as retrieving objects for booking an appointment even though they could not say the function of the objects in a test beforehand. Patients with damage to the prefrontal cortex show difficulties in assembling and ordering the events of a script, even if they can still recall the events themselves and can catch errors in meaning in the event descriptions. The prefrontal cortex is associated with executive function, which include attention and goal-orientation, the latter of which is important for carrying out scripts. So, from these studies of brain-damaged patients, we can see concepts and schemas are distinct.
Schemas are useful in many ways. Because we have expectation about certain things or events, we are able to fill in gaps while reading or listening, and make inferences for information the writer or speaker is not making explicit.
This is also the case for visual information. In one study, participants were briefly presented pictures of objects (e.g. loaf of bread). Participants were better at identifying the object when a relevant scene (e.g. kitchen) was presented beforehand. With the appropriate schema activated, inferences can accurately be made using a small amount of visual information.
As well, schema help us avoid cognitive overload. Rather than processing all of the information of a thing or situation, we recall our expectations of the situation and just focus on what is novel or different.
This adds to my argument about the importance of knowledge. I already specifically highlighted the role of background knowledge in reading comprehension. The idea of cognitive overload when dealing with novel information also supports my point that students can’t ‘just look it up’. If they have to look up everything, then they are overloaded with novel information. Schemas help reduce processing load, by reducing the amount of novel information, but we need to have schemas in the first place for this to work.
However, schemas can create problems as well. Stereotypes are a form of schema, and while our minds use them to reduce detailed processing of each new person we meet, they can also result in harmful generalizations.
Schema can also result in errors in recall, which was first explored by Bartlett, researching how people recall stories from other cultures. To quote Eysenck (author of the chapter), who was paraphrasing Bartlett:
Suppose people read a story taken from a different culture. Their prior knowledge might produce distortions in the remembered version of the story, making it more conventional and acceptable from their own cultural background.
The longer the time interval between reading the story and being tested for recall, the greater the number of errors that change the story to better align with cultural expectations of narrative.
In a future Memory Mondays post, I’ll talk about forgetting, and how it helps us, until it doesn’t. Schemas are similar. We have them for good reasons, and they work, until they don’t.
We need to forgive our minds for their shortcomings, are they are pretty extraordinary for what they can do. And knowing the problems that can occur when we process information can help us recognize and compensate for errors when they do happen.
Which is a pretty good segue for the next chapter.
Next time: retrieval!