Recognition Memory Requires Less Brain Activity Than Recall - But Don't Tell Your Instructor
Cabeza, R., Kapur, S., Craik, F.I.M., McIntosh, A.R., Houle, S., and Tulving, E. (1997) Functional neuroanatomy of recall and recognition: A PET study of episodic memory. Journal of Cognitive Neuroscience, 9, 254-265.
Studying Recognition Memory and Recall
Have you ever read a question on a multiple choice test and thought "Did I learn anything about this?" but still were able to identify the correct answer from among the options? Have you ever left an essay exam wishing you had taken a multiple choice test instead? Most of us agree that multiple choice tests are easier than essays.
Multiple choice, matching, and true-false questions require you to recognize the correct answer. Essay, fill-in-the-blank, and short answer questions require you to recall the information. Recognition and recall are generally considered to be different psychological processes but they also share a lot in common. Cognitive psychologists have studied the differences and similarities in recall versus recognition by presenting different types of learning tasks and examining the effects on recognition and recall performance.
Cognitive neuroscientists have begun to examine recall and recognition memory as well. Cognitive neuroscientists merge interests in how individuals think, learn, and remember with special techniques for studying the how brain and nervous system work. Two common methods used by cognitive neuroscientists are lesion studies and neuroimaging studies.
Lesion studies involve investigating cognitive processes in individuals with damage to different parts of their brain. In humans, damage can be caused by head injuries, brain tumors, stokes, and other types of trauma. People with memory loss, amnesiacs, often have damage to their medial temporal lobes, to deeper structures like the thalamus and hippocampus, and sometimes to the frontal lobes.
Cabeza, Kapur, Craik, McIntosh, Houle, and Tulving studied recognition memory and recall in people without brain damage using neuroimaging. Although studies of people with brain damage can provide a lot of information about brain function, it is limited because many people with brain damage tend to have a large number of cognitive problems. Neuroimaging can not only be used with normal people but it can be used to identify brain activity. These images of brain activity can then be compared with cognitive processes as they occur over time to obtain a better picture of brain-behavior relationships.
Most of us agree that recognition tests are easier than recall tests. Why? In recognition tests, the information to be recalled is present, along with attractive alternatives. The task of recognition is to select the correct answer from among the alternatives. In recall, on the other hand, no possible answers are available (unless you cheat) but have to be generated. Unless you have absolutely no memory for the information, recognition is "easier" than recall.
Does the brain find recognition easier? What parts of the brain are active when recognizing information? What same and different parts of the brain are active while recalling information? Cabeza, Kapur, Craik, McIntosh, Houle, and Tulving used PET to address these questions.
Healthy young university students performed recognition and recall tasks while lying in the PET scan machine. In the study phase, the participant was shown word pairs, such as parents-piano. In the recognition condition, the participant was shown either the correct word pair again, such as parents-piano, or a pair in which the second word was incorrect, such as parents-wall. The participant was instructed to say the second word if he or she thought it was the correct one or else say "pass." In the recall condition, the participant was shown the first word paired with word, such as parents-word?, and was instructed to say the second word out loud if it was remembered or else say "pass."
The PET scans identified all brain activity occurring during these tasks. Brain activity during the recognition tasks would include that related to recognition but also activity related to reading, speaking, and generally attending to the task. To eliminate activity not related to the specific type of memory they were studying, Cabeza et al. had participants also read word pairs while being scanned. They subtracted brain activity resulting from this task from each of the memory tasks to obtain measures of brain activity related to specific recognition and recall processes.
The right prefrontal cortex was active for both recognition and recall. This part of the brain is important for attempting to recover information in memory and is activated in all memory attempts, regardless of the success of the attempt. Similarly the anterior cingulate area was active for both memory tasks. The anterior cingulate cortex has been implicated in the initiation of many types of human behavior.
One brain area was more active during recognition than recall. The right inferior parietal cortex, which is important for perceptual processing was activated during recognition. Cabeza et al. argued that this area should be active if recognition involves comparing the presented word, such as piano, with perceptual information from the study phase, such as whether the physical stimulus, piano, was seen during the study phase.
Four brain areas were more active during recall than during recognition. Although the anterior cingulate area was activated in both memory tasks, it was more active during recall. Cabeza et al. speculated that this is because recall involves more initiation of activity to generate a response.
Three additional brain areas were more active during recall than recognition. These were the left cerebellum, the right thalamus area, and the right globus palladius. These three brain areas, along with the right prefrontal cortex, form a cognitive cerebello-thalamo-cortical pathway that has recently been traced out in the brain. Cabeza et al. conclude that the pathway that goes from the left cerebellum to the right frontal cortex is important for recall of episodic memory, such as the words in their paired associate task.
Cabeza et al. identified more brain areas that were important for recognition and recall than have been found with lesion studies. They concluded that lesion studies and neuroimaging studies complement each other and need to be integrated in cognitive neuroscience investigations of memory. In addition, they concluded that, since fewer brain areas are activated by recognition than recall, that recognition is indeed easier than recall.
Links to the Lecture
A Quick Experiment
Replicate the recognition versus recall experiment. Generate 2 lists of 20 word pairs. To demonstrate recognition, project the one list of word pairs and allow students 1 minute to study them. After the study phase, project a list of word pairs with 10 from the studied list and 10 with a different second word. Have students record which word pairs were correct. To demonstrate recall, project the other word pair list and allow students to study them for 1 minute. After study, project a list of the first words in the pairs. Have students write as many of the second words as they can. Compare recognition scores with recall scores and discuss the different demands of the two tasks.
About the Authors
Roberto Cabeza is at the University of Alberta, Shitij Kapur, Fergus Craik, Anthony McIntosh, and Endel Tulving are from the Rotman Research Institute of Baycrest Centre, University of Toronto, and Sylvain Houle is at the PET Centre, Clark Institute of Psychiatry, University of Toronto. Dr. Cabeza has begun putting color pictures of his brain images on his homepage. Check out some PET scans from recent publications.
About the Journal
The Journal of Cognitive Neuroscience is a premier journal in cognitive neuroscience. Check out some of the abstracts and sample reports.
Links to Life
Positron Emission Tomography (PET) scans allow us to see brain activity during cognition. PET is very complicated but this description from the PET Centre of the Clark Institute of Psychiatry, where the research was conducted, and this description that also compares different types of neuroimaging are very clear and concise.
The Virtual Hospital is an electronic multimedia library from the Department of Radiology at the University of Iowa College of Medicine. It includes a fascinating textbook on brain anatomy. A photograph of a portion of the brain is displayed; underneath the photo is a line drawing that is labeled with the brain structures that are visible. This photo shows the lateral (outside) surface of the right cerebral cortex (with the cerebellum tucked up underneath). This photo shows the medial (inside) surface of the right cortex.
Here are the same brain structures shown in magnetic resonance images (MRI) from The Whole Brain Atlas. Click on a name to view a brain structure. The globus pallidus is shown in the image with the thalamus. This site also has some interesting MRIs of diseased brains.
How much do you know about brain structures? Here is a sagital (horizontal) slice through a brain scan created by PET. Click on a part of the slice to get the label for the brain structure. When you are really confident, try this test of brain structures (HINT: Click on one of the labeled options in the table at the bottom of the page).
Recognition is easier than recall. Multiple-choice tests are generally easier than fill-in-the-blanks tests or essays because it is easier to recognize the correct answer out of a group of possibilities than it is to have to dredge up the answer out of one’s own head.
Still, in order to be able to recognize the correct multiple-choice answer it has to be “somewhere” in one’s brain; otherwise there’s nothing to recognize. Someone with zero knowledge of a topic does no better than random chance on a multiple-choice test because all of the answer choices are equally meaningless to him. And someone with mastery of a topic can fill-in-the-blanks or can write an essay.
Think of your brain like a file cabinet, with tons of information stored in it. When you recognize a piece of information, it’s like the tab on a file folder in your head; the whole file folder now gets pulled up. By writing down anything you know about a problem, getting started in any possible way, you are hopefully going to write something that you then recognize, and your brain is going to pull the tab and bring up the rest of the folder.
Your brain contains over four terabytes of information (which is way too big a number to imagine), yet your working memory, the part of your brain that consciously works on a problem, can only hold about seven bits at any time. It’s as if your brain is a library, full of knowledge, yet you’re restricted to using a table only as big as a postage stamp.
Think about how impossible it is to multiply big numbers in your head, but how easy it is on paper. Your brain knows how to multiply, but it can’t keep track of all those digits.
This is why writing was invented in the first place. People found themselves with way more knowledge than they could hold and work with in their heads, and so they invented a way to put information “out there;” they scratched it in the dirt or into clay tablets or they inked it onto papyrus or paper.
Once people invented writing they could work with far more than just seven bits of information at a time. Writing taps into the powers of recognition instead of relying on recall.
Students, parents and educators: Please sign up for Learning Something New, my free newsletter containing articles, videos, study tips and food for thought about learning better and living better, delivered to your inbox twice each month.
Photo by candycanedisco