Retrieval From Semantic Memory

February 29, 2008

When we speak, listen, read, or write, when we give someone directions to a favorite restaurant; or, when we think about a friend, we are retrieving information from semantic memory. The distinction between episodic and semantic memory was proposed by Endel Tulving in 1972; since then close to a hundred studies have focused specifically on semantic memory in context of aging or dementia. The findings are quite clear: Normal aging does not impair semantic memory, while Alzheimer patient’s ability retrieve (remember) this kind of information deteriorates steadily over the course of the disease.Retrieval from semantic memory can be tested in a number of ways.At most basic level, people are simply asked what day it is, where they are how to spell a word, and the like. Some studies have examined retrieval of general knowledge, such as historical or current events,vocabulary, names of famous people, and geography . Laboratory studies typically employ tasks that yield easily quantifiable results, such as naming items from specified categories (e.g. list all the fruits you can think of, list as many U.S. presidents as you can ) or simply naming specific objects or pictures (e.g., a key, a picture of on airplane).

Healthy older adults, as a group, rarely do worse than young adults on the foregoing kinds of tasks. In fact, a number of recent studies have found that older adults often perform significantly better than young adults on semantic memory tasks. For instance, in one study we conducted at Southern Methodist University (with Alan Brown, Todd Jones, and Laura Steen-Patterson) with current students (average age 19) and older alumni (average age 74), the older group’s vocabulary scores were nearly 20 percent above those of the young. Other studies have found that memory for facts and world knowledge (e.g. history, entertainment, politics) also tends to actually increase with age.

In contrast, Alzheimer patients typically experience difficulty retrieving information from semantic memory. Early on, patients simply have problems retrieving specific names, a common memory complaint among healthy individuals as well. Frederick Schmitt and I found that relatively mild Alzheimer patients had more tip-of-the-tongues than age-matched healthy adults (e.g., shown a picture of an artichoke, the patients knew what it was, but couldn’t retrieve the name). Another study by Jacob Huff and his colleagues at Massachusetts Institute of Technology also found naming impairments in Alzheimer patients. However, when given a name recognition test, as in “Is this called an x?” where x was the name of the accompanying line drawing, these patients performed as well as healthy control subjects. This suggests that the Alzheimer patients have not lost their semantic memory; rather, their access to information in semantic memory is impaired. There is further evidence of an access problem. Herbert Weingartner and his colleagues reported that more than two years before the onset of Alzheimer’s individuals in the baltimore Longitudinal study of Aging exhibited difficulty generating less common instances from certain categories such as fruits and vegetables. Even reading reflects this phenomenon. Karolyn Patterson and colleagues at the MRC Applied Psychology Unit in Cambridge, England, found impaired pronunciation of words with atypical spelling-sound correspondence.

A number of studies have analyzed the nature of the naming impairment in Alzheimer’s disease. For some time, it was thought that the problem was primarily perceptual; in other words, the Alzheimer patients could not accurately interpret what an object was, which in turn pre­vented them from retrieving the correct name. For example, Howard Kirshner and his colleagues at Vanderbilt University found that Alzheimer patients made more naming errors for photographs of objects than for the actual objects. On the other hand, the healthy control adults actually made more perceptual errors than the Alzheimer patients did.

One problem is that many naming errors are not exclusively perceptual or semantic. For example, when a picture of a pen is called a “pencil” or a chisel is called an “ice pick,” these names share both perceptual (long, thin, sticklike cylinders) and semantic (writing instruments, tools) attributes. Perceptual Horse that have no apparent semantic component are actually quite rare (such as “bucket” for thimble, “wheel” for button). Like­wise, semantic errors that have no visual or perceptual component (such as “saxophone” for a picture of a harp) are rare as well. However, pure semantic errors that are very common involve producing category super­ordinate names: “worm” for caterpillar, “vegetable” for asparagus, “bird” for penguin, and “bug” for grasshopper. Similarly, semantic errors recalling the function, but not the name, of the object are very common. For example, when asked to name pictured objects, an Alzheimer patient said that sock was “for a foot,” gun was “pull trigger, say bam with that,” and ashtray was a “cigarette dish.” These functional errors tend to be more prominent and occur more often beyond the mild stages of the disease.

In light of the foregoing distinctions, studies by Kathryn Bayles and Cheryl Tomoeda at the University of Arizona and by myself and Frederick Schmitt have led to the conclusion that the source of the naming errors is semantic rather than perceptual. In other words, Alzheimer patients (at least in the mild to moderate range) can perceive objects, but have trouble accessing the specific name. Some investigators have interpreted the impaired access to semantic memory as involving effortful memory processes. In the next section, we will see evidence that implicit memory seems to be spared in the course of both normal aging and Alzheimer’s disease.

Posted in Alzheimer's Disease Disorder