Working Memory: Animals
Animals have good memories. There are anecdotal reports of dogs greeting their masters after years of absence and of bears and elephants squaring old scores after many years. Birds, too, can have remarkable recall. The Clark's nutcracker recovers more than 30,000 pine seeds buried in some 3,000 cache sites in the forest (Vander Wall, 1982). It must remember site locations after snow covers the forest and even burrow at an angle through the snow to arrive at the location. They use new cache sites each year and must overcome potential confusion (i.e., interference) over current and previous sites.
The recall of cache sites for six months certainly qualifies as long-term memory (LTM). But the borderline between long-and short-term memory (STM) is far from clear; in fact, the distinction between STM and LTM may have outlived its usefulness for behavioral analyses of memory (Crowder, 1993). Perhaps a more useful distinction is that between "working" and "reference" memory (Honig, 1978). These terms are free of the theoretical baggage carried by the terms STM/LTM or the similar primary/secondary memory. In the nutcracker example, reference memory would be the skill-learning involved in storing and retrieving seeds (Balda and Kamil, 1992). Working memory would apply to this year's cache sites.
Remarkable though the capacity and duration of nutcracker may be, the underlying processes and mechanisms are what compel the attention of researchers who seek to understand how memory works. Such insight requires laboratory investigations that manipulate independent variables critical to memory. Such work is proceeding briskly along two avenues of research: neurobiological and behavioral.
Investigations of neural mechanisms of memory use simple, quick—almost instantaneous—learning procedures. A repeated application of a stimulus (e.g., tactile) to an organism (e.g., sea slug) can produce habituation—a diminished response (e.g., siphon withdrawal)—or sensitization—a heightened response. This is a primitive form of learning: modification of a response through experience. Since all learning depends on memory, tests of learning are also tests of memory. Neuroplasticity studies have shown changes in cellular mechanisms, membrane channels, second-messengers, and protein syntheses related to memory (Squire and Kandel, 2001).
Another nearly instantaneous learning procedure is fear conditioning, a pairing of two stimuli (classical or Pavlovian conditioning). Rats learn to freeze (remain immobile) when encountering a stimulus paired with electric shock. Fear conditioning has shown that dorsal hippocampal lesions disrupt recent memories of the conditioning context (e.g., test chamber) but not those of the conditioned stimuli (e.g., tones) themselves (Anagnostaras et al., 2001).
Other neurobiological memory procedures require instrumental responses to escape unpleasant situations or to obtain rewards. Rats readily learn the location of a hidden platform in order to rest from swimming in a Morris water maze. Water maze studies have shown that LTP, NMDA receptors, and the dorsal (but not ventral) hippocampus are related to memory (Morris et al., 1986). Rats also readily learn the locations of food in radial-arm mazes. They are very good at remembering, for example, which arms have not been visited on a particular trial and thus still contain rewards. Four-arm "plus" mazes have played a key role in revealing the mnemonic components of the place cells of rats (O'Keefe and Speakman, 1987). Strategies used to solve water and radialarm maze tasks occasionally make interpretation of results complex, and such strategies (i.e., reference memory) can confound measures of working memory. Other memory results raise questions about the universality of the hippocampus in rats' place memory. In delayed nonmatching to sample (DNMS), monkeys choose a stimulus that does not match a previously seen sample. DNMS in monkeys depends less on the hippocampus than place memory in rats. Some researchers are trying to determine whether the hippocampus will prove necessary for general "relational" memory (Cohen and Eichenbaum, 1994) or just place memory. A newer, more sensitive procedure for identifying the role of the hippocampus in nonplace memory is a habituation procedure called preferential viewing (Alvarado and Bachevalier, 2000). Monkeys (or children) view and habituate to a picture, then are tested with that picture presented with a novel picture. The subject viewing mainly the novel picture shows good memory of habituation.
Investigations of behavioral mechanisms of animal memory have traditionally used single-memory items on each trial. Single-memory-item tests may be well suited to neurobiological investigations (e.g., treatment versus normal groups) but by themselves do not reveal much about memory mechanisms (Shettleworth, 1998, p. 257; Olson et al., 1995; Kamil, 1988). Explorations of human memory mechanisms have used list-memory procedures for a long time (Ebbinghaus, 1902). Memory is typically best for items at the beginning (primacy effect) and at the end (recency effect) of each list, producing a U-shaped serial-position function (SPF). Delaying recall (or recognition) eliminates the recency effect. This selective elimination supported the hypothesis that the recency effect represented STM (Glanzer and Cunitz, 1966), and helped stimulate the cognitive revolution of the 1960s. The degree to which animals share similar STM processes with humans became testable only in the 1990s because of the extreme difficulty of training animals in list-memory tasks. Nevertheless, the list memory of rats, pigeons, dolphins, squirrel monkeys, capuchin monkeys, rhesus monkeys, and chimpanzees has been tested.
Two experiments with rhesus monkeys show similarities to human memory processing and expand upon these findings. In a visual-memory experiment, monkeys saw lists of four different "travel-slide" pictures. After a delay (0, 1, 2, 10, 20, or 30 seconds later) following each list, the monkeys indicated whether a test picture presented in a different location was or was not in the list (Wright et al., 1985). Figure 1 (upper portion) shows the changes in the monkeys' four-item serial-position functions for three of the retention delays. Like humans, the monkeys' recency effect dissipated with retention delay. A new finding was the gradual appearance of the primacy effect with delay, and this finding was made possible by using much shorter lists than those used with humans. The primacy and recency effects were separately manipulated and dissociated by retention interval, and the benchmark U-shaped function emerged as a transitional function. Similar changes have been shown for humans (e.g., using four-or five-item lists of snowflake or kaleidoscope patterns), capuchin monkeys, and pigeons (Korsnes, 1995; Neath, 1993; Wright, 1999a; Wright et al., 1985). One issue was whether SPFs for other types of memory (e.g., auditory) would change with delay like those for visual memory.
In an auditory memory experiment, monkeys heard lists of four natural/environmental sounds. After a delay (0, 1, 2, 10, 20, or 30 seconds later) following each list, they had to touch one of two side speakers to indicate whether or not a test sound was in the list (Wright, 1998). As in the visual memory experiment, researchers drew from a large pool of sounds (520 in this case), each one unique to each day's trials. The most striking feature of the auditory memory SPFs (see Figure 1) is that their shape is opposite to that of the visual memory results. Initially, there was no recency effect. The recency effect appeared to lengthen with delay, and the primacy effect dissipated. Thus, evidence about how one memory system works may not reveal much about the workings of other memory systems: vision, auditory, taste, smell, and touch.
Another finding is the counterintuitive one that memory—specifically, visual primacy and auditory recenty—can improve with time. Thus, mechanisms other than memory decay must be responsible for the SPFs. To explore possible mechanisms, researchers manipulated interference among auditory list items. The first list items were shown to interfere (proactively, i.e., forward in time) with the monkey's memory of the last list items (Wright, 1999b). Diminishing or eliminating this interference greatly enhanced memory for the last items, a result that ruled out lack of consolidation as an explanation of poor memory of the last list items. With long delays, the last list items interfered (retroactively, i.e., backward in time) with the monkey's memory of the first list items. Diminishing or eliminating this interference greatly improved memory of the first item, a result that ruled out forgetting (i.e., memory decay) as an explanation for poor memory of the first list items.
Researchers face the challenge of closing the gaps between neurobiological and behavioral approaches to animal and human memory. This interdisciplinary approach will require a blend of molecular, cellular, neurochemical, anatomical, imaging, behavioral, and computational techniques.
Alvarado, M. C., and Bachevalier, J. (2000). Revisiting the maturation of medial temporal lobe memory functions in primates. Learning and Memory 7, 244-256.
Anagnostaras, S. G., Gale, G. D., and Fanselow, M. S. (2001). Hippocampus and contextual fear conditioning: Recent controversies and advances. Hippocampus 11, 8-17.
Balda, R. P., and Kamil, A. C. (1992). Long-term spatial memory in Clark's nutcracker, Nucifraga columbiana. Animal Behaviour 44,761-769.
Cohen, N. J., and Eichenbaum, H. (1994). Memory, amnesia, and the hippocampal system. Cambridge, MA: MIT Press.
Crowder, R. G. (1993). Short-term memory: Where do we stand?Memory & Cognition 21, 142-145.
Ebbinghaus, H. E. (1902). Grundzuge der Psychologie. Leipzig: VonVeit.
Glanzer, M., and Cunitz, A. R. (1966). Two storage mechanisms in free recall. Journal of Verbal Learning and Verbal Behavior 5, 351-360.
Honig, W. K. (1978). Studies of working memory in the pigeon. InS. H. Hulse, H. Fowler, and W. K. Honig, eds., Cognitive processes in animal behavior. Hillsdale, NJ: Erlbaum.
Kamil, A. C. (1988). A synthetic approach to the study of animal intelligence. In D. W. Leger, ed., Nebraska Symposium on Motivation, Vol. 35. Lincoln: University of Nebraska Press.
Korsnes, M. S. (1995). Retention intervals and serial list memory. Perceptual and Motor Skills 80, 723-731.
Morris, R. G. M., Anderson, E., Lynch, G. S., and Baudry, M.(1986). Selective impairment of learning and blockade of long-term potentiation by an N-methyl-D-aspartate receptor antagonist, AP5. Nature 319, 774-776.
Neath, I. (1993). Distinctiveness and serial position effects in recognition. Memory & Cognition 21, 689-698.
O'Keefe, J., and Speakman, A. (1987). Single unit activity in the rat hippocampus during a spatial memory task. Experimental Brain Research 68, 1-27.
Olson, D. J., Kamil, A. C., Balda, R. P., and Mims, P. J. (1995). Performance of four-seed caching corvid species in operant tests of nonspatial and spatial memory. Journal of Comparative Psychology 109,173-181.
Squire, L. R., and Kandel, E. R. (1998). Memory from mind to molecules. New York: W. H. Freeman and Co.
Vander Wall, S. B. (1982). An experimental analysis of cache recovery in Clark's nutcracker. Animal Behaviour 30, 84-94.
Wright, A. A. (1998). Auditory list memory in rhesus monkeys. Psychological Science 9, 91-98.
—— (1999a). Visual list memory in capuchin monkeys (Cebus apella). Journal of Comparative Psychology 113, 74-80.
—— (1999b). Auditory list memory and interference in monkeys. Journal of Experimental Psychology: Animal Behavior Processes 25, 284-296.
Wright, A. A., Santiago, H. C., Sands, S. F., Kendrick, D. F., and Cook, R. G. (1985). Memory processing of serial lists by pigeons, monkeys, and people. Science 229, 287-289.