Modeling the effects of irrelevant speech on memory.

The feature model (Nairne, 1990) is extended to account for the effects of irrelevant speech and concomitant interactions in immediate serial recall. In the feature model, both articulatory suppression and irrelevant speech are seen as adding noise to the memory representation, the difference being that articulatory suppression diverts more resources than does irrelevant speech. The addition of noise impairs recall because it reduces the probability of successful redintegration. When a competitor is incorrectly recalled, rather than the correct item, this competitor is recalled out of order, producing an increase in order errors. Six simulations are reported that show that the model accounts for (1) the impairment by both irrelevant speech and articulatory suppression, (2) the irrelevance of the phonological and semantic composition of the irrelevant speech, (3) greater disruption when the irrelevant speech tokens vary, (4) the abolition of the phonological similarity effect for visual, but not for auditory, items, (5) the abolition of the word length effect for both visual and auditory items, and (6) the abolition of the irrelevant speech effect under articulatory suppression for both visual and auditory items. The feature model is compared with the two other major views of irrelevant speech, the phonological store hypothesis and the changing state hypothesis.

Manipulations of irrelevant information: suffix effects with articulatory suppression and irrelevant speech.

Numerous studies have demonstrated impaired recall when the to-be-remembered information is accompanied or followed by irrelevant information. However, no current theory of immediate memory explains all three common methods of manipulating irrelevant information: requiring concurrent articulation, presenting irrelevant speech, and adding a stimulus suffix. Five experiments combined these manipulations to determine how they interact and which theoretical framework most accurately and completely accounts for the data. In Experiments 1 and 2, a list of auditory items was followed by an irrelevant speech sound (the suffix) while subjects engaged in articulatory suppression. Although articulatory suppression reduced overall recall compared to a control condition, comparable suffix effects were seen in both conditions. Experiments 3 and 4 found reliable suffix effects when list presentation was accompanied by irrelevant speech. Experiment 5 found a suffix effect even when the irrelevant speech was composed of a set of different items. Implications for working memory, precategorical acoustic store, the changing-state hypothesis, and the feature model are discussed.

Computer simulations of global memory models.

Computer simulation models of memory--including the so-called global memory models--have had and continue to have a profound impact on current memory research and theory. Unfortunately, no memory textbooks published before 1998 present these models to students. The arguments for and against these types of models are summarized along with reasons why the models should be taught at both the undergraduate and graduate levels. A set of freely available computer programs that can facilitate classroom presentation is then briefly described.

The world wide web: exploring a new advertising environment.

The World Wide Web currently boasts millions of users in the United States alone and is likely to continue to expand both as a marketplace and as an advertising environment. Three experiments explored advertising in the Web environment, in particular memory for ads as they appear in everyday use across the Web. Experiments 1 and 2 examined the effect of advertising repetition on the retention of familiar and less familiar brand names, respectively. Experiment 1 demonstrated that repetition of a banner ad within multiple web pages can improve recall of familiar brand names, and Experiment 2 demonstrated that repetition can improve recognition of less familiar brand names. Experiment 3 directly compared the retention of familiar and less familiar brand names that were promoted by static and dynamic ads and demonstrated that the use of dynamic advertising can increase brand name recall, though only for familiar brand names. This study also demonstrated that, in the Web environment, much as in other advertising environments, familiar brand names possess a mnemonic advantage not possessed by less familiar brand names. Finally, data regarding Web usage gathered from all experiments confirm reports that Web usage among males tends to exceed that among females.

Irrelevant speech eliminates the word length effect.

The word length effect refers to the observation that memory is better for short than for long words. The irrelevant speech effect refers to the finding that memory is better when items are presented against a quiet background than against one with irrelevant speech. According to Baddeley's (1986, 1994) working memory, these variables should not interact: The word length effect arises from rehearsal by the articulatory control process, whereas irrelevant speech reduces recall through interference in the phonological store. Four experiments demonstrate that, like articulatory suppression, irrelevant speech eliminates the word length effect for both visual and auditory items. These results (1) provide further evidence against the ability of working memory to explain the word length and irrelevant speech effects and (2) confirm a specific prediction of Nairne's (1990) feature model.

Modality, concreteness, and set-size effects in a free reconstruction of order task.

Would informing subjects which items were presented on the current list remove effects of presentation modality, concreteness, and set size in a long-term free reconstruction of order task? In Experiment 1, a typical modality effect was found: memory for the final item in a list was enhanced when the item was presented auditorily rather than visually. In Experiment 2, order memory was better for concrete than for abstract items. And in Experiment 3, order memory was better when the same six items were presented on every trial than when a unique set of six items was presented. In all conditions in all experiments, the to-be-remembered items were given to the subject at test. These results suggest that contrary to a popular assumption, the reconstruction of order task does not provide a functionally pure measure of order memory.

The relation between discriminability and memory for vowels, consonants, and silent-center vowels.

People remember lists of vowel-contrasting syllables better than lists that vary only in stop consonant identity. Most views suggest that this difference is due to the structure of immediate memory and the greater discriminability of vowels compared with consonants. In all of these views, there is a presumed systematic relationship between discriminability and recall so that the more discriminable an item, the better that item should be recalled. The 11 experiments reported here measured the relative discriminability of and compared serial recall for (1) intact syllables that varied only in the medial vowel, (2) intact syllables that varied only in the initial consonant, and (3) syllables with the center vowel replaced by silence (so-called silent-center vowels). When item discriminability, as measured by identification, was equated for consonant-contrasting and silent-center lists, serial recall performance was also equal. However, even when the vowels were less discriminable than the consonants or silent-center vowels, serial recall performance for the vowels was still better. These results are problematic for theories based on acoustic discriminability but can be explained parsimoniously by Nairne's (1990) feature model.

Distinctiveness and very short-term serial position effects.

The serial position function reflects better memory for the first and last few items in a list than for the middle items. Four experiments examined the effects of temporal spacing on the serial position function for five-item lists that took between 0.5 seconds and 1.1 seconds to present. As with recall of far longer-lasting lists, recency and other robust serial position effects were observed with both free and serial recall. We demonstrate that temporal schedules of presentation control recall probability in predictable ways, and conclude that very fleeting lists obey similar principles as do longer-lasting lists. We compare both sets of findings with predictions from the dimensional distinctiveness framework.

Remembering and forgetting as context discrimination.

In 1929, H.C. Blodgett reported the results of a seminal maze learning experiment using rats. In that experiment, hungry rats ran in a complex maze but were not rewarded on reaching the goal box. Not surprisingly, the performance of the hungry rats did not improve over trials. However, with the introduction of reward, the error scores of the rats suddenly dropped to the level of the control rats that were rewarded from the outset. This finding indicates that the experimental group had learned the maze despite the absence of reward but that the learning was latent rather than manifest. With Blodgett's findings, the distinction between learning and performance became firmly established, if not as widely appreciated as it might be. Blodgett's (1929) early experimental finding of latent learning could well serve as a paradigm for the approach taken here. That is, we have emphasized the principle that a lack of performance does not necessarily indicate a lack of either learning or memory. This principle is much more than an empty admonition: We have shown it can have a firm theoretical basis, one that has been confirmed repeatedly by experiments cited throughout this paper. That is, it has been shown numerous times that a failure to perform either in a Pavlovian or instrumental learning task or to remember in an animal or human memory task under one set of conditions could be alleviated under another set of conditions. Forgetting was viewed here as a failure of performance resulting from the cues at test retrieving a memory other than the target memory or retrieving no memory at all. According to this view, memory involves discrimination learning. Essentially, memories are stored in the presence of an elaborate set of interoceptive and exteroceptive stimuli, a context. Whether at test the target memory is retrieved depends on how well the cues at test discriminate between the target memory and other memories. This approach suggests that forgetting does not occur: There is only a failure to perform because of a difference between the stimulus conditions prevailing at encoding and at test. It was demonstrated that this approach is at least as reasonable as that which suggests that true forgetting occurs, but certainly more useful. At least three advantages adhere to our view that memory is a discrimination problem. First, in almost numberless cases, it has been shown that failure of performance under one set of stimulus conditions can be alleviated under some other set of stimulus conditions. Second, the proposition that altered stimulus conditions are responsible for forgetting is one of wide generality. Thus, the altered stimulus conditions approach can serve as an explanation not only for various human memory findings but also for various animal memory and learning findings. Finally, and perhaps most importantly, the present approach provides investigators with a powerful and proven working hypothesis. It tells us not to accept failures of performance as indicating an absence of learning or a loss of memory but rather to seek conditions favorable to improving performance, a strategy that should lead to a better fundamental understanding of memory and learning. This position is obviously a type of optimality theory, of which evolutionary theory is one of the more outstanding examples. In optimality theory, any deviation from some ideal state or condition prompts the investigator to seek the reasons for deviation. This approach may prove as successful when applied to learning and memory as it has to other areas of science.

Word-length effects in immediate memory: Overwriting trace decay theory.

Memory is worse for items that take longer to pronounce, even when the items are equated for frequency, number of syllables, and number of phonemes. Current explanations of the word-length effect rely on a time-based decay process within the articulatory loop structure in working memory. Using an extension of Nairne's (1990) feature model, we demonstrate that the approximately linear relationship between span and pronunciation rate can be observed in a model that does not use the concept of decay. Moreover, the feature model also correctly predicts the effects of modality, phonological similarity, articulatory suppression, and serial position on memory for items of different lengths. We argue that word-length effects do not offer sufficient justification for including time-based decay components in theories of memory.

Distinctiveness and serial position effects in recognition.

Digitized photographs of snowflakes were presented for a recognition test after retention intervals of varying durations. While overall accuracy and discrimination remained constant, as the retention interval increased, primacy increased from chance to reliably better than chance while recency decreased to chance levels. A variation of Murdock's (1960) distinctiveness model accounted for the changing primacy and recency effects observed in both between- and within-subjects designs. The generality of the model was examined in two different paradigms: lexical access during sentence processing, and free recall in the continual distractor paradigm. In both cases, the model made accurate qualitative predictions for both latency and accuracy measures.

Schedules of presentation and temporal distinctiveness in human memory.

Recency, in remembering a series of events, reflects the simple fact that memory is vivid for what has just happened but deteriorates over time. Theories based on distinctiveness, an alternative to the multistore model, assert that the last few events in a series are well remembered because their times of occurrence are more highly distinctive than those of earlier items. Three experiments examined the role of temporal and ordinal factors in auditorily and visually presented lists that were temporally organized by distractor materials interpolated between memory items. With uniform distractor periods, the results were consistent with Glenberg's (1987) temporal distinctiveness theory. When the procedure was altered so that distractor periods became progressively shorter from the beginning to the end of the list, the results were consistent for only the visual modality; the auditory modality produced a different and unpredicted (by the theory) pattern of results, thus falsifying the claim that the auditory modality derives more benefit from temporal information than the visual modality. We distinguish serial order information from specifically temporal information, arguing that the former may be enhanced by auditory presentation but that the two modalities are more nearly equal with respect to the latter.

Recency effect in recall of a word list when an immediate memory task is performed after each word presentation.

Subjects studied 12-word lists for free recall. During presentation of the lists, each word was followed by a supraspan sequence of digits, which the subjects tried to reproduce. This task, unlike those used in previous research with this continual distractor procedure, presumably taxed immediate memory capacity to the full. Nevertheless, the word recall data showed a pronounced recency effect. Moreover, the magnitude of the recency effect was found to be just as great with this task as with a more typical task in which the demands on immediate memory are likely to be fewer. These findings reinforce the emerging view that the recency effect need not be the product of immediate memory.