Figure 910

A schematic outline of the major components of dual-coding theory. The two main symbolic systems—the verbal and non-verbal systems—are connected to distinct input and output systems. Within the two systems are associative structures (involving logogens and imagens) that are linked to one another by referential connections. Reproduced with permission from Mental representations: A dual coding approach, by Allan Paivio, 1986, Oxford University Press.

Consider an experiment in which subjects are given either a set of pictures or a list of words to memorise. If the pictures are of common objects then subjects are likely to name them spontaneously while memorising them (see Paivio, 1971). So, people should encode them using both verbal and non-verbal systems. In contrast, the words are more likely to be memorised using the verbal system alone (assuming that subjects do not spontaneously image the objects referred to by the words). Memory for pictures should, therefore, be better than that for words because of the joint influence of both systems in the former case. Paivio (1971) found that pictures were remembered, in both free-recall and recognition tasks, more readily than words. In fact, pictures are recalled so much more easily than words that Paivio has proposed that the image code is mnemonically superior to the verbal code, although exactly why this should be so is not clear.

These joint effects are not only found for pictures and words. Initial results indicated that they could also be found between different classes of words. Some words are concrete and evoke images more readily than other words. If words are concrete, in the sense of denoting things that can be perceived by one of the sense modalities, rather than abstract, they appear to be retrieved more easily (see Paivio, Yuille, & Madigan, 1968, for evidence of this). As in the case of the picture-word differences, words that are rated as being high in their image-evoking value or concreteness (or both) are likely to be encoded using two codes rather than just one (for reviews of the results of item-memory tasks see Cornoldi & Paivio, 1982; Richardson, 1999). So, again there seems to be a joint contribution to performance when both systems are involved in the task.

However, there is some controversy on the dual-code explanation of recall differences for concrete and abstract words. Part of the problem is that the results are of a correlational nature, they merely show that the imagibility/concreteness of words correlates with good recall performance. They do not show a causal connection between concreteness and recall. We can test for such a causal connection by varying the instructions given to subjects when they are memorising the words. If you employ interactive-imagery instructions (e.g., form images depicting objects interacting in some way), then it is typically found that performance is improved for concrete material but not for abstract materials (see Richardson, 1999). This is perfectly consistent with dual-coding theory because the imagery instructions should involve both coding systems for the concrete words but not for the abstract words.

Unfortunately, similar instructions that do not involve imaging have similar effects; verbal mediation instructions (e.g., form short phrases including the list of items) result in concrete materials being recalled more readily than abstract materials. On the basis of these results, Bower (1970, 1972) proposed that interactive imagery and verbal mediation instructions were both effective in that they increased the organisation and cohesion of the to-be-remembered information. To test this hypothesis, Bower presented subjects with pairs of concrete words using three different types of instructions for different groups: interactive-imagery instructions, separation-imagery instructions (i.e., construct an image of two objects separated in space), or instructions to memorise by rote. On a subsequent cued-recall task, the interactive-imagery subjects performed much better than the separation-imagery subjects, who in turn performed no better than subjects instructed to use rote memorisation. In other words, interactive imagery instructions are effective because they enhance relational organisation. So, recall differences between concrete and abstract words create some difficulties for Paivio's theory. However, we should point out that Paivio has gone some way towards accounting for these results by including organisational assumptions within each of his symbolic systems, which account for differences between interactive-imagery and separation-imagery instructions (see Paivio, 1986, Chapters 4 and 8). Having said this, the issue has not been fully resolved. Recent research has shown that concreteness effects are not due solely to the effects of imagery but may also involve factors like distinctiveness and relational information (see Marschark & Cornoldi, 1990; Marschark & Hunt, 1989; Marschark & Surian, 1992; Plaut & Shallice, 1993). Furthermore, in a review of the literature, Marschark, Richman, Yuille, and Hunt (1987) have rejected the proposal that imaginal codes are stored in long-term memory, arguing instead that verbal and imaginal processing systems operate on a more generic, conceptual memory.

Studies of free recall also support the additivity and functional independence of the two systems (Paivio, 1975; Paivio & Csapo, 1973). In these experiments, subjects were shown a series of concrete nouns and asked to either image to the presented noun or to pronounce it. During the five-second intervals between items they were asked to rate the difficulty of imaging or pronouncing the word. In one manipulation, subjects were presented with a given word repeatedly. In some cases, the repetition encouraged dual coding, in that subjects had to image it on one occurrence and pronounce it on the next. In other cases, the repetition merely promoted encoding in a single code when subjects either imaged or pronounced the word again. After doing this task, without prior warning, subjects were asked to recall the presented words.

Several interesting results were found to support dual-coding theory. First, the probability of imaged words being recalled was twice as high as that for pronounced words, indicating the superiority of nonverbal codes in recall. Second, the imagery-instructions raised the level of recall to the same high level that is normally seen for the encoding of pictures under comparable conditions. Third, in the conditions that

Business Correspondence

Business Correspondence

24 chapters on preparing to write the letter and finding the proper viewpoint how to open the letter, present the proposition convincingly, make an effective close how to acquire a forceful style and inject originality how to adapt selling appeal to different prospects and get orders by letter proved principles and practical schemes illustrated by extracts from 217 actual letter.

Get My Free Ebook


Post a comment